HydrologicalDroughtAssessmentofEnergy-BasedWater … · 2019. 6. 3. · Tsetserleg 47.45°N...
Transcript of HydrologicalDroughtAssessmentofEnergy-BasedWater … · 2019. 6. 3. · Tsetserleg 47.45°N...
Research ArticleHydrological Drought Assessment of Energy-Based WaterDeficit Index (EWDI) at Different Geographical Regions
Chanyang Sur1 Dongkyun Kim 2 Joo-Heon Lee3 Muhammad Mazhar Iqbal 4
and Minha Choi 4
1Drought Research Center Joongbu University Goyang Gyeonggi Republic of Korea2Department of Civil Engineering Hongik University Seoul Republic of Korea3Department of Civil Engineering Joongbu University Goyang Gyeonggi Republic of Korea4Department of Water Resources Graduate School of Water Resources Sungkyunkwan University Suwon Republic of Korea
Correspondence should be addressed to Minha Choi mhchoiskkuedu
Received 29 November 2018 Revised 25 February 2019 Accepted 23 April 2019 Published 29 May 2019
Academic Editor Giacomo Gerosa
Copyright copy 2019 Chanyang Sur et alis is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
is study applied the remote sensing-based drought index namely the Energy-Based Water Deficit Index (EWDI) acrossMongolia Australia and Korean Peninsula for the period between 2000 and 2010 e EWDI is estimated based on the hy-drometeorological variables such as evapotranspiration soil moisture solar radiation and vegetation activity which are derivedfrom the Moderate Resolution Imaging Spectroradiometer (MODIS) imageries e estimated EWDI was compared with theEvaporative Stress Index (ESI) the Vegetation Condition Index (VCI) and the Standardized Precipitation Index (SPI) ecorrelation coefficients between the drought indices are as follows 073ndash076 (EWDI vs ESI) 064ndash071 (EWDI vs VCI) 054ndash064(EWDI vs SPI-3) 069ndash071 (ESI vs VCI) 055ndash062 (ESI vs SPI-3) and 053ndash057 (VCI vs SPI-3)e drought prediction accuracyof each index according to error matrix analysis is as follows 8333ndash9417 (EWDI) 7000ndash9167 (ESI) 4750ndash8500 (VCI)and 6167ndash8833 (SPI-3) Based on the results the EWDI and ESI were found to be more accurate in capturingmoderate droughtconditions than the SPI at different geographical regions
1 Introduction
In general drought is considered from numerous percep-tions Firstly meteorological drought which is usuallyinterpreted by degree of aridness and duration of aridnessand its extent which shows anomalies correspond to cu-mulative precipitation Secondly hydrological drought isrelated to the precipitation deficits on water supply which isquantified by short runoff deepened ground water level andwater resource deficiencies irdly agriculture droughtaccounts for the variable susceptibility of vegetation duringdifferent statuses of vegetation development which is esti-mated by measuring diminution in crop yield and soilmoistness as well as differences among actual and potentialevapotranspiration
Due to these assorted definitions of drought as well astroublesome in estimating the precise commencement range
level and end of drought substantial efforts have been utilizedto delineate techniques for investigation and monitoring ofdrought However the conventional drought has been con-sidered based on the hydrometeorological variables measuredby the network of in situ data tools whereas remote sensingtechnology is robust substituted by providing decisive hy-drometeorological variables for drought analysis at theenormously higher spatial scale than the capability of in situnetwork devices Several studies have presented remotesensing-based drought indices Likewise Kogan [1] in-troduced the Vegetation Condition Index (VCI) by usingremote sensing-based Normalized Difference VegetationIndex (NDVI) data and Kogan [2] developed the VegetationHealth Index (VHI) by using remotely sensed data of TIRimageries to monitor variation in canopy temperaturesAnderson et al [3ndash5] developed a new drought index knownas ldquoEvaporative Stress Indexrdquo (hereafter ESI) ey assessed
HindawiAdvances in MeteorologyVolume 2019 Article ID 8512727 11 pageshttpsdoiorg10115520198512727
the ESI crosswise the globe based on water vapor and tem-perature that is attained from the remote sensing modelnamed ldquoAtmosphere-Land Exchange Inverse (ALEXI) re-mote sensing modelrdquo [3ndash5] ey described that the ESIcorrelated soundly with the Palmer Drought Severity Index(PDSI) and Standardized Precipitation Index (SPI) Mu et al[6] proposed the Drought Severity Index (DSI) which wasbased on MODIS data eir proposed DSI matched well notonly with the Palmer Drought Severity Index (PDSI) andStandardized Precipitation Index (SPI) but also with thevegetation net primary production (NPP) data which des-ignated that the index was useful for assessing drought stimulion crop production and forest growth Keshavarz et al [7]proposed and evaluated a new drought index Soil WaterDeficit Index (SWDI) to study the agricultural drought Hereit is remarkable that almost all of these drought indices focuson specific aspects of various and complex drought conditionsin reality For example SPI ESI EDI and PDSI were esti-mated primarily based on meteorological variables so theseindices did not reflect the level of soil moisture that couldmainly influence crop growth and ecology Moreover VHIand SWDI are mainly estimated based on other variablesrelated to vegetative greening conditions so they cannotaccurately reveal the instant of meteorological phenomenathat can improve drought severity For tenacity of this issueSur et al [8] assessed a progressive drought index namedldquoEnergy-Based Water Deficit Indexrdquo (EWDI) which con-currently considered the circulation of energy water andcarbon across the soil surface and atmosphere to reflect thecomplex conditions of droughts related to atmosphere andvegetationey estimated this index across Korean Peninsulausing MODIS-based datasets and exposed that the EWDIperformed well and showed favorable association with the ESI(correlation coefficient within 073 and 076 based on theirspecific study area) as well as the conventional drought indicessuch as PDSI (correlation coefficient within 057 and 067)and SPI (correlation coefficient within 061 and 064) Asoutcomes of their research were achieved based on the data ofKorean Peninsula which is located on the northeastern brinkof the Asian continent they could not interpret an overallconclusion for the applicability and legitimacy of the EWDIon a wide range of spatial scales across the globe
In this view the main purpose of this research is toenhance the application of EWDI by validating the EWDI atother geographical locations with greater spatial scales thatare prone to drought To attain this goal the EWDI ESIVCI and SPI were estimated across Mongolia (north-centralAsia) the Australian continent and the remaining part ofKorean Peninsula for the duration of 2000ndash2010 Linearregression and error matrices were used to compare esti-mated indices with each other
2 Study Area and Datasets
21 Study Area
211 Mongolia e first study area was the north-centralAsian country Mongolia located between 42ndash51degN (latitude)and 85ndash120degE (longitude) (Figure 1) e total area of
Mongolia is nearly 16 million square kilometers and theelevation ranges from 1000m to 2500m above themean sealevel e country is divided into six types of natural zoneshaving different soil types and plant life in each zone In thisstudy meteorological data are obtained from six selectedstations which cover the entire country e climate ofMongolia is described by a dry and hot summer a long-lasting cold winter high temperature variations low pre-cipitation and a relatively high total of sunny days (onaverage 260 days per year) [9] Mongolia which is relativelya dry region has a less mean annual precipitation accu-mulating to approximately 100ndash200mm in the dry southernmountainous regions and 200ndash350mm in the northernmountainous regions [10] e entire area has a total annualprecipitation about 90mme northern part of Mongolia ismountainous ranges characterized by dense forests in a drysubhumid climate whereas the southern region is the GobiDesert characterized by a drier climate at lower elevations[9] e above-mentioned climatic array as a function oflatitude also described the vegetation pattern athwartMongolia
212 Australia e second study area was Australia locatedbetween 10ndash40degS (latitude) and 113ndash153degE (longitude)(Figure 1)e total area of Australia is 7617930 km2 settingon the Indo-Australian Plate Australia is separated intoeight climate zones which are defined by the Building Codeof Australia (BCA) Based on the local geographical varietiesincluding wind patterns and elevation above the mean sealevel each climate zone is further subdivided into manysubzones
213 Korean Peninsula Korean Peninsula is located on thenortheast brink of Asia at 33ndash43degN (latitude) and 124ndash132degE(longitude) (Figure 1) Even though Korean Peninsula hasbeen previously investigated by Sur et al [8] this studypresents the result of the additional analysis performed inthe north Korean region Korean Peninsula covers an area of219020 km2 located in the Asian monsoon region having amean annual precipitation approximately 1100mm (NorthKorea 9197mm South Korea 13077mm) e topogra-phy of the study area represents an elevation range about0ndash1915m [11] e land use is mainly composed of crop-lands (297) mixed forests (396) deciduous broadleafforests (144) woody savannas (63) and residential andcommercial areas (52) Table 1 describes the geographicalfeatures of meteorological observations
Validations of various drought indices were performedfor the selected three meteorological measurement sites(Hamheung Anju and Kimchaek) e meteorologicalobservation stations and flux tower are presented in Figure 1and the features of each meteorological station are describedin Table 1
22 Datasets In this study to calculate drought indicesinput datasets were obtained from the MODIS satellite andground observation for the duration of 2000ndash2010 e
2 Advances in Meteorology
MODIS multispectral sensor that is the scientific instrumentsent off into the earth circle by NASArsquos Earth ObservingSystem (EOS) was developed to observe the atmosphereland and ocean e temporal resolution is 1 day and thespatial resolutions of the sensor estimations are 1 km 500mand 250m e sensors are found on board the Terra andAqua satellites which were launched in December 1999 andMay 2002 respectively e Terra satellite has an overpasstime around 10 30 PM when ascending and 10 30 AM
when descending e Aqua satellitersquos overpass time isroughly 1 30 PM when ascending and 1 30 AM whiledescending
MODIS has been widely used in the field of energybalance studies since it gives a firm footing for spatiotem-porally continuous information over the entire surface of theglobe [12] MODIS information from the Terra spacecraft(10 30 overpass) is used to estimate ET using variousequations MOD07 having a spatial resolution of 5 km is
Darwinndash10
ndash15
ndash20
ndash25
ndash30
ndash35
ndash40
115 120 125 130 135 140 145 150
Brisbane
Melbourne
AdelaidePerth
Giles
ForestsRice paddy
CroplandsUrban
Temperate(dry summers)Arid
TropicalTemperate(nondry summers)
HamheungAnju
Kimchaek
42degN
39degN
36degN
33degN
42degN
39degN
36degN
33degN
123degE 126degE 129degE 132degE
123degE
95deg0prime0PrimeE 100deg0prime0PrimeE 105deg0prime0PrimeE 110deg0prime0PrimeE 115deg0prime0PrimeE
0 65Projection WGS_1984 Central Merdian 105 (Zone48) Datum WGS_1984 130 260 390 520
Suitability classification maps for roplands of MongoliaA
95deg0prime0Prime
E
90deg0prime0Prime
E
51deg0prime0PrimeN
48deg0prime0PrimeN
45deg0prime0PrimeN
42deg0prime0PrimeN
51deg0prime0PrimeN
48deg0prime0PrimeN
45deg0prime0PrimeN
Integrated evaluation MCDMmethod and Boolean login theory
42deg0prime0PrimeN
100deg
0prime0Prime
E
105deg
0prime0Prime
E
110deg
0prime0Prime
E
115deg
0prime0Prime
E
120deg
0prime0Prime
E
126degE 129degE 132degE
Border of countryBorder of aimagLake
Highly suitableSuitableModerately suitableUnsuitableHighly unsuitableLand use-constraint area
N
EW
S
Figure 1 Geographical locations of study areas
Advances in Meteorology 3
chosen among all current MODIS items given by NASA as itincorporates air and dew point temperatures e MOD07provides instantaneous geophysical variables such as lati-tude longitude dew point and air temperatures surfacepressure solar zenith angle and brightness temperature andtotal perceptible water vapor with moderate resolution[13 14] Together with these geophysical variables air anddew point temperatures were considered a part of this studye MODIS cloud product (MOD06) was used for calcu-lation of radiation under cloudy sky conditions e cloudparameters including cloud fraction cloud top temperatureand cloud optical thickness with 1 km spatial determinationand cloud emissivity with 5 km spatial resolution [15] wereused in this study e sinusoidal projection was imple-mented to the land products to peruse various needs formajor discipline groups Korean Peninsula placed on hor-izontal tile numbers 27 to 28 and vertical tile numbers 4 to 5(H27V04 H27V05 H28V04 and H28V05) in the sinusoidalprojection In the case of Mongolia the HDF tiles areH23V03 H23V04 H24V03 H24V04 H25V03 H25V04H26V03 and H26V04 For Australia the HDF tiles areH27V12 H28V11 H28V12 H29V10 H29V11 H29V12H30V10 H30V11 H30V12 H31V10 H31V11 andH31V12 e MOD13A2 provides NDVI and the EnhancedVegetation Index (EVI) at a temporal resolution of 16 daysand spatial resolution of 1 km [16] and the MOD15A2provides Leaf Area Index (LAI) and the Fraction of Pho-tosynthetically Active Radiation (fPAR) for every 8 dayseMOD17A2 provides vegetation information at 1 km spatialresolution in every 8-day intervals through gross primaryproductivity (GPP) and net primary productivity (NPP)products Hemispherical reflectance (white sky albedo) andbihemispherical reflectance (black sky albedo) were offeredby the MOD43 albedo product Reflected solar radiation wasestimated by using the shortwave infrared band (10th band)of the white sky albedo from the MCD43B3 albedo product[17]
For estimation of SPI the climatological ground mea-surement data were obtained from the weather stationsBecause SPI is calculated from more than 30 years of data wehave chosen a location that provides over 30 years of
precipitation data in Korean Peninsula Australia andMongolia (12 sites from Mongolia httpworldweatherwmointencountryhtmlcountryCodeMNG 32 sites from Aus-tralia httpwwwbomgovauclimatedatastations 60 sitesfrom Korean Peninsula httpwwwkmagokrweatherclimatepast_caljsp) e streamflow data were obtainedfrom Global Land Data Assimilation System (GLDAS) fordrought status validation [18]eGLDASNoah dataset having25km spatial resolution and 1-month estimated data were usedas the ground basis observation for validation purposes
3 Methodology
In this study the following four drought indices wereestimated and compared EWDI ESI SPI-3 and VCI Sinceselected drought indices have different data ranges theywere normalized for more intuitive comparison withEWDI Following sections briefly describe the four droughtindices
31 Evaporative Stress Index (ESI) ESI is calculated by usingAET-PET ratios denoted by fPET
fPET AETPET
(1)
A well-known PriestleyndashTaylor algorithm (Priestley andTaylor 1972) was used for calculations of potential evapo-transpiration (PET) For PET calculation all hydrometeo-rological data were derived from satellite observations Wemodified the algorithm of Cleugh et al [19] and Mu et al[17] which estimates AET based on the followingPenmanndashMonteith equation [20]
λE Δ RN minusG( 1113857 + ρCp esat minus e( 11138571113872 1113873raΔ + c 1 + rs( 1113857ra( 1113857
(2)
where λE is the latent heat flux (Wmiddotmminus2) λ is the latent heatof vaporization (Jmiddotkgminus1) Δ is the slope of the curve relatingthe saturated water vapor pressure to temperature(kPamiddotKminus1) RN is the net radiation flux (Wmiddotmminus2) G is the soilheat flux (Wmiddotmminus2) ρ is the air density (kgmiddotmminus3) CP is the
Table 1 General characteristics of the study sites
ID Latitude Longitude Altitude (m) Temperature (degC) Precipitation (mm)Hamheung 3993degN 12755degE 13 98 8903Anju 3962degN 12565degE 125 92 10720Kimchaek 4067degN 12920degE 540 84 7000Tsetserleg 4745degN 10147degE 1695 05 3360Darkhan 4947degN 10598degE 709 minus06 3090Sainshand 4490degN 11012degE 915 40 1190Dalanzadgad 4358degN 10442degE 1469 46 1270Khovd 4802degN 9165degE 1405 03 1190Murun 4963degN 10017degE 1288 minus13 2070Darwin 1246degS 13105degE 34 274 1694Giles 2503degS 12830degE 598 227 2917Perth 3188degS 11613degE 25 187 8070Brisbane 2748degS 15304degE 8 203 11680Adelaide 3487degS 13887degE 48 164 5360Melbourne 3788degS 14504degE 49 148 6660
4 Advances in Meteorology
specific heat capacity of air (Jmiddotkgminus1middotKminus1) esat is the saturatedwater vapor pressure (Pa) e is the actual water vaporpressure (Pa) ra is the aerodynamic resistance (smiddotmminus1) c isthe psychrometric constant and is set as a constant with avalue of 066 Pa Kminus1 and rs is the surface resistance (smiddotm
minus1)All required parameters of equation (2) were obtained
from satellite observations using the algorithms of Cleughet al [19] and Mu et al [17] In this study the only dif-ference was EVI instead of NDVI because the EVI mightenhance the accuracy of the estimated AET value [17] isstudy enhanced the accuracy of the AET estimation byintroducing the gross primary productivity (GPP) valuesfor the estimation of the surface resistance (rs) e GPPcan be derived from the MOD17 product and it is knownto precisely reflect the impact of photosynthesis which EVIand NDVI cannot reflect [21] To enhance the accuracy ofthe algorithm this study revised the equation for calcu-lating the vegetation cover fraction by the followingequation
Fcci12
EVIi minusEVImin
EVImax minusEVImin+
GPPi minusGPPmin
GPPmax minusGPPmin1113890 1113891 (3)
where Fccistands for the ith dayrsquos vegetation cover fraction
and the subscripts max and min symbolize the maximumand minimum value of all GPP and EVI values attained forall observation periods
is calculated value of vegetation cover fraction issubsequently used as the input of the set of equations [8] toestimate the surface resistance value (rs) to be used inequation (2) Figure 2 shows the comparison among the AETvalues estimated based on the method of this study (y) to thereference flux tower ET value (x) at Cheongmicheon (CFC)and Seolmacheon (SMC) gages located in Korean Peninsulae AET value based on the method of Mu et al [17] isshown for comparison It can be noted that the method ofthis study has greater accuracy compared to the method ofMu et al [17] in terms of correlation coefficient (Figure 2)
Lastly the Evaporative Stress Index (ESI) is obtained bynormalizing the estimated fPET value
ESI fPET minus μfPET
σfPET
(4)
where μfPETand σfPET
represent the mean and the standarddeviation of all fPET values estimated for the entire studyperiod at a given grid cell location
32 Energy-Based Water Deficit Index (EWDI) e waterstatus of the land surface under different conditions can beestimated by considering the EWDI which represents thewater deficit condition Based on apparent thermal inertia(ATI) the EWDI was corporated using the ESI and SoilMoisture Saturation Index (SMSI) e ATI evaluates thespatiotemporal variability of soil moisture and is deriveddirectly from multispectral satellite imageries [22]
Using the differences in land surface temperature(ΔLST) and land surface albedo the ATI is calculated asfollows
ATI 1minus αΔLST
Z(ATI)ij SMSI ATIij minusATImin
ATImax minusATImin
(5)
where α represents the land surface albedo and ΔLST is thediurnal land surface temperature which is the difference oftemperature between the daytime and nighttime Since theATI represents the sum of canopy and soil moisture vari-ability the higher the value the higher the soil water contentof the land surface [22 23] ATI values are normalized byusing SMSI for the purpose of calculating EWDI ATIijrepresents the ATI value at the ith latitude and jth longitude
After calculation of the EWDI the ESI and SMSI termsare differences in the standardized anomaly
EWDIij Z ESIij + SMSIij1113872 1113873ij
(6)
where the EWDI is a dimensionless index ranging frominfinite negative values (drier than normal) to infinitepositive values (wetter than normal) Surface albedo prod-ucts (MCD43B3) in 8 days were used in this study Leaf AreaIndex (MOD15A2) NDVI and EVI (from the MOD13A2product) and atmospheric products (MOD07_L2 atmo-sphere product) were used for calculating EWDI
33 Standardized Precipitation Index (SPI) e SPI wasestablished by McKee et al [24] e SPI is assessed by usingthe monthly average precipitation dataset for a continuousperiod of at least 30 years Because SPI is calculated frommore than 30 years of data we have chosen a location thatprovides over 30 years of monthly precipitation data eSPI uses monthly precipitation aggregated at various timescales (1month 3months 6months 12months etc) Ingeneral gamma fitting function is applied for each dataset todescribe probability interactions e distinction of the SPIis that it does not depend upon the model A straightforwardvaluation of precipitation is the input disparate with thePDSI which makes assumptions about water storage anddeficit
34 Vegetation Condition Index (VCI) e VegetationCondition Index (VCI hereafter) [1] is the most widely usedsatellite-based drought index to monitor vegetation condi-tions VCI is determined based on the Normalized Differ-ence Vegetation Index (NDVI) which assesses live greenvegetation for the observed target VCI is determined usingthe following equation
VCI ZNDVIminusNDVImin
NDVImax minusNDVImin1113890 1113891 (7)
where NDVI NDVImin and NDVImax are the smoothedmonthly normalized difference vegetation index at a givengrid cell location its multiyear maximum and its multi-year minimum respectively and Z is the meaning ofstandardized normalization Since the VCI is an indexvalue which normalizes the value varying between 0 and 1
Advances in Meteorology 5
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x ndash 007
R = 087
Mu et al [17]y = 114x ndash 094
R = 079
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x + 014
R = 090
Mu et al (2007)y = 107x + 013
R = 085
Figure 2 Comparison of the actual evapotranspiration estimated by the method of this study
Table 2 Error matrix result for Korean Peninsula (Hamheung site) during 2001ndash2010
HamheungStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 003137 cfs)ESI
5458 9310 110120 9167Drought 54 4 58No drought 6 56 62Sum 60 60 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
VCI
3553 6604 85120 7083Drought 35 18 53No drought 17 50 67Sum 52 68 120
Table 3 Error matrix result for Australia (Brisbane site) during 2001ndash2010
BrisbaneStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004784 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
2957 5088 91120 7583Drought 29 28 57No drought 1 62 63Sum 30 90 120
6 Advances in Meteorology
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
the ESI crosswise the globe based on water vapor and tem-perature that is attained from the remote sensing modelnamed ldquoAtmosphere-Land Exchange Inverse (ALEXI) re-mote sensing modelrdquo [3ndash5] ey described that the ESIcorrelated soundly with the Palmer Drought Severity Index(PDSI) and Standardized Precipitation Index (SPI) Mu et al[6] proposed the Drought Severity Index (DSI) which wasbased on MODIS data eir proposed DSI matched well notonly with the Palmer Drought Severity Index (PDSI) andStandardized Precipitation Index (SPI) but also with thevegetation net primary production (NPP) data which des-ignated that the index was useful for assessing drought stimulion crop production and forest growth Keshavarz et al [7]proposed and evaluated a new drought index Soil WaterDeficit Index (SWDI) to study the agricultural drought Hereit is remarkable that almost all of these drought indices focuson specific aspects of various and complex drought conditionsin reality For example SPI ESI EDI and PDSI were esti-mated primarily based on meteorological variables so theseindices did not reflect the level of soil moisture that couldmainly influence crop growth and ecology Moreover VHIand SWDI are mainly estimated based on other variablesrelated to vegetative greening conditions so they cannotaccurately reveal the instant of meteorological phenomenathat can improve drought severity For tenacity of this issueSur et al [8] assessed a progressive drought index namedldquoEnergy-Based Water Deficit Indexrdquo (EWDI) which con-currently considered the circulation of energy water andcarbon across the soil surface and atmosphere to reflect thecomplex conditions of droughts related to atmosphere andvegetationey estimated this index across Korean Peninsulausing MODIS-based datasets and exposed that the EWDIperformed well and showed favorable association with the ESI(correlation coefficient within 073 and 076 based on theirspecific study area) as well as the conventional drought indicessuch as PDSI (correlation coefficient within 057 and 067)and SPI (correlation coefficient within 061 and 064) Asoutcomes of their research were achieved based on the data ofKorean Peninsula which is located on the northeastern brinkof the Asian continent they could not interpret an overallconclusion for the applicability and legitimacy of the EWDIon a wide range of spatial scales across the globe
In this view the main purpose of this research is toenhance the application of EWDI by validating the EWDI atother geographical locations with greater spatial scales thatare prone to drought To attain this goal the EWDI ESIVCI and SPI were estimated across Mongolia (north-centralAsia) the Australian continent and the remaining part ofKorean Peninsula for the duration of 2000ndash2010 Linearregression and error matrices were used to compare esti-mated indices with each other
2 Study Area and Datasets
21 Study Area
211 Mongolia e first study area was the north-centralAsian country Mongolia located between 42ndash51degN (latitude)and 85ndash120degE (longitude) (Figure 1) e total area of
Mongolia is nearly 16 million square kilometers and theelevation ranges from 1000m to 2500m above themean sealevel e country is divided into six types of natural zoneshaving different soil types and plant life in each zone In thisstudy meteorological data are obtained from six selectedstations which cover the entire country e climate ofMongolia is described by a dry and hot summer a long-lasting cold winter high temperature variations low pre-cipitation and a relatively high total of sunny days (onaverage 260 days per year) [9] Mongolia which is relativelya dry region has a less mean annual precipitation accu-mulating to approximately 100ndash200mm in the dry southernmountainous regions and 200ndash350mm in the northernmountainous regions [10] e entire area has a total annualprecipitation about 90mme northern part of Mongolia ismountainous ranges characterized by dense forests in a drysubhumid climate whereas the southern region is the GobiDesert characterized by a drier climate at lower elevations[9] e above-mentioned climatic array as a function oflatitude also described the vegetation pattern athwartMongolia
212 Australia e second study area was Australia locatedbetween 10ndash40degS (latitude) and 113ndash153degE (longitude)(Figure 1)e total area of Australia is 7617930 km2 settingon the Indo-Australian Plate Australia is separated intoeight climate zones which are defined by the Building Codeof Australia (BCA) Based on the local geographical varietiesincluding wind patterns and elevation above the mean sealevel each climate zone is further subdivided into manysubzones
213 Korean Peninsula Korean Peninsula is located on thenortheast brink of Asia at 33ndash43degN (latitude) and 124ndash132degE(longitude) (Figure 1) Even though Korean Peninsula hasbeen previously investigated by Sur et al [8] this studypresents the result of the additional analysis performed inthe north Korean region Korean Peninsula covers an area of219020 km2 located in the Asian monsoon region having amean annual precipitation approximately 1100mm (NorthKorea 9197mm South Korea 13077mm) e topogra-phy of the study area represents an elevation range about0ndash1915m [11] e land use is mainly composed of crop-lands (297) mixed forests (396) deciduous broadleafforests (144) woody savannas (63) and residential andcommercial areas (52) Table 1 describes the geographicalfeatures of meteorological observations
Validations of various drought indices were performedfor the selected three meteorological measurement sites(Hamheung Anju and Kimchaek) e meteorologicalobservation stations and flux tower are presented in Figure 1and the features of each meteorological station are describedin Table 1
22 Datasets In this study to calculate drought indicesinput datasets were obtained from the MODIS satellite andground observation for the duration of 2000ndash2010 e
2 Advances in Meteorology
MODIS multispectral sensor that is the scientific instrumentsent off into the earth circle by NASArsquos Earth ObservingSystem (EOS) was developed to observe the atmosphereland and ocean e temporal resolution is 1 day and thespatial resolutions of the sensor estimations are 1 km 500mand 250m e sensors are found on board the Terra andAqua satellites which were launched in December 1999 andMay 2002 respectively e Terra satellite has an overpasstime around 10 30 PM when ascending and 10 30 AM
when descending e Aqua satellitersquos overpass time isroughly 1 30 PM when ascending and 1 30 AM whiledescending
MODIS has been widely used in the field of energybalance studies since it gives a firm footing for spatiotem-porally continuous information over the entire surface of theglobe [12] MODIS information from the Terra spacecraft(10 30 overpass) is used to estimate ET using variousequations MOD07 having a spatial resolution of 5 km is
Darwinndash10
ndash15
ndash20
ndash25
ndash30
ndash35
ndash40
115 120 125 130 135 140 145 150
Brisbane
Melbourne
AdelaidePerth
Giles
ForestsRice paddy
CroplandsUrban
Temperate(dry summers)Arid
TropicalTemperate(nondry summers)
HamheungAnju
Kimchaek
42degN
39degN
36degN
33degN
42degN
39degN
36degN
33degN
123degE 126degE 129degE 132degE
123degE
95deg0prime0PrimeE 100deg0prime0PrimeE 105deg0prime0PrimeE 110deg0prime0PrimeE 115deg0prime0PrimeE
0 65Projection WGS_1984 Central Merdian 105 (Zone48) Datum WGS_1984 130 260 390 520
Suitability classification maps for roplands of MongoliaA
95deg0prime0Prime
E
90deg0prime0Prime
E
51deg0prime0PrimeN
48deg0prime0PrimeN
45deg0prime0PrimeN
42deg0prime0PrimeN
51deg0prime0PrimeN
48deg0prime0PrimeN
45deg0prime0PrimeN
Integrated evaluation MCDMmethod and Boolean login theory
42deg0prime0PrimeN
100deg
0prime0Prime
E
105deg
0prime0Prime
E
110deg
0prime0Prime
E
115deg
0prime0Prime
E
120deg
0prime0Prime
E
126degE 129degE 132degE
Border of countryBorder of aimagLake
Highly suitableSuitableModerately suitableUnsuitableHighly unsuitableLand use-constraint area
N
EW
S
Figure 1 Geographical locations of study areas
Advances in Meteorology 3
chosen among all current MODIS items given by NASA as itincorporates air and dew point temperatures e MOD07provides instantaneous geophysical variables such as lati-tude longitude dew point and air temperatures surfacepressure solar zenith angle and brightness temperature andtotal perceptible water vapor with moderate resolution[13 14] Together with these geophysical variables air anddew point temperatures were considered a part of this studye MODIS cloud product (MOD06) was used for calcu-lation of radiation under cloudy sky conditions e cloudparameters including cloud fraction cloud top temperatureand cloud optical thickness with 1 km spatial determinationand cloud emissivity with 5 km spatial resolution [15] wereused in this study e sinusoidal projection was imple-mented to the land products to peruse various needs formajor discipline groups Korean Peninsula placed on hor-izontal tile numbers 27 to 28 and vertical tile numbers 4 to 5(H27V04 H27V05 H28V04 and H28V05) in the sinusoidalprojection In the case of Mongolia the HDF tiles areH23V03 H23V04 H24V03 H24V04 H25V03 H25V04H26V03 and H26V04 For Australia the HDF tiles areH27V12 H28V11 H28V12 H29V10 H29V11 H29V12H30V10 H30V11 H30V12 H31V10 H31V11 andH31V12 e MOD13A2 provides NDVI and the EnhancedVegetation Index (EVI) at a temporal resolution of 16 daysand spatial resolution of 1 km [16] and the MOD15A2provides Leaf Area Index (LAI) and the Fraction of Pho-tosynthetically Active Radiation (fPAR) for every 8 dayseMOD17A2 provides vegetation information at 1 km spatialresolution in every 8-day intervals through gross primaryproductivity (GPP) and net primary productivity (NPP)products Hemispherical reflectance (white sky albedo) andbihemispherical reflectance (black sky albedo) were offeredby the MOD43 albedo product Reflected solar radiation wasestimated by using the shortwave infrared band (10th band)of the white sky albedo from the MCD43B3 albedo product[17]
For estimation of SPI the climatological ground mea-surement data were obtained from the weather stationsBecause SPI is calculated from more than 30 years of data wehave chosen a location that provides over 30 years of
precipitation data in Korean Peninsula Australia andMongolia (12 sites from Mongolia httpworldweatherwmointencountryhtmlcountryCodeMNG 32 sites from Aus-tralia httpwwwbomgovauclimatedatastations 60 sitesfrom Korean Peninsula httpwwwkmagokrweatherclimatepast_caljsp) e streamflow data were obtainedfrom Global Land Data Assimilation System (GLDAS) fordrought status validation [18]eGLDASNoah dataset having25km spatial resolution and 1-month estimated data were usedas the ground basis observation for validation purposes
3 Methodology
In this study the following four drought indices wereestimated and compared EWDI ESI SPI-3 and VCI Sinceselected drought indices have different data ranges theywere normalized for more intuitive comparison withEWDI Following sections briefly describe the four droughtindices
31 Evaporative Stress Index (ESI) ESI is calculated by usingAET-PET ratios denoted by fPET
fPET AETPET
(1)
A well-known PriestleyndashTaylor algorithm (Priestley andTaylor 1972) was used for calculations of potential evapo-transpiration (PET) For PET calculation all hydrometeo-rological data were derived from satellite observations Wemodified the algorithm of Cleugh et al [19] and Mu et al[17] which estimates AET based on the followingPenmanndashMonteith equation [20]
λE Δ RN minusG( 1113857 + ρCp esat minus e( 11138571113872 1113873raΔ + c 1 + rs( 1113857ra( 1113857
(2)
where λE is the latent heat flux (Wmiddotmminus2) λ is the latent heatof vaporization (Jmiddotkgminus1) Δ is the slope of the curve relatingthe saturated water vapor pressure to temperature(kPamiddotKminus1) RN is the net radiation flux (Wmiddotmminus2) G is the soilheat flux (Wmiddotmminus2) ρ is the air density (kgmiddotmminus3) CP is the
Table 1 General characteristics of the study sites
ID Latitude Longitude Altitude (m) Temperature (degC) Precipitation (mm)Hamheung 3993degN 12755degE 13 98 8903Anju 3962degN 12565degE 125 92 10720Kimchaek 4067degN 12920degE 540 84 7000Tsetserleg 4745degN 10147degE 1695 05 3360Darkhan 4947degN 10598degE 709 minus06 3090Sainshand 4490degN 11012degE 915 40 1190Dalanzadgad 4358degN 10442degE 1469 46 1270Khovd 4802degN 9165degE 1405 03 1190Murun 4963degN 10017degE 1288 minus13 2070Darwin 1246degS 13105degE 34 274 1694Giles 2503degS 12830degE 598 227 2917Perth 3188degS 11613degE 25 187 8070Brisbane 2748degS 15304degE 8 203 11680Adelaide 3487degS 13887degE 48 164 5360Melbourne 3788degS 14504degE 49 148 6660
4 Advances in Meteorology
specific heat capacity of air (Jmiddotkgminus1middotKminus1) esat is the saturatedwater vapor pressure (Pa) e is the actual water vaporpressure (Pa) ra is the aerodynamic resistance (smiddotmminus1) c isthe psychrometric constant and is set as a constant with avalue of 066 Pa Kminus1 and rs is the surface resistance (smiddotm
minus1)All required parameters of equation (2) were obtained
from satellite observations using the algorithms of Cleughet al [19] and Mu et al [17] In this study the only dif-ference was EVI instead of NDVI because the EVI mightenhance the accuracy of the estimated AET value [17] isstudy enhanced the accuracy of the AET estimation byintroducing the gross primary productivity (GPP) valuesfor the estimation of the surface resistance (rs) e GPPcan be derived from the MOD17 product and it is knownto precisely reflect the impact of photosynthesis which EVIand NDVI cannot reflect [21] To enhance the accuracy ofthe algorithm this study revised the equation for calcu-lating the vegetation cover fraction by the followingequation
Fcci12
EVIi minusEVImin
EVImax minusEVImin+
GPPi minusGPPmin
GPPmax minusGPPmin1113890 1113891 (3)
where Fccistands for the ith dayrsquos vegetation cover fraction
and the subscripts max and min symbolize the maximumand minimum value of all GPP and EVI values attained forall observation periods
is calculated value of vegetation cover fraction issubsequently used as the input of the set of equations [8] toestimate the surface resistance value (rs) to be used inequation (2) Figure 2 shows the comparison among the AETvalues estimated based on the method of this study (y) to thereference flux tower ET value (x) at Cheongmicheon (CFC)and Seolmacheon (SMC) gages located in Korean Peninsulae AET value based on the method of Mu et al [17] isshown for comparison It can be noted that the method ofthis study has greater accuracy compared to the method ofMu et al [17] in terms of correlation coefficient (Figure 2)
Lastly the Evaporative Stress Index (ESI) is obtained bynormalizing the estimated fPET value
ESI fPET minus μfPET
σfPET
(4)
where μfPETand σfPET
represent the mean and the standarddeviation of all fPET values estimated for the entire studyperiod at a given grid cell location
32 Energy-Based Water Deficit Index (EWDI) e waterstatus of the land surface under different conditions can beestimated by considering the EWDI which represents thewater deficit condition Based on apparent thermal inertia(ATI) the EWDI was corporated using the ESI and SoilMoisture Saturation Index (SMSI) e ATI evaluates thespatiotemporal variability of soil moisture and is deriveddirectly from multispectral satellite imageries [22]
Using the differences in land surface temperature(ΔLST) and land surface albedo the ATI is calculated asfollows
ATI 1minus αΔLST
Z(ATI)ij SMSI ATIij minusATImin
ATImax minusATImin
(5)
where α represents the land surface albedo and ΔLST is thediurnal land surface temperature which is the difference oftemperature between the daytime and nighttime Since theATI represents the sum of canopy and soil moisture vari-ability the higher the value the higher the soil water contentof the land surface [22 23] ATI values are normalized byusing SMSI for the purpose of calculating EWDI ATIijrepresents the ATI value at the ith latitude and jth longitude
After calculation of the EWDI the ESI and SMSI termsare differences in the standardized anomaly
EWDIij Z ESIij + SMSIij1113872 1113873ij
(6)
where the EWDI is a dimensionless index ranging frominfinite negative values (drier than normal) to infinitepositive values (wetter than normal) Surface albedo prod-ucts (MCD43B3) in 8 days were used in this study Leaf AreaIndex (MOD15A2) NDVI and EVI (from the MOD13A2product) and atmospheric products (MOD07_L2 atmo-sphere product) were used for calculating EWDI
33 Standardized Precipitation Index (SPI) e SPI wasestablished by McKee et al [24] e SPI is assessed by usingthe monthly average precipitation dataset for a continuousperiod of at least 30 years Because SPI is calculated frommore than 30 years of data we have chosen a location thatprovides over 30 years of monthly precipitation data eSPI uses monthly precipitation aggregated at various timescales (1month 3months 6months 12months etc) Ingeneral gamma fitting function is applied for each dataset todescribe probability interactions e distinction of the SPIis that it does not depend upon the model A straightforwardvaluation of precipitation is the input disparate with thePDSI which makes assumptions about water storage anddeficit
34 Vegetation Condition Index (VCI) e VegetationCondition Index (VCI hereafter) [1] is the most widely usedsatellite-based drought index to monitor vegetation condi-tions VCI is determined based on the Normalized Differ-ence Vegetation Index (NDVI) which assesses live greenvegetation for the observed target VCI is determined usingthe following equation
VCI ZNDVIminusNDVImin
NDVImax minusNDVImin1113890 1113891 (7)
where NDVI NDVImin and NDVImax are the smoothedmonthly normalized difference vegetation index at a givengrid cell location its multiyear maximum and its multi-year minimum respectively and Z is the meaning ofstandardized normalization Since the VCI is an indexvalue which normalizes the value varying between 0 and 1
Advances in Meteorology 5
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x ndash 007
R = 087
Mu et al [17]y = 114x ndash 094
R = 079
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x + 014
R = 090
Mu et al (2007)y = 107x + 013
R = 085
Figure 2 Comparison of the actual evapotranspiration estimated by the method of this study
Table 2 Error matrix result for Korean Peninsula (Hamheung site) during 2001ndash2010
HamheungStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 003137 cfs)ESI
5458 9310 110120 9167Drought 54 4 58No drought 6 56 62Sum 60 60 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
VCI
3553 6604 85120 7083Drought 35 18 53No drought 17 50 67Sum 52 68 120
Table 3 Error matrix result for Australia (Brisbane site) during 2001ndash2010
BrisbaneStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004784 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
2957 5088 91120 7583Drought 29 28 57No drought 1 62 63Sum 30 90 120
6 Advances in Meteorology
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
MODIS multispectral sensor that is the scientific instrumentsent off into the earth circle by NASArsquos Earth ObservingSystem (EOS) was developed to observe the atmosphereland and ocean e temporal resolution is 1 day and thespatial resolutions of the sensor estimations are 1 km 500mand 250m e sensors are found on board the Terra andAqua satellites which were launched in December 1999 andMay 2002 respectively e Terra satellite has an overpasstime around 10 30 PM when ascending and 10 30 AM
when descending e Aqua satellitersquos overpass time isroughly 1 30 PM when ascending and 1 30 AM whiledescending
MODIS has been widely used in the field of energybalance studies since it gives a firm footing for spatiotem-porally continuous information over the entire surface of theglobe [12] MODIS information from the Terra spacecraft(10 30 overpass) is used to estimate ET using variousequations MOD07 having a spatial resolution of 5 km is
Darwinndash10
ndash15
ndash20
ndash25
ndash30
ndash35
ndash40
115 120 125 130 135 140 145 150
Brisbane
Melbourne
AdelaidePerth
Giles
ForestsRice paddy
CroplandsUrban
Temperate(dry summers)Arid
TropicalTemperate(nondry summers)
HamheungAnju
Kimchaek
42degN
39degN
36degN
33degN
42degN
39degN
36degN
33degN
123degE 126degE 129degE 132degE
123degE
95deg0prime0PrimeE 100deg0prime0PrimeE 105deg0prime0PrimeE 110deg0prime0PrimeE 115deg0prime0PrimeE
0 65Projection WGS_1984 Central Merdian 105 (Zone48) Datum WGS_1984 130 260 390 520
Suitability classification maps for roplands of MongoliaA
95deg0prime0Prime
E
90deg0prime0Prime
E
51deg0prime0PrimeN
48deg0prime0PrimeN
45deg0prime0PrimeN
42deg0prime0PrimeN
51deg0prime0PrimeN
48deg0prime0PrimeN
45deg0prime0PrimeN
Integrated evaluation MCDMmethod and Boolean login theory
42deg0prime0PrimeN
100deg
0prime0Prime
E
105deg
0prime0Prime
E
110deg
0prime0Prime
E
115deg
0prime0Prime
E
120deg
0prime0Prime
E
126degE 129degE 132degE
Border of countryBorder of aimagLake
Highly suitableSuitableModerately suitableUnsuitableHighly unsuitableLand use-constraint area
N
EW
S
Figure 1 Geographical locations of study areas
Advances in Meteorology 3
chosen among all current MODIS items given by NASA as itincorporates air and dew point temperatures e MOD07provides instantaneous geophysical variables such as lati-tude longitude dew point and air temperatures surfacepressure solar zenith angle and brightness temperature andtotal perceptible water vapor with moderate resolution[13 14] Together with these geophysical variables air anddew point temperatures were considered a part of this studye MODIS cloud product (MOD06) was used for calcu-lation of radiation under cloudy sky conditions e cloudparameters including cloud fraction cloud top temperatureand cloud optical thickness with 1 km spatial determinationand cloud emissivity with 5 km spatial resolution [15] wereused in this study e sinusoidal projection was imple-mented to the land products to peruse various needs formajor discipline groups Korean Peninsula placed on hor-izontal tile numbers 27 to 28 and vertical tile numbers 4 to 5(H27V04 H27V05 H28V04 and H28V05) in the sinusoidalprojection In the case of Mongolia the HDF tiles areH23V03 H23V04 H24V03 H24V04 H25V03 H25V04H26V03 and H26V04 For Australia the HDF tiles areH27V12 H28V11 H28V12 H29V10 H29V11 H29V12H30V10 H30V11 H30V12 H31V10 H31V11 andH31V12 e MOD13A2 provides NDVI and the EnhancedVegetation Index (EVI) at a temporal resolution of 16 daysand spatial resolution of 1 km [16] and the MOD15A2provides Leaf Area Index (LAI) and the Fraction of Pho-tosynthetically Active Radiation (fPAR) for every 8 dayseMOD17A2 provides vegetation information at 1 km spatialresolution in every 8-day intervals through gross primaryproductivity (GPP) and net primary productivity (NPP)products Hemispherical reflectance (white sky albedo) andbihemispherical reflectance (black sky albedo) were offeredby the MOD43 albedo product Reflected solar radiation wasestimated by using the shortwave infrared band (10th band)of the white sky albedo from the MCD43B3 albedo product[17]
For estimation of SPI the climatological ground mea-surement data were obtained from the weather stationsBecause SPI is calculated from more than 30 years of data wehave chosen a location that provides over 30 years of
precipitation data in Korean Peninsula Australia andMongolia (12 sites from Mongolia httpworldweatherwmointencountryhtmlcountryCodeMNG 32 sites from Aus-tralia httpwwwbomgovauclimatedatastations 60 sitesfrom Korean Peninsula httpwwwkmagokrweatherclimatepast_caljsp) e streamflow data were obtainedfrom Global Land Data Assimilation System (GLDAS) fordrought status validation [18]eGLDASNoah dataset having25km spatial resolution and 1-month estimated data were usedas the ground basis observation for validation purposes
3 Methodology
In this study the following four drought indices wereestimated and compared EWDI ESI SPI-3 and VCI Sinceselected drought indices have different data ranges theywere normalized for more intuitive comparison withEWDI Following sections briefly describe the four droughtindices
31 Evaporative Stress Index (ESI) ESI is calculated by usingAET-PET ratios denoted by fPET
fPET AETPET
(1)
A well-known PriestleyndashTaylor algorithm (Priestley andTaylor 1972) was used for calculations of potential evapo-transpiration (PET) For PET calculation all hydrometeo-rological data were derived from satellite observations Wemodified the algorithm of Cleugh et al [19] and Mu et al[17] which estimates AET based on the followingPenmanndashMonteith equation [20]
λE Δ RN minusG( 1113857 + ρCp esat minus e( 11138571113872 1113873raΔ + c 1 + rs( 1113857ra( 1113857
(2)
where λE is the latent heat flux (Wmiddotmminus2) λ is the latent heatof vaporization (Jmiddotkgminus1) Δ is the slope of the curve relatingthe saturated water vapor pressure to temperature(kPamiddotKminus1) RN is the net radiation flux (Wmiddotmminus2) G is the soilheat flux (Wmiddotmminus2) ρ is the air density (kgmiddotmminus3) CP is the
Table 1 General characteristics of the study sites
ID Latitude Longitude Altitude (m) Temperature (degC) Precipitation (mm)Hamheung 3993degN 12755degE 13 98 8903Anju 3962degN 12565degE 125 92 10720Kimchaek 4067degN 12920degE 540 84 7000Tsetserleg 4745degN 10147degE 1695 05 3360Darkhan 4947degN 10598degE 709 minus06 3090Sainshand 4490degN 11012degE 915 40 1190Dalanzadgad 4358degN 10442degE 1469 46 1270Khovd 4802degN 9165degE 1405 03 1190Murun 4963degN 10017degE 1288 minus13 2070Darwin 1246degS 13105degE 34 274 1694Giles 2503degS 12830degE 598 227 2917Perth 3188degS 11613degE 25 187 8070Brisbane 2748degS 15304degE 8 203 11680Adelaide 3487degS 13887degE 48 164 5360Melbourne 3788degS 14504degE 49 148 6660
4 Advances in Meteorology
specific heat capacity of air (Jmiddotkgminus1middotKminus1) esat is the saturatedwater vapor pressure (Pa) e is the actual water vaporpressure (Pa) ra is the aerodynamic resistance (smiddotmminus1) c isthe psychrometric constant and is set as a constant with avalue of 066 Pa Kminus1 and rs is the surface resistance (smiddotm
minus1)All required parameters of equation (2) were obtained
from satellite observations using the algorithms of Cleughet al [19] and Mu et al [17] In this study the only dif-ference was EVI instead of NDVI because the EVI mightenhance the accuracy of the estimated AET value [17] isstudy enhanced the accuracy of the AET estimation byintroducing the gross primary productivity (GPP) valuesfor the estimation of the surface resistance (rs) e GPPcan be derived from the MOD17 product and it is knownto precisely reflect the impact of photosynthesis which EVIand NDVI cannot reflect [21] To enhance the accuracy ofthe algorithm this study revised the equation for calcu-lating the vegetation cover fraction by the followingequation
Fcci12
EVIi minusEVImin
EVImax minusEVImin+
GPPi minusGPPmin
GPPmax minusGPPmin1113890 1113891 (3)
where Fccistands for the ith dayrsquos vegetation cover fraction
and the subscripts max and min symbolize the maximumand minimum value of all GPP and EVI values attained forall observation periods
is calculated value of vegetation cover fraction issubsequently used as the input of the set of equations [8] toestimate the surface resistance value (rs) to be used inequation (2) Figure 2 shows the comparison among the AETvalues estimated based on the method of this study (y) to thereference flux tower ET value (x) at Cheongmicheon (CFC)and Seolmacheon (SMC) gages located in Korean Peninsulae AET value based on the method of Mu et al [17] isshown for comparison It can be noted that the method ofthis study has greater accuracy compared to the method ofMu et al [17] in terms of correlation coefficient (Figure 2)
Lastly the Evaporative Stress Index (ESI) is obtained bynormalizing the estimated fPET value
ESI fPET minus μfPET
σfPET
(4)
where μfPETand σfPET
represent the mean and the standarddeviation of all fPET values estimated for the entire studyperiod at a given grid cell location
32 Energy-Based Water Deficit Index (EWDI) e waterstatus of the land surface under different conditions can beestimated by considering the EWDI which represents thewater deficit condition Based on apparent thermal inertia(ATI) the EWDI was corporated using the ESI and SoilMoisture Saturation Index (SMSI) e ATI evaluates thespatiotemporal variability of soil moisture and is deriveddirectly from multispectral satellite imageries [22]
Using the differences in land surface temperature(ΔLST) and land surface albedo the ATI is calculated asfollows
ATI 1minus αΔLST
Z(ATI)ij SMSI ATIij minusATImin
ATImax minusATImin
(5)
where α represents the land surface albedo and ΔLST is thediurnal land surface temperature which is the difference oftemperature between the daytime and nighttime Since theATI represents the sum of canopy and soil moisture vari-ability the higher the value the higher the soil water contentof the land surface [22 23] ATI values are normalized byusing SMSI for the purpose of calculating EWDI ATIijrepresents the ATI value at the ith latitude and jth longitude
After calculation of the EWDI the ESI and SMSI termsare differences in the standardized anomaly
EWDIij Z ESIij + SMSIij1113872 1113873ij
(6)
where the EWDI is a dimensionless index ranging frominfinite negative values (drier than normal) to infinitepositive values (wetter than normal) Surface albedo prod-ucts (MCD43B3) in 8 days were used in this study Leaf AreaIndex (MOD15A2) NDVI and EVI (from the MOD13A2product) and atmospheric products (MOD07_L2 atmo-sphere product) were used for calculating EWDI
33 Standardized Precipitation Index (SPI) e SPI wasestablished by McKee et al [24] e SPI is assessed by usingthe monthly average precipitation dataset for a continuousperiod of at least 30 years Because SPI is calculated frommore than 30 years of data we have chosen a location thatprovides over 30 years of monthly precipitation data eSPI uses monthly precipitation aggregated at various timescales (1month 3months 6months 12months etc) Ingeneral gamma fitting function is applied for each dataset todescribe probability interactions e distinction of the SPIis that it does not depend upon the model A straightforwardvaluation of precipitation is the input disparate with thePDSI which makes assumptions about water storage anddeficit
34 Vegetation Condition Index (VCI) e VegetationCondition Index (VCI hereafter) [1] is the most widely usedsatellite-based drought index to monitor vegetation condi-tions VCI is determined based on the Normalized Differ-ence Vegetation Index (NDVI) which assesses live greenvegetation for the observed target VCI is determined usingthe following equation
VCI ZNDVIminusNDVImin
NDVImax minusNDVImin1113890 1113891 (7)
where NDVI NDVImin and NDVImax are the smoothedmonthly normalized difference vegetation index at a givengrid cell location its multiyear maximum and its multi-year minimum respectively and Z is the meaning ofstandardized normalization Since the VCI is an indexvalue which normalizes the value varying between 0 and 1
Advances in Meteorology 5
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x ndash 007
R = 087
Mu et al [17]y = 114x ndash 094
R = 079
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x + 014
R = 090
Mu et al (2007)y = 107x + 013
R = 085
Figure 2 Comparison of the actual evapotranspiration estimated by the method of this study
Table 2 Error matrix result for Korean Peninsula (Hamheung site) during 2001ndash2010
HamheungStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 003137 cfs)ESI
5458 9310 110120 9167Drought 54 4 58No drought 6 56 62Sum 60 60 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
VCI
3553 6604 85120 7083Drought 35 18 53No drought 17 50 67Sum 52 68 120
Table 3 Error matrix result for Australia (Brisbane site) during 2001ndash2010
BrisbaneStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004784 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
2957 5088 91120 7583Drought 29 28 57No drought 1 62 63Sum 30 90 120
6 Advances in Meteorology
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
chosen among all current MODIS items given by NASA as itincorporates air and dew point temperatures e MOD07provides instantaneous geophysical variables such as lati-tude longitude dew point and air temperatures surfacepressure solar zenith angle and brightness temperature andtotal perceptible water vapor with moderate resolution[13 14] Together with these geophysical variables air anddew point temperatures were considered a part of this studye MODIS cloud product (MOD06) was used for calcu-lation of radiation under cloudy sky conditions e cloudparameters including cloud fraction cloud top temperatureand cloud optical thickness with 1 km spatial determinationand cloud emissivity with 5 km spatial resolution [15] wereused in this study e sinusoidal projection was imple-mented to the land products to peruse various needs formajor discipline groups Korean Peninsula placed on hor-izontal tile numbers 27 to 28 and vertical tile numbers 4 to 5(H27V04 H27V05 H28V04 and H28V05) in the sinusoidalprojection In the case of Mongolia the HDF tiles areH23V03 H23V04 H24V03 H24V04 H25V03 H25V04H26V03 and H26V04 For Australia the HDF tiles areH27V12 H28V11 H28V12 H29V10 H29V11 H29V12H30V10 H30V11 H30V12 H31V10 H31V11 andH31V12 e MOD13A2 provides NDVI and the EnhancedVegetation Index (EVI) at a temporal resolution of 16 daysand spatial resolution of 1 km [16] and the MOD15A2provides Leaf Area Index (LAI) and the Fraction of Pho-tosynthetically Active Radiation (fPAR) for every 8 dayseMOD17A2 provides vegetation information at 1 km spatialresolution in every 8-day intervals through gross primaryproductivity (GPP) and net primary productivity (NPP)products Hemispherical reflectance (white sky albedo) andbihemispherical reflectance (black sky albedo) were offeredby the MOD43 albedo product Reflected solar radiation wasestimated by using the shortwave infrared band (10th band)of the white sky albedo from the MCD43B3 albedo product[17]
For estimation of SPI the climatological ground mea-surement data were obtained from the weather stationsBecause SPI is calculated from more than 30 years of data wehave chosen a location that provides over 30 years of
precipitation data in Korean Peninsula Australia andMongolia (12 sites from Mongolia httpworldweatherwmointencountryhtmlcountryCodeMNG 32 sites from Aus-tralia httpwwwbomgovauclimatedatastations 60 sitesfrom Korean Peninsula httpwwwkmagokrweatherclimatepast_caljsp) e streamflow data were obtainedfrom Global Land Data Assimilation System (GLDAS) fordrought status validation [18]eGLDASNoah dataset having25km spatial resolution and 1-month estimated data were usedas the ground basis observation for validation purposes
3 Methodology
In this study the following four drought indices wereestimated and compared EWDI ESI SPI-3 and VCI Sinceselected drought indices have different data ranges theywere normalized for more intuitive comparison withEWDI Following sections briefly describe the four droughtindices
31 Evaporative Stress Index (ESI) ESI is calculated by usingAET-PET ratios denoted by fPET
fPET AETPET
(1)
A well-known PriestleyndashTaylor algorithm (Priestley andTaylor 1972) was used for calculations of potential evapo-transpiration (PET) For PET calculation all hydrometeo-rological data were derived from satellite observations Wemodified the algorithm of Cleugh et al [19] and Mu et al[17] which estimates AET based on the followingPenmanndashMonteith equation [20]
λE Δ RN minusG( 1113857 + ρCp esat minus e( 11138571113872 1113873raΔ + c 1 + rs( 1113857ra( 1113857
(2)
where λE is the latent heat flux (Wmiddotmminus2) λ is the latent heatof vaporization (Jmiddotkgminus1) Δ is the slope of the curve relatingthe saturated water vapor pressure to temperature(kPamiddotKminus1) RN is the net radiation flux (Wmiddotmminus2) G is the soilheat flux (Wmiddotmminus2) ρ is the air density (kgmiddotmminus3) CP is the
Table 1 General characteristics of the study sites
ID Latitude Longitude Altitude (m) Temperature (degC) Precipitation (mm)Hamheung 3993degN 12755degE 13 98 8903Anju 3962degN 12565degE 125 92 10720Kimchaek 4067degN 12920degE 540 84 7000Tsetserleg 4745degN 10147degE 1695 05 3360Darkhan 4947degN 10598degE 709 minus06 3090Sainshand 4490degN 11012degE 915 40 1190Dalanzadgad 4358degN 10442degE 1469 46 1270Khovd 4802degN 9165degE 1405 03 1190Murun 4963degN 10017degE 1288 minus13 2070Darwin 1246degS 13105degE 34 274 1694Giles 2503degS 12830degE 598 227 2917Perth 3188degS 11613degE 25 187 8070Brisbane 2748degS 15304degE 8 203 11680Adelaide 3487degS 13887degE 48 164 5360Melbourne 3788degS 14504degE 49 148 6660
4 Advances in Meteorology
specific heat capacity of air (Jmiddotkgminus1middotKminus1) esat is the saturatedwater vapor pressure (Pa) e is the actual water vaporpressure (Pa) ra is the aerodynamic resistance (smiddotmminus1) c isthe psychrometric constant and is set as a constant with avalue of 066 Pa Kminus1 and rs is the surface resistance (smiddotm
minus1)All required parameters of equation (2) were obtained
from satellite observations using the algorithms of Cleughet al [19] and Mu et al [17] In this study the only dif-ference was EVI instead of NDVI because the EVI mightenhance the accuracy of the estimated AET value [17] isstudy enhanced the accuracy of the AET estimation byintroducing the gross primary productivity (GPP) valuesfor the estimation of the surface resistance (rs) e GPPcan be derived from the MOD17 product and it is knownto precisely reflect the impact of photosynthesis which EVIand NDVI cannot reflect [21] To enhance the accuracy ofthe algorithm this study revised the equation for calcu-lating the vegetation cover fraction by the followingequation
Fcci12
EVIi minusEVImin
EVImax minusEVImin+
GPPi minusGPPmin
GPPmax minusGPPmin1113890 1113891 (3)
where Fccistands for the ith dayrsquos vegetation cover fraction
and the subscripts max and min symbolize the maximumand minimum value of all GPP and EVI values attained forall observation periods
is calculated value of vegetation cover fraction issubsequently used as the input of the set of equations [8] toestimate the surface resistance value (rs) to be used inequation (2) Figure 2 shows the comparison among the AETvalues estimated based on the method of this study (y) to thereference flux tower ET value (x) at Cheongmicheon (CFC)and Seolmacheon (SMC) gages located in Korean Peninsulae AET value based on the method of Mu et al [17] isshown for comparison It can be noted that the method ofthis study has greater accuracy compared to the method ofMu et al [17] in terms of correlation coefficient (Figure 2)
Lastly the Evaporative Stress Index (ESI) is obtained bynormalizing the estimated fPET value
ESI fPET minus μfPET
σfPET
(4)
where μfPETand σfPET
represent the mean and the standarddeviation of all fPET values estimated for the entire studyperiod at a given grid cell location
32 Energy-Based Water Deficit Index (EWDI) e waterstatus of the land surface under different conditions can beestimated by considering the EWDI which represents thewater deficit condition Based on apparent thermal inertia(ATI) the EWDI was corporated using the ESI and SoilMoisture Saturation Index (SMSI) e ATI evaluates thespatiotemporal variability of soil moisture and is deriveddirectly from multispectral satellite imageries [22]
Using the differences in land surface temperature(ΔLST) and land surface albedo the ATI is calculated asfollows
ATI 1minus αΔLST
Z(ATI)ij SMSI ATIij minusATImin
ATImax minusATImin
(5)
where α represents the land surface albedo and ΔLST is thediurnal land surface temperature which is the difference oftemperature between the daytime and nighttime Since theATI represents the sum of canopy and soil moisture vari-ability the higher the value the higher the soil water contentof the land surface [22 23] ATI values are normalized byusing SMSI for the purpose of calculating EWDI ATIijrepresents the ATI value at the ith latitude and jth longitude
After calculation of the EWDI the ESI and SMSI termsare differences in the standardized anomaly
EWDIij Z ESIij + SMSIij1113872 1113873ij
(6)
where the EWDI is a dimensionless index ranging frominfinite negative values (drier than normal) to infinitepositive values (wetter than normal) Surface albedo prod-ucts (MCD43B3) in 8 days were used in this study Leaf AreaIndex (MOD15A2) NDVI and EVI (from the MOD13A2product) and atmospheric products (MOD07_L2 atmo-sphere product) were used for calculating EWDI
33 Standardized Precipitation Index (SPI) e SPI wasestablished by McKee et al [24] e SPI is assessed by usingthe monthly average precipitation dataset for a continuousperiod of at least 30 years Because SPI is calculated frommore than 30 years of data we have chosen a location thatprovides over 30 years of monthly precipitation data eSPI uses monthly precipitation aggregated at various timescales (1month 3months 6months 12months etc) Ingeneral gamma fitting function is applied for each dataset todescribe probability interactions e distinction of the SPIis that it does not depend upon the model A straightforwardvaluation of precipitation is the input disparate with thePDSI which makes assumptions about water storage anddeficit
34 Vegetation Condition Index (VCI) e VegetationCondition Index (VCI hereafter) [1] is the most widely usedsatellite-based drought index to monitor vegetation condi-tions VCI is determined based on the Normalized Differ-ence Vegetation Index (NDVI) which assesses live greenvegetation for the observed target VCI is determined usingthe following equation
VCI ZNDVIminusNDVImin
NDVImax minusNDVImin1113890 1113891 (7)
where NDVI NDVImin and NDVImax are the smoothedmonthly normalized difference vegetation index at a givengrid cell location its multiyear maximum and its multi-year minimum respectively and Z is the meaning ofstandardized normalization Since the VCI is an indexvalue which normalizes the value varying between 0 and 1
Advances in Meteorology 5
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x ndash 007
R = 087
Mu et al [17]y = 114x ndash 094
R = 079
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x + 014
R = 090
Mu et al (2007)y = 107x + 013
R = 085
Figure 2 Comparison of the actual evapotranspiration estimated by the method of this study
Table 2 Error matrix result for Korean Peninsula (Hamheung site) during 2001ndash2010
HamheungStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 003137 cfs)ESI
5458 9310 110120 9167Drought 54 4 58No drought 6 56 62Sum 60 60 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
VCI
3553 6604 85120 7083Drought 35 18 53No drought 17 50 67Sum 52 68 120
Table 3 Error matrix result for Australia (Brisbane site) during 2001ndash2010
BrisbaneStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004784 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
2957 5088 91120 7583Drought 29 28 57No drought 1 62 63Sum 30 90 120
6 Advances in Meteorology
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
specific heat capacity of air (Jmiddotkgminus1middotKminus1) esat is the saturatedwater vapor pressure (Pa) e is the actual water vaporpressure (Pa) ra is the aerodynamic resistance (smiddotmminus1) c isthe psychrometric constant and is set as a constant with avalue of 066 Pa Kminus1 and rs is the surface resistance (smiddotm
minus1)All required parameters of equation (2) were obtained
from satellite observations using the algorithms of Cleughet al [19] and Mu et al [17] In this study the only dif-ference was EVI instead of NDVI because the EVI mightenhance the accuracy of the estimated AET value [17] isstudy enhanced the accuracy of the AET estimation byintroducing the gross primary productivity (GPP) valuesfor the estimation of the surface resistance (rs) e GPPcan be derived from the MOD17 product and it is knownto precisely reflect the impact of photosynthesis which EVIand NDVI cannot reflect [21] To enhance the accuracy ofthe algorithm this study revised the equation for calcu-lating the vegetation cover fraction by the followingequation
Fcci12
EVIi minusEVImin
EVImax minusEVImin+
GPPi minusGPPmin
GPPmax minusGPPmin1113890 1113891 (3)
where Fccistands for the ith dayrsquos vegetation cover fraction
and the subscripts max and min symbolize the maximumand minimum value of all GPP and EVI values attained forall observation periods
is calculated value of vegetation cover fraction issubsequently used as the input of the set of equations [8] toestimate the surface resistance value (rs) to be used inequation (2) Figure 2 shows the comparison among the AETvalues estimated based on the method of this study (y) to thereference flux tower ET value (x) at Cheongmicheon (CFC)and Seolmacheon (SMC) gages located in Korean Peninsulae AET value based on the method of Mu et al [17] isshown for comparison It can be noted that the method ofthis study has greater accuracy compared to the method ofMu et al [17] in terms of correlation coefficient (Figure 2)
Lastly the Evaporative Stress Index (ESI) is obtained bynormalizing the estimated fPET value
ESI fPET minus μfPET
σfPET
(4)
where μfPETand σfPET
represent the mean and the standarddeviation of all fPET values estimated for the entire studyperiod at a given grid cell location
32 Energy-Based Water Deficit Index (EWDI) e waterstatus of the land surface under different conditions can beestimated by considering the EWDI which represents thewater deficit condition Based on apparent thermal inertia(ATI) the EWDI was corporated using the ESI and SoilMoisture Saturation Index (SMSI) e ATI evaluates thespatiotemporal variability of soil moisture and is deriveddirectly from multispectral satellite imageries [22]
Using the differences in land surface temperature(ΔLST) and land surface albedo the ATI is calculated asfollows
ATI 1minus αΔLST
Z(ATI)ij SMSI ATIij minusATImin
ATImax minusATImin
(5)
where α represents the land surface albedo and ΔLST is thediurnal land surface temperature which is the difference oftemperature between the daytime and nighttime Since theATI represents the sum of canopy and soil moisture vari-ability the higher the value the higher the soil water contentof the land surface [22 23] ATI values are normalized byusing SMSI for the purpose of calculating EWDI ATIijrepresents the ATI value at the ith latitude and jth longitude
After calculation of the EWDI the ESI and SMSI termsare differences in the standardized anomaly
EWDIij Z ESIij + SMSIij1113872 1113873ij
(6)
where the EWDI is a dimensionless index ranging frominfinite negative values (drier than normal) to infinitepositive values (wetter than normal) Surface albedo prod-ucts (MCD43B3) in 8 days were used in this study Leaf AreaIndex (MOD15A2) NDVI and EVI (from the MOD13A2product) and atmospheric products (MOD07_L2 atmo-sphere product) were used for calculating EWDI
33 Standardized Precipitation Index (SPI) e SPI wasestablished by McKee et al [24] e SPI is assessed by usingthe monthly average precipitation dataset for a continuousperiod of at least 30 years Because SPI is calculated frommore than 30 years of data we have chosen a location thatprovides over 30 years of monthly precipitation data eSPI uses monthly precipitation aggregated at various timescales (1month 3months 6months 12months etc) Ingeneral gamma fitting function is applied for each dataset todescribe probability interactions e distinction of the SPIis that it does not depend upon the model A straightforwardvaluation of precipitation is the input disparate with thePDSI which makes assumptions about water storage anddeficit
34 Vegetation Condition Index (VCI) e VegetationCondition Index (VCI hereafter) [1] is the most widely usedsatellite-based drought index to monitor vegetation condi-tions VCI is determined based on the Normalized Differ-ence Vegetation Index (NDVI) which assesses live greenvegetation for the observed target VCI is determined usingthe following equation
VCI ZNDVIminusNDVImin
NDVImax minusNDVImin1113890 1113891 (7)
where NDVI NDVImin and NDVImax are the smoothedmonthly normalized difference vegetation index at a givengrid cell location its multiyear maximum and its multi-year minimum respectively and Z is the meaning ofstandardized normalization Since the VCI is an indexvalue which normalizes the value varying between 0 and 1
Advances in Meteorology 5
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x ndash 007
R = 087
Mu et al [17]y = 114x ndash 094
R = 079
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x + 014
R = 090
Mu et al (2007)y = 107x + 013
R = 085
Figure 2 Comparison of the actual evapotranspiration estimated by the method of this study
Table 2 Error matrix result for Korean Peninsula (Hamheung site) during 2001ndash2010
HamheungStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 003137 cfs)ESI
5458 9310 110120 9167Drought 54 4 58No drought 6 56 62Sum 60 60 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
VCI
3553 6604 85120 7083Drought 35 18 53No drought 17 50 67Sum 52 68 120
Table 3 Error matrix result for Australia (Brisbane site) during 2001ndash2010
BrisbaneStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004784 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
2957 5088 91120 7583Drought 29 28 57No drought 1 62 63Sum 30 90 120
6 Advances in Meteorology
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x ndash 007
R = 087
Mu et al [17]y = 114x ndash 094
R = 079
20
16
12
8
4
0201612840
ETM
OD
IS (m
md
ay)
ETFlux tower (mmday)
This studyy = 097x + 014
R = 090
Mu et al (2007)y = 107x + 013
R = 085
Figure 2 Comparison of the actual evapotranspiration estimated by the method of this study
Table 2 Error matrix result for Korean Peninsula (Hamheung site) during 2001ndash2010
HamheungStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 003137 cfs)ESI
5458 9310 110120 9167Drought 54 4 58No drought 6 56 62Sum 60 60 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
VCI
3553 6604 85120 7083Drought 35 18 53No drought 17 50 67Sum 52 68 120
Table 3 Error matrix result for Australia (Brisbane site) during 2001ndash2010
BrisbaneStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004784 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
2957 5088 91120 7583Drought 29 28 57No drought 1 62 63Sum 30 90 120
6 Advances in Meteorology
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
approximately 95 percent of the VCI at a given grid celllocation varies from minus2 (harsh drought) to 2 (healthyvegetation condition)
35 Error Matrix Method To correctly detect the droughtevent and severity numerous drought indices were eval-uated using an error matrix method [25 26] An error
Table 4 Error matrix result for Mongolia (Tsetserleg site) during 2001ndash2010
TsetserlegStreamflow
Sum Drought accuracy () Overall accuracy ()Drought No drought
Moderate drought (Streamflowlower quartile 004314 cfs)ESI
4552 8654 100120 8333Drought 45 7 52No drought 13 55 68Sum 58 62 120
EWDI
5457 9474 113120 9417Drought 54 3 57No drought 4 59 63Sum 58 62 120
SPI-3
3956 6964 84120 7000Drought 39 17 56No drought 19 45 64Sum 58 62 120
VCI
1750 3400 74120 6167Drought 17 33 50No drought 13 57 70Sum 30 90 120
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
ndash1Dry
126degE 129degE
2000
2001
20012000(C)
(B)
(A)
126degE
N
S
EW
129degE
42degN
39degN
42degN
39degN
42degN
39degN
42degN
39degN
0
(a) (b)1 (unitless)
Wet
0
200
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
1200
PrecipitationEWDI
Figure 3 (a) Spatial distributions of EWDI for 2000 and 2001 in North Korea (b) temporal distributions of EWDI and precipitation in the(A) Anju (B) Hamheung and (C) Kimchaek sites
Advances in Meteorology 7
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
matrix method is a formed array that consists of drought orwet condition as compared to the category of droughtsuggested by the observation data such as streamflow andsoil moisture When the value of the standardizedstreamflow is less than 0 it is defined as a drought statuse fallouts of matrix are represented to obtain the ac-curacy of drought and the accuracy of drought is the ratioof all observation datasets that are certificated as drought byboth the index and the observation datasets to the totalnumber of drought statuses
4 Results and Discussions
41 Analysis ofDroughtAccuracyUsing ErrorMatrixMethodError matrix derived from GLDAS streamflow dataset inTables 24 showed overall accuracy from 75 to 92 withapproximately 90 accuracy during dry season is resultindicates that all four drought indices (EWDI ESI VCI andSPI-3) have a good degree of reliability for analyzing thedrought status at each site e applicability of the EWDIwas best compared to that of another drought index underdrought conditions For streamflow the EWDI and ESI hadmarkedly better results than did the VCI and SPI-3 withdrought and overall accuracies ranging between 75 and
90 at every study site e patterns of SPI-3 were relativelyslow because it related to the precipitation variation
ese intercomparison fallouts prove the applicability ofthe satellite-based drought indices and the impact of droughton streamflow [25] However some limitations may exist Asnoted by Karnieli et al [27] and Choi et al [25] thevegetation-based drought indices such as the VCI andEvapotranspiration-Based Drought Index (ESI) might beless appropriate for the time and the places where droughtconditions cannot be fully represented by vegetation con-ditions such as dormant season and the area with highlatitudes and elevations For the aforementioned results theEWDI precisely predicts the drought status compared to theSPI-3 e EWDI ESI VCI and SPI-3 were good indices ofextreme drought status
42 Spatiotemporal Patterns of Various Drought IndicesIn Section 42 the spatiotemporal patterns of EWDI andactual drought situations were compared In the case ofNorth Korea several previous studies have reported thatthere has been a serious drought situation from February toMay in 2000 and from March to June in 2001 respectively[28ndash30] Jang et al [28] examined the cases of drought
(a)
(b)
(c)
(e)
(f)
2002
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
115degE 120degE 125degE 130degE 135degE 140degE 145degE 150degE
1200(F)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(A)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(D)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
1200(C)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
PrecipitationEWDI
1200(E)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
200
400
600
800
1000
PrecipitationEWDI
Prec
ipita
tion
(mm
)
1200(B)
1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
0
200
400
600
800
1000
N
S
EW
15degS
20degS
25degS
30degS
35degS
15degS
20degS
25degS
30degS
35degS
Dry Wet(unitless)ndash1 0 1
Figure 4 Spatial distributions of EWDI for 2002 in Australia (central part) temporal distributions of EWDI and precipitation in the (A)Darwin (B) Giles (C) Perth (D) Adelaide (E) Melbourne and (F) Brisbane sites (external parts)
8 Advances in Meteorology
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
damage in North Korea from 2000 to 2001 and showed thespatial distribution of drought damage regions Nam et al[30] reported that severe drought conditions occurred in thewestern part of North Korea in 2000 and in the eastern partof North Korea in 2001
Figure 3 represents the annual mean spatial distributionsof EWDI for 2000 and 2001 in North Korea It also shows thetemporal distribution of EWDI and precipitation in theAnju Hamheung and Kimchaek sites
Comparing the results of this study with those of pre-vious studies it can be seen that a more severe droughtcondition occurred during 2000 than 2001 at Anju which islocated in the western part of North Korea On the contraryin Kimchaek and Hamheung areas located in the easternpart of the country a more severe drought occurred in 2001than in 2000 e main reason of this phenomenon can beexplained by the lack of precipitation In 2000 the amount ofprecipitation in the western region was 20 of that in thenormal year while in 2001 precipitation in the easternregion was only 17 lower than that in the normal year Forthis reason the spatial distribution of drought was showndifferently by year [30]
In the case of Australia Horridge et al [31] reported thatthere has been a serious drought situation from April to
December 2002 Horridge et al [31] showed the cases ofdrought damage in Australia in 2002 and reported the spatialdistribution of drought damage regions
Figure 4 represents the annual mean spatial distribu-tions of EWDI for 2002 in Australia It also shows thetemporal distribution of EWDI and precipitation in theDarwin Giles Perth Adelaide Melbourne and Brisbanesites Among the six validation sites Perth had the lowestannual average precipitation of 688mm in 2002 which was32 lower than that in the normal year [31] e value ofthe drought index for that period indicated the droughtstatus In the case of the Darwin site there was an extremedrought situation from April to August but there was a lotof precipitation during the rest of the year For this reasonthe annual average drought condition was expressedmoderately (Figure 4)
Finally in the case of Mongolia several previous re-searches have reported that there has been a seriousdrought situation from February to October 2001 and fromMarch to December 2002 respectively [32 33] Bayarjargalet al [32] examined the cases of drought damage inMongolia in 2001 and 2002 ey reported that the overalldrought condition was serious in 2001 but in 2002 therewas extreme drought condition in the southern region
N
S
EW
Dry Wet(unitless)ndash1 0 1
2001
2002
(a) (b) (c)
(d)
(e)
(f)
90degE 95degE 100degE 105degE 110degE 115degE 120degE
90degE51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
51degN
48degN
45degN
42degN51degN
48degN
45degN
42degN
95degE 100degE 105degE 110degE 115degE 120degE
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400(A)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
Prec
ipita
tion
(mm
)
0
100
200
300
(D)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
2001 2002
400
Prec
ipita
tion
(mm
)
0
100
200
300
400
Prec
ipita
tion
(mm
)
0
100
200
300
(C)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
(F)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(E)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
2001 2002 400
Prec
ipita
tion
(mm
)
0
100
200
300
(B)
215
105
0ndash05
ndash1ndash15
ndash2
Dro
ught
indi
ces
PrecipitationEWDI
Figure 5 Spatial distributions of EWDI for 2001 and 2002 in Mongolia (central part) temporal distributions of EWDI and precipitation inthe (A) Khovd (B) Murun (C) Darkhan (D) Tsetserleg (E) Dalanzadgad and (F) Sainshand sites (external parts)
Advances in Meteorology 9
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
Davi et al [33] showed the spatial distribution of droughtcondition regions
Figure 5 represents the annual mean spatial distributionsof EWDI for 2001 and 2002 in Mongolia It also shows thetemporal distributions of EWDI and precipitation in theKhovd Murun Darkhan Tsetserleg Dalanzadgad andSainshand sites Among the six validation sites the Dalan-zadgad site had the lowest annual average precipitation of95mm in 2001 and 76mm in 2002 which were 58 lower thanthat in the normal year [32]e value of the drought index forthat period also indicated the drought status In the case of theKhovd andMurun sites there were extreme drought situationsfrom February to April but there was a lot of precipitationduring the rest of the year For this reason the annual averagedrought condition was expressed moderately (Figure 5)
5 Conclusions
In this study conventional and satellite-based drought in-dices were compared over drought-vulnerable sites (KoreanPeninsula Mongolia and Australia) from 2000 to 2010 eEWDI showed the highest drought accuracy through theerrormatrixmethod (drought accuracy ranged from 8571 to9474 Hamheung in North Korea Brisbane in Australiaand Tsetserleg in Mongolia 9474)
e applicability of the EWDI was determined by com-paring the estimated EWDI with actual drought conditionse results of the EWDI and the spatiotemporal distributionof the actual drought situation showed a similar tendency Inthe case of North Korea there has been a serious droughtsituation from February to May in 2000 and from March toJune in 2001 respectively More severe drought conditionoccurred during 2000 than 2001 at the Anju region which islocated in the western part of North Korea On the one handin Kimchaek and Hamheung areas located in the eastern partof the country a more severe drought occurred in 2001 thanin 2000 In the case of Australia there has been a seriousdrought situation from April to December 2002 e mostsevere droughts in Perth were examined because the pre-cipitation was only about 32 lower than that in the normalyear On the other hand Darwin was relatively less droughtprone due to heavy rainfall during the summer season (fromSeptember to March) Finally in the case of Mongolia therehas been a serious drought situation from February to Oc-tober 2001 and from March to December 2002 respectivelyDuring the drought period the most serious droughts oc-curred in the Dalanzadgad region and the less severe droughtconditions in the Khovd and Murun regions
rough the above-mentioned results the applicabilityof EWDI was good compared to that of the other droughtindices Based on the results RS-based indices were iden-tified as good indicators for detecting the drought statusespecially when climate data were not available or weresparsely distributed
Data Availability
e data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
e authors declare that they have no conflicts of interest
Acknowledgments
is research was supported by Basic Science ResearchProgram through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science ICT andFuture Planning (2018R1C1B6008805) and the Ministry ofEducation Science and Technology (Project No NRF-2017R1C1B2003927)
References
[1] F N Kogan ldquoDroughts of the late 1980s in the United Statesas derived from NOAA polar-orbiting satellite datardquo Bulletinof the American Meteorological Society vol 76 no 5pp 655ndash668 1995
[2] F N Kogan ldquoGlobal drought watch from spacerdquo Bulletin ofthe American Meteorological Society vol 78 no 4 pp 621ndash636 1997
[3] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 1 Model formulationrdquo Journal ofGeophysical Research Atmospheres vol 112 no 10 2007
[4] M C Anderson J Norman J Mecikalski J Otkin andW P Kustas ldquoA climatological study of evapotranspirationand moisture stress across the continental United States basedon thermal remote sensing 2 Surface moisture climatologyrdquoJournal of Geophysical Research vol 112 no 11 2007
[5] M C Anderson C Hain B Wardlow A PimsteinJ R Mecikalski and W P Kustas ldquoEvaluation of droughtindices based on thermal remote sensing of evapotranspira-tion over the continental United Statesrdquo Journal of Climatevol 24 no 8 pp 2025ndash2044 2011
[6] Q Mu M Zhao J S Kimball N G McDowell andS W Running ldquoA remotely sensed global terrestrial droughtseverity indexrdquo Bulletin of the American Meteorological So-ciety vol 94 no 1 pp 83ndash98 2013
[7] M R Keshavarz M Vazifedoust and A Alizadeh ldquoDroughtmonitoring using a soil wetness deficit index (SWDI) derivedfromMODIS satellite datardquo Agricultural Water Managementvol 132 pp 37ndash45 2014
[8] C Sur J Hur K Kim W Choi and M Choi ldquoAn evaluationof satellite-based drought indices on a regional scalerdquo In-ternational Journal of Remote Sensing vol 36 no 22pp 5593ndash5612 2015
[9] P Batima Climate Change Vulnerability and Adaptation inthe Livestock Sector of Mongolia A Final Report SubmittedAssessments of Impacts and Adaptations to Climate Change(AIACC) e International Start Secretariat WashingtonDC USA 2006
[10] T Sato F Kimura and A Kitoh ldquoProjection of globalwarming onto regional precipitation over Mongolia using aregional climate modelrdquo Journal of Hydrology vol 333 no 1pp 144ndash154 2007
[11] K Hwang M Choi S O Lee and J-W Seo ldquoEstimation ofinstantaneous and daily net radiation from MODIS dataunder clear sky conditions a case study in East Asiardquo Irri-gation Science vol 31 pp 1173ndash1184 2013
10 Advances in Meteorology
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
[12] B Tang and Z Li ldquoEstimation of instantaneous net surfacelongwave radiation from MODIS cloud-free datardquo RemoteSensing of Environment vol 112 no 9 pp 3482ndash3492 2008
[13] W P Menzel S W Seemann J Li and L E GumleyMODISAtmospheric Profile Retrieval Algorithm eoretical BasisDocument (MOD07) University of Winconsin-MadisonPress Madison WI USA 2002
[14] S W Seemann J Li W P Menzel and L E GumleyldquoOperational retrieval of atmospheric temperature moistureand ozone from MODIS infrared radiancesrdquo in Journal ofApplied Meteorology vol 42 pp 1072ndash1091 2003
[15] G Bisht and R L Bras ldquoEstimation of net radiation from theMODIS data under all sky conditions southern great plainscase studyrdquo Remote Sensing of Environment vol 114pp 1522ndash1534 2010
[16] A R Huete K Didan T Miura E P Rodriguez X Gao andL G Ferreira ldquoOverview of the radiometric and biophysicalperformance of the MODIS vegetation indicesrdquo RemoteSensing of Environment vol 83 no 1-2 pp 195ndash213 2002
[17] Q Mu F A Heinsch M Zhao and S W Running ldquoDe-velopment of a global evapotranspiration algorithm based onMODIS and global meteorology datardquo Remote Sensing ofEnvironment vol 111 no 4 pp 519ndash536 2007
[18] M Rodell P R Houser U Jambor et al ldquoe global land dataassimilation systemrdquo Bulletin of the American MeteorologicalSociety vol 85 no 3 pp 381ndash394 2004
[19] H A Cleugh R Leuning Q Mu and S W RunningldquoRegional evaporation estimates from flux tower and MODISsatellite datardquo Remote Sensing of Environment vol 106 no 3pp 285ndash304 2007
[20] J LMonteith ldquoEvaporation and environmentrdquo Symposia of theSociety for Experimental Biology vol 19 pp 205ndash234 1965
[21] W Yuan S Liu G Yu et al ldquoGlobal estimates of evapo-transpiration and gross primary production based on MODISand global meteorology datardquo Remote Sensing of Environ-ment vol 114 no 7 pp 1416ndash1431 2010
[22] W W Verstraeten F Veroustraete C J van der SandeI Grootaers and J Feyen ldquoSoil moisture retrieval usingthermal inertia determined with visible and thermal space-borne data validated for European forestsrdquo Remote Sensing ofEnvironment vol 101 no 3 pp 299ndash314 2006
[23] T Y Chang Y C Wang C C Feng A D ZieglerT W Giambelluca and Y A Liou ldquoEstimation of root zonesoil moisture using apparent thermal inertia with MODISimagery over a tropical catchment in Northern ailandrdquoIEEE Journal of Selected Topics in Applied Earth Observationsand Remote Sensing vol 5 no 3 pp 752ndash761 2012
[24] T B McKee N J Doesken and J Kleist ldquoe relationship ofdrought frequency and duration to time scalesrdquo in Pro-ceedings of the Preprints of the 8th Conference on AppliedClimatology vol 17ndash22 pp 179ndash184 Anaheim CA USAJanuary 1993
[25] M Choi J M Jacobs M C Anderson and D D BoschldquoEvaluation of drought indices via remotely sensed data withhydrological variablesrdquo Journal of Hydrology vol 476pp 265ndash273 2013
[26] R G Congalton ldquoA review of assessing the accuracy ofclassifications of remotely sensed datardquo Remote Sensing ofEnvironment vol 37 no 1 pp 35ndash46 1991
[27] A Karnieli N Agam and R T Pinker ldquoUse of NDVI andland surface temperature for drought assessment merits andlimitationsrdquo Journal of Climate vol 23 no 3 pp 618ndash6332010
[28] M W Jang S H Yoo and J Y Choi ldquoAnalysis of springdrought using NOAAAVHRR NDVI for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 49no 6 pp 21ndash33 2007
[29] S U Kang and J W Moon ldquoDrought analysis using SC-PDSIand derivation of drought severity-duration-frequency curvesin North Koreardquo Journal of Korea Water Resources Associ-ation vol 47 no 9 pp 813ndash824 2014
[30] W H Nam S H Yoo M W Jang and J Y Choi ldquoAp-plication of meteorological drought indices for North KoreardquoJournal of the Korean Society of Agricultural Engineers vol 50no 3 pp 3ndash15 2008
[31] M Horridge J Madden and G Wittwer ldquoUsing a highlydisaggregated multi-regional single-country model to analysethe impacts of the 2002-03 drought on Australiardquo Journal ofPolicy Modelling vol 27 pp 285ndash308 2005
[32] A Bayarjargal A Karnieli M Bayasgalan S KhudulmurC Gandusha and T J Tucker ldquoA comparative study ofNOAAndashAVHRR derived drought indices using change vectoranalysisrdquo Remote Sensing of Environment vol 105 no 1pp 9ndash22 2006
[33] N K Davi G C Jacoby R D DrsquoArrigo N BaatarbilegL Jinbao and A E Curtis ldquoA tree-ring-based drought indexreconstruction for far-Western Mongolia 1565ndash2004rdquo In-ternational Journal of Climatology vol 29 no 10 pp 1508ndash1514 2009
Advances in Meteorology 11
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom
Hindawiwwwhindawicom Volume 2018
Journal of
ChemistryArchaeaHindawiwwwhindawicom Volume 2018
Marine BiologyJournal of
Hindawiwwwhindawicom Volume 2018
BiodiversityInternational Journal of
Hindawiwwwhindawicom Volume 2018
EcologyInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Applied ampEnvironmentalSoil Science
Volume 2018
Forestry ResearchInternational Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Geophysics
Environmental and Public Health
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
International Journal of
Microbiology
Hindawiwwwhindawicom Volume 2018
Public Health Advances in
AgricultureAdvances in
Hindawiwwwhindawicom Volume 2018
Agronomy
Hindawiwwwhindawicom Volume 2018
International Journal of
Hindawiwwwhindawicom Volume 2018
MeteorologyAdvances in
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018Hindawiwwwhindawicom Volume 2018
ChemistryAdvances in
ScienticaHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Geological ResearchJournal of
Analytical ChemistryInternational Journal of
Hindawiwwwhindawicom Volume 2018
Submit your manuscripts atwwwhindawicom