CGMS Anhui & Yield estimation with RS CGMS Anhui & Yield estimation with RS.

CGMS Anhui & Yield estimation with RSCGMS Anhui & Yield estimation with RS

Our workOur work

We participate the following work package

WP 21 Ground data collection

WP 24 CGMS pilot in CHINA

WP 41 Official yield data collection

WP 44 Wheat Yield estimation based on remote

sensing for HUAIBEI Plain

WP 7 Networking and Sustainable partnership

Our workOur work

Organizing a CGMS workshop in China, 2011

Part 1 Part 1 CGMS-AnhuiCGMS-Anhui

Level 1Level 1

Weather Station

Level 1Level 1

Update the METDATA

The data from meteorological department (Archive data, from 1990 to 2012) The data from NOAA GSOD, now we can download the real-time data from the NOAA GSOD FTP everyday.

Level 1Level 1

Interpolation Grid Weather

The batch model give us a easy way to interpolation weather

Level 1Level 1

Grid Weather (Average daily temperature, 31/12/2012 )

Level 2Level 2

Crop simulation- using the batch model 1.Calculate the crop yield2. Aggregation the grid yield

Level 3Level 3

Yield Forecast 1.Aggregation the Nuts yield2. Prepare for forecast

Level 3Level 3

CGMS Statistical Tool

Level 3Level 3

CGMS Statistical Tool But only use the potential yield storage to estimate the crop yield, the result is not very good

Resent works and Resent works and further worksfurther works

Update the CGMS dataset

Prepared some NDVI, fAPAR and DMP data, plan to add these data into CST

Integrate the CGMS Anhui (further work)

Part2 Part2 Prediction of Wheat Yields Using Prediction of Wheat Yields Using Multiple Linear Regression Models Multiple Linear Regression Models

in the Huaibei Plain of Chinain the Huaibei Plain of China

Beier Zhang (AIER - China )Qinhan Dong (VITO - Belgium)

ContentsContents

Study area

Phenology

Trends of yields

Data sets and methods

Results of prediction

Validation

Discussions

Study areaStudy area

Huaibei Plain (include 6 prefectures)

Area:64154 km2

Arable area: 20905 km2

Main soil type :Cambosols & VertisolsMain crop type: Winter wheat & Maize

PhenologyPhenology

SowingEmergence

Tiller

Wintering period

Turning green

Jointing Heading Maturity

Harvest

10/12 10/19 12/1 12/20 2/10 3/10 4/22 5/15 6/1

Wheat: October to next year June

Maize or soybeans: June to October

Trends of yieldsTrends of yieldsThere are significant yearly trend of wheat yield in every prefectures from 2000 to 2011, so the trend must be considered the trend must be considered in the prediction

Data setsData sets

I. Biophysical variables based on RS: using SPOT-VGT

Ten-daily series : every dekad from 1999 to 2012

Variables: Smoothed k-NDVI

Building data sets of RS:The cumulative NDVI for all possible combinations (at

least 2, at most 9, because the one phenological stage is less than 3 month) of consecutive dekads within the wheat growing period (2nd dekad of Oct to 3rd dekad of Jun).

Data setsData sets

For example

Year O2 O3 …… O2O3 O3N1 …… O2N1 O3N2 ……

2000200120022003200420052006200720082009201020112012

Data setsData sets

III. Meteorology data sets

Variables: include rainfall, temperature and duration of sunshine, from 1999 to 2012

Interpolation method: CGMS Level-1 give us the values of every grid (25km x 25km) in the study area.

Calculate average values in every prefecture

Building data sets of Meteorology data sets:The average rainfall, temperature and solar radiation of

every phonelogical stage of wheat in every prefecture.

MethodsMethods

Detrend method.We use two different methods

1.Add year as a variables into the model.2.Separate the trend yield from real yield, and build the regression model with ΣNDVI and residual error

First predicting the yield using regression to obtain the inter-annual trend (PT)

Calculate the residual error (official yield - PT )Using ΣNDVI and meteorology data to predicting residual

error(PR)PT+PR

MethodsMethods

Precision validationLeave-one-out (leave one year data out; regression model

building using the rest of data to predict the left year; corellating the official yield with the predicted ones)

ResultsResults

Prefecture

Models

R2

Absolute Error

Constant year ΣNDVI Meteorology

Bengbu +0.645 +0.101*year +2.099*O2N2+1.152*F2F3 0.630 0.372

Bozhou +3.414 +0.244*year +2.417*N1 0.823 0.224

Fuyang -0.156 +0.236*year +4.606*O3N1 +0.257*Sg 0.822 0.278

Huaibei +0.101 +0.174*year +0.621*J3M3 -0.284*Sj 0.831 0.363

Huainan +0.854 +0.305*year +0.287*Sg+0.296*Em 0.838 0.322

Suzhou +2.118 +0.151*year +0.527*J2M3 +0.064*Es 0.778 0.277

Regression models Using year ， k-NDVI ， and Meteorology Data

Results-detrendResults-detrend

Prefecture Trend Model R2

Bengbu 3.580+0.214*year 0.667

Bozhou 3.97+0.275*year 0.912

Fuyang 3.765+0.218*year 0.741

Huaibei 3.985+0.239*year 0.838

Huainan 2.859+0.283*year 0.751

Suzhou 3.900+0.196*year 0.754

Trend

Results-detrendResults-detrend

PrefectureModels

R2

Constant ΣNDVI Meteorology

Bengbu -7.178 1.724*M2+0.721*Sj+0.201*Em+0.293*Ee

+0.129*Sm0.667

Bozhou -2.304+0.279*Sm+0.101*Em+0.

298*Ew0.753

Fuyang -3.720 +4.575*O3N1 +0.244*Sg 0.681

Huaibei -1.147 +0.01*J3M3 +0.263*Sm-0.152*Se 0.585

Huainan -2.877 +3.047*N1 +0.328*Sg+0.283*Em 0.736

Suzhou -257 +3.135*O2+1.414*M3 +0.331*Ee 0.592

Regression models after detrend

Validation Validation

Using Jack-knife method, comparing absolute error of different methods

Prefecture Year & k-NDVI &Meteorologyk-NDVI & Meteorology

after Detrend

Bengbu 0.372 0.167

Bozhou 0.224 0.159

Fuyang 0.278 0.238

Huaibei 0.363 0.289

Huainan 0.322 0.276

Suzhou 0.277 0.27

Mean Error 0.306 0.233


After detrend Bengbu


After detrend Bozhou


After detrend Fuyang


After detrend Huaibei


After detrend Huainan


After detrend Suzhou


An example ,If we want to estimate the yield of 2012.

Building the trend model using the data from 2000 to 2011Calculating residual error .Building the model using the above variables.Then Calculating the yield.


The result of year 2012.

Prefecture Real yield Estimate yield absolute error

Bengbu 5.94 6.21 0.27

Bozhou 7.53 7.64 0.11

Fuyang 6.48 6.76 0.28

Huaibei 7.17 7.58 0.41

Huainan 6.02 5.93 0.33

Suzhou 6.31 6.79 0.47

Mean Error 0.31

Discussions Discussions

The method We think the method using k-NDVI & Meteorology

after detrend is better This method consider the fact of yield trend, RS and

Meteorology. The average error of six prefecture in Huaibei Plain is

about 0.233 ton per ha, this is a quite good result.

Discussions Discussions

Suggestion for further study Add these data into CST Add a new dataset from CGMS Level2 Do some field work, get the real crop yield about the

field level, then build the model of this level. This work I think can adjust our method and make the result more accurately.

CGMS Anhui & Yield estimation with RS CGMS Anhui & Yield estimation with RS.

Documents

Transcript of CGMS Anhui & Yield estimation with RS CGMS Anhui & Yield estimation with RS.