B13E 0583M i l k i CH CO d H O fl ith th di t diff i th d

B13E-0583Measuring lake-air CH CO and H O fluxes with the gradient diffusion methodMeasuring lake-air CH4, CO2 and H2O fluxes with the gradient diffusion methodg 4, 2 2 g

W i Xi 1 X h i L 1 2 Sh d Li 1 W i W 1 H h Li1 Qit Xi 1Wei Xiao1 Xuhui Lee1,2 Shoudong Liu1 Wei Wang1 Hanchao Li1 Qitao Xiao1Wei Xiao , Xuhui Lee , , Shoudong Liu , Wei Wang , Hanchao Li , Qitao Xiao, , g , g , , Q1Yale NUIST Center on Atmospheric Environment Nanjing University of Information Science & Technology Nanjing 210044 China; wei xiao@nuist edu cn1Yale-NUIST Center on Atmospheric Environment, Nanjing University of Information Science & Technology, Nanjing 210044, China; wei.xiao@nuist.edu.cnYale NUIST Center on Atmospheric Environment, Nanjing University of Information Science & Technology, Nanjing 210044, China; wei.xiao@nuist.edu.cn

2School of Forestry and Environmental Studies Yale University New Haven Connecticut 06511 USA2School of Forestry and Environmental Studies, Yale University, New Haven, Connecticut 06511, USASchool of Forestry and Environmental Studies, Yale University, New Haven, Connecticut 06511, USA

I t d tiIntroductionIntroduction In order to measure the smallG di t t In order to measure the small Gradient measurement

Lakes are an important component of the global greenhouse gas fl t th l k t hGradient measurement 2384pb Lakes are an important component of the global greenhouse gas flux at the lake-atmosphere2384

ppp p g g gcycles Several methods are available for the measurement of lake air

flux at the lake atmosphere pcycles. Several methods are available for the measurement of lake-air interface using the GD method4,cyc es Se e a et ods a e a a ab e o t e easu e e t o a e afl f h h fl ti t ti h b t

interface using the GD method, Filt H

fluxes of greenhouse gases such as floating static chamber water we reduced turbulentFilter 2376Cfluxes of greenhouse gases, such as floating static chamber, water we reduced turbulent 2376

equilibrium method eddy covariance (EC) and gradient diffusion (GD) fl t ti ith b ff lequilibrium method, eddy covariance (EC) and gradient diffusion (GD). fluctuations with buffer volumes424m

q , y ( ) g ( ) fluctuations with buffer volumes Flow meterFilterGas intake 424m

M i th l k i h fl i h ll i f l and minimized the transientFlow meterFilter 424

pp Measuring the lake-air greenhouse gas fluxes is challenging for several and minimized the transient 1 5L/ i

p Measuring the lake air greenhouse gas fluxes is challenging for severaltime between valve switching

1.5L/min 2,

reasons including small signal to noise ratios bubble ebullition that time between valve switching BypasGas intake Solenoid 423O2

reasons, including small signal to noise ratios, bubble ebullition that gi ll d h t t b

Bypass

Gas intake Solenoid 423

COg g

evades equilibrium method detection large density corrections on the using small and short tubesFlow meters

Pump

C

evades equilibrium method detection, large density corrections on the using small and short tubes. Flow meter Pumpevades equilibrium method detection, large density corrections on theEC fl d i f i h h h b h d 1 5L/min 3.7

%v

EC flux and wave interference with the chamber method 1.5L/min 3.7

%EC flux, and wave interference with the chamber method. A 3 l it h d4L O

, %

A 3-way valve switched4LB ff B ff O

In this study a trace gas analyzer based on the cavity ring downA 3 way valve switched

AnalyzerBuffer Buffer3 5H

2

In this study, a trace gas analyzer based on the cavity ring-down between the two intakes everyAnalyzer0.2L/min 3.5Hy, g y y g

t (CRDS) l d i th GD d t thbetween the two intakes every 0.2L/min

spectroscopy (CRDS) was employed in the GD mode to measure the 60 s and the measurement11 53 43 11 55 43 11 57 44 11 59 47 12 01 50spectroscopy (CRDS) was employed in the GD mode to measure the 60 s and the measurement 11:53:43 11:55:43 11:57:44 11:59:47 12:01:50H O CO and CH fluxes at a lake atmosphere interface h d t d t t i lTi DOY 227H2O, CO2 and CH4 fluxes at a lake-atmosphere interface. approached steady state in lessTime, DOY 2272 2 4 p approached steady state in less ,

O t d t b th fi t tt t t i th fl than 10 s after each switchingSchematic design diagram of the gradient measurement system Step changes in the CO CH and H O mixing ratios in response to valve switching Our study appears to be the first attempt at measuring these fluxes than 10 s after each switching. Schematic design diagram of the gradient measurement system Step changes in the CO2, CH4 and H2O mixing ratios in response to valve switching Our study appears to be the first attempt at measuring these fluxessimultaneously using one instrument and in a long term uninterruptedsimultaneously using one instrument and in a long-term uninterruptedy g g poperationoperation.ope a o

Latent heat and CO flux comparison The goal of this paper is to present a performance evaluation of this Zero-gradient test Latent heat and CO2 flux comparison The goal of this paper is to present a performance evaluation of this Zero-gradient test Latent heat and CO2 flux comparisonmeasurement system at Lake Taihu from May 11 to August 31 2012

gmeasurement system at Lake Taihu from May 11 to August 31, 2012

160500406030y y g ,160

EC all w ind direction500 1200406030

140EC, all w ind directionGD, all w ind direction

400All w ind direction

G d f t h30yyy 140 EC, good fetch400 Good fetch 100030

ncy

40ncy20ncy

120GD, good fetch

300m-220uen

uen

uen

120

-2

300

W m 800cy

20

equ

20equ

10equ

m-

200, W enc

10Fre20Fre10Fre

100W 200

EC 600que10

Site and Instruments LE,

100of E

Freq000Site and Instruments 80L100

LE o 400F-1 0 1 2

0-0.2 0 0.20-0.01 0 0.01

H O di t %0

LCH4 gradient, ppbCO2 gradient, ppmH2O gradient, %v600 20042

Th i t it (31o24′N 120o13′E) i th th t f L k 1006030 60

The experiment site (31o24′N 120o13′E) was in the north part of Lake 0 4 8 12 16 20 2440 -100 0 100 200 300 400 500

-100 100 0 100 200 300 400 5000The experiment site (31 24 N, 120 13 E) was in the north part of Lake 0 4 8 12 16 20 24

Time Hour100 0 100 200 300 400 500

LE of GD W m-2-100 0 100 200 300 400 500

LE f GD W 240cy20cy 40cy

Taihu which is the third largest freshwater lake in China with a water Time, HourLE of GD, W m 2LE of GD, W m-240

enc20

enc 40

enTaihu, which is the third largest freshwater lake in China, with a water

que

que

20equg

surface area of 2400 km2 and a mean water depth of 1 9 m Wi d di ti f lli i th f 180 345o id d20req

10req 20

Fre

surface area of 2400 km2 and a mean water depth of 1.9 m. Wind direction falling in the range of 180-345o was considered as20F10F Fp Wind direction falling in the range of 180 345 was considered as 0

good fetch wherein the signal was least interfered by land0 05 0 0 050

20 10 0 10 200 -0.1 0 0.1 0.2

2 1 good fetch, wherein the signal was least interfered by land -0.05 0 0.05CO fl -2 -1

-20 -10 0 10 20LE W -2 CH4 flux, g m-2 s-1 g g y

ecosystemsCO2 flux, mg m-2 s-1LE, W m-2 4

A CRDS analyzer (Model ecosystems.A CRDS analyzer (Model Mixing ratio gradient 30-min Fluxecosystems.

G1301 Picarro Inc CAMixing ratio gradient 30 min Flux

I i t th EC t th GD th d d dG1301, Picarro Inc., CA, In comparison to the EC measurement, the GD method recorded, , ,USA) d t H2O CO2 CH4 LE CO2 flux CH4 flux

In comparison to the EC measurement, the GD method recorded USA) was used to H2O

%CO2 CH4

bLE

W 2CO2 flux

2 1CH4 flux

2 1 higher LE values under good fetch conditions indicating possibleUSA) was used to %v ppm ppb W m-2 mg m-2 s-1 μg m-2 s-1 higher LE values under good fetch conditions, indicating possible EC

simultaneously measureg g

bias errors in the Monin Obukhov turbulent diffusivity formulationEC

simultaneously measure Mean value 7 36×10-5 0 0332 0 203 0 231 0 00914 0 0214 bias errors in the Monin-Obukhov turbulent diffusivity formulation. GD ythe mixing ratio of CO

Mean value -7.36×10-5 -0.0332 0.203 -0.231 -0.00914 0.0214 ythe mixing ratio of CO2, Standard deviation 0 00202 0 0410 0 356 5 496 0 0113 0 0379

g 2,CH d H O Ai

Standard deviation 0.00202 0.0410 0.356 5.496 0.0113 0.03790 1 600CH4 and H2O Air was 0.5

All i d di ti

0.1 600CH4 and H2O. Air was A di t t t f d ith th t i t k iti d s-1

All w ind directionGood fetch

drawn from two intakes at A zero-gradient test, performed with the two intakes positioned -2 s Good fetch

0.051 y

200 m drawn from two intakes at A zero gradient test, performed with the two intakes positioned

m-

2 s-

400cyhttp://ditu.google.cn

th h i ht f 1 1 d next to each other indicates very small bias errors mg

0m-2 400

enc

the heights of 1.1 m and next to each other, indicates very small bias errors.0C

, m 0

mg

m

uethe heights of 1.1 m and 0EC

, m eq3 5 m above the water of

-0.05ux 200Fre

GD EC 3.5 m above the water

ux

O2 fl F

surface and measurement 2 fl

0 1CO EC, all w ind directionsurface, and measurement C

O -0.1C

GD, all w ind directionEC d f t h 0,

ti 1 i h 0 5

C EC, good fetchGD good fetch -0 1 -0 05 0 0 05 0 1

0

CH fl time was 1 min on each -0.5 0 0.5-0.5

0 4 8 12 16 20 24-0.15

GD, good fetch 0.1 0.05 0 0.05 0.12 1CH flux time was 1 min on each 0.5 0 0.5

CO flux of GD mg m-2 s-1 0 4 8 12 16 20 24Time Hour CO2 flux of GD, mg m-2 s-1CH4 flux

intake CO2 flux of GD, mg m s Time, Hour 2 , g4intake. 0 5 5003

F th CO fl b i diff i t b t th GD d th EC0.5

1

All wind direction

5003

-1

For the CO2 flux obvious difference exists between the GD and the EC0 4s-1 Good fetch 4002 s-

A dd i t For the CO2 flux, obvious difference exists between the GD and the EC 0.4

m-2

00

cy

2

m-2

An eddy covariance system method The EC system recorded negative flux in the daytime whileg m 300enc

1g mAn eddy covariance system method. The EC system recorded negative flux in the daytime, while 0.3g qu

e1

g

was installed at the samey g y

the GD method shows a weak positive flux0.3

ux, 200req

0ux,

was installed at the same the GD method shows a weak positive flux. 0 2

flu

100

Fflu

site ith a distance of 100t e G et od s o s a ea pos t e u0.2H

4 100-1H4

site, with a distance of 100 Th diff i th CO fl h b d b i thCH

0CHsite, with a distance of 100

f h GD i k The difference in the CO2 flux may have been caused by errors in the0 4 8 12 16 20 24

0.1 -1 -0 5 0 0 5 10

130 150 170 190 210 230 250-2

m from the GD intakesThe difference in the CO2 flux may have been caused by errors in the

C f ff0 4 8 12 16 20 24 1 0.5 0 0.5 1

CH flux of GD g m-2 s-1130 150 170 190 210 230 250DOY f 2012m from the GD intakes. density corrections to the EC flux and by different algal activities withinTime, Hour CH4 flux of GD, g m sDOY of 2012 density corrections to the EC flux and by different algal activities within

their respective source footprintsCH flux was positive most of the time except for occasional their respective source footprints. CH4 flux was positive most of the time except for occasional p pnegative values at nightnegative values at night. g g

Acknowledgement: This research was supported by the Natural Science Foundation of Jiangsu Province China (grant BK2010572) the Ministry of Education of China (grant PCSIRT) and the Priority Academic Program Development of Jiangsu Higher Education Institutions (grant PAPD)Acknowledgement: This research was supported by the Natural Science Foundation of Jiangsu Province, China (grant BK2010572), the Ministry of Education of China (grant PCSIRT) and the Priority Academic Program Development of Jiangsu Higher Education Institutions (grant PAPD).pp y g (g ) y (g ) y g p g g (g )