Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

31
Gap Filling Comparison Workshop, September 18-20, 2006, Jena, German Corinna Rebmann Olaf Kolle Max-Planck-Institute for Biogeochemistry Jena, Germany Eddy covariance measurements and their shortcomings for the determination of the net ecosystem exchange of carbon dioxide

description

Eddy covariance measurements and their shortcomings for the determination of the net ecosystem exchange of carbon dioxide. Corinna Rebmann Olaf Kolle Max-Planck-Institute for Biogeochemistry Jena, Germany. Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany. Outline. - PowerPoint PPT Presentation

Transcript of Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Page 1: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Corinna RebmannOlaf Kolle

Max-Planck-Institute for BiogeochemistryJena, Germany

Eddy covariance measurements and their shortcomings for the

determination of the net ecosystem exchange of carbon dioxide

Page 2: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Outline

• Introduction of measurement site and advection experiment

• Reasons for data gapsSpecial features of open path analyser

• Consequences for final flux data

• Summary

Page 3: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Measurement Site:

Wetzstein, Thuringia, Germany,

flux measurements established end of 2001

main towertower C tower B

tower D

tower A

measuring heights: Main tower: 30.0mTower C: 29.4m

ADVEX’06 (April 11– June 19, 2006)flux measurements for , H, E, CO2

CO2, wind and temperature profiles

Page 4: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

The ADVEX Experiment

Advection experiment CarboEurope-IP:4 towers around the main tower:A, B, C, D: profiles of [CO2], T, u‘, v‘, w‘, T‘,tower B with CO2-fluxes below canopy,tower C and main tower with CO2-fluxes above canopy

60m

Page 5: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Why care about advection?

Eddy covariance theory is derived from tracer conservation equation with many simplifications which are only valid under homogeneous conditions

Page 6: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Data gaps are due to

• Maintenance interruptions, power failures, ice coating

• Instrumental problems• Non-turbulent conditions• Unfavoured wind directions (tower effects,

heterogeneous terrain)• Precipitation, fog events (open path

analyser)• high wind speeds

Page 7: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main towerdata gaps (closed-path analyser)

Jan 1 – Aug 24, 2006

00:00

12:00

00:00

Tim

e

1.1. 1.3. 1.5. 1.7.

caused by maintainance, power failure, ice coating

1.9% gaps

00:00

12:00

00:00

1.1. 1.3. 1.5. 1.7.

after pre-selection ( high-frequency data)

3.5% gaps

00:00

12:00

00:00

Tim

e

1.1. 1.3. 1.5. 1.7.

Date

after stationarity test 1 ( diff >50%)

25.9% gaps

00:00

12:00

00:00

1.1. 1.3. 1.5. 1.7.

Date

after stationarity test 2 ( 30%<diff<50%)

42.1% gaps

data available

no data

Page 8: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower CApr 11 – Jun 19, 2006

data gaps caused by maintenance, power failures etc.

-100

-80

-60

-40

-20

0

20

40

60

FC

O(µ

mol

ms

)2

-2-1

0.0

1.0

2.0

3.0

4.0

prec

ipita

tion

(mm

)

15.4. 22.4. 29.4.1.5. 8.5. 15.5. 22.5. 29.5. 1.6. 8.6. 15.6.

Date

FCO2 TM

FCO2 TC

precip (mm)

Page 9: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower CApr 11 – Jun 19, 2006

data gaps caused by maintenance, power failures etc.

step Main tower (TM)

Tower C (TC)

1(maintenance etc)

3.6% 4.8%

2(after pre-selection)

3(after stationarity test 1)

Page 10: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower CApr 11 – Jun 19, 2006

time series of CO2-fluxes after pre-selection(eg Vickers & Mahrt 1997, JAOT14)

-40

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

0.0

1.0

2.0

3.0

4.0

rain

(mm

)

15.4. 22.4. 29.4.1.5. 8.5. 15.5. 22.5. 29.5. 1.6. 8.6. 15.6.

Date

FCO2 TM

FCO2 TC

rain (mm)

Page 11: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower CApr 11 – Jun 19, 2006

data gaps after pre-selection

step Main tower (TM)

Tower C (TC)

1(maintenance etc)

3.6% 4.8%

2(after pre-selection)

4.5% 30.2%

3(after stationarity test 1)

30.2%

Page 12: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower C

which data are rejected in case of open path-analyser?

-40

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

0.0

1.0

2.0

3.0

4.0

rain

(mm

)

1.5. 8.5. 15.5. 22.5. 29.5. 1.6. 8.6.

Date

FCO2 TM

FCO2 TC

rain (mm)

-40

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

0.0

1.0

2.0

3.0

4.0

rain

(mm

)

1.5. 2.5. 3.5. 4.5. 5.5. 6.5. 7.5. 8.5. 9.5. 10.5. 11.5. 12.5. 13.5.

Date

FCO2 TM

FCO2 TC

rain (mm)

24 of 624 half-hours (3.8%) rejected

April 30 – May 12, 2006, dry period

Page 13: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower C

which data are rejected in case of open path-analyser?

-40

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

0.0

1.0

2.0

3.0

4.0

rain

(mm

)

1.5. 8.5. 15.5. 22.5. 29.5. 1.6. 8.6.

Date

FCO2 TM

FCO2 TC

rain (mm)

-40

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

0.0

1.0

2.0

3.0

4.0

rain

(mm

)

14.5. 15.5. 16.5. 17.5. 18.5. 19.5. 20.5. 21.5. 22.5. 23.5. 24.5. 25.5. 26.5. 27.5. 28.5.

Date

FCO2 TM

FCO2 TC

rain (mm)

263 of 630 half-hours (41.7%) rejected!!!

May 13 – 28, 2006, rainy period

Page 14: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower C Apr 11 – Jun 19, 2006

consequences for dependencies on meteorological variables

GGsatGPP

GsatGPPday R

RaF

RFaNEE

,

,Michalis-Menten-relationship:see Falge et al. 2001, AFM107

NEE: net ecosystem exchange (µmol CO2 m−2 s−1)

PPFD: photosynthetic photon flux density (µmol quantum m−2 s−1)

a: ecosystem quantum yield (µmol CO2) / (µmol quantum)

FGPP,sat: gross primary productivity at saturating light (µmol CO2 m−2 s−1)

Rday: ecosystem respiration during the day (µmol CO2 m−2 s−1)

-30

-20

-10

0

10

NE

E(µ

mo

lms

)-2

-1

0 200 400 600 800 1000 1200 1400 1600 1800 2000

PPFD (µmol m s )-2 -1

day-time NEE, TM

day-time NEE, TC TM TC

a 0.057 0.061

FGPP,sat 24.6 27.6

Rday 5.4 4.1

r2 0.66 0.58

Page 15: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower C Apr 11 – Jun 19, 2006

consequences for dependencies on meteorological variables

GGsatGPP

GsatGPPday R

RaF

RFaNEE

,

,Michalis-Menten-relationship:see Falge et al. 2001, AFM107

NEE: net ecosystem exchange (µmol CO2 m−2 s−1)

PPFD: photosynthetic photon flux density (µmol quantum m−2 s−1)

a: ecosystem quantum yield (µmol CO2) / (µmol quantum)

FGPP,sat: gross primary productivity at saturating light (µmol CO2 m−2 s−1)

Rday: ecosystem respiration during the day (µmol CO2 m−2 s−1)

-30

-20

-10

0

10

NE

E(µ

mol

ms

)-2

-1

0 200 400 600 800 1000 1200 1400 1600 1800 2000

PPFD (µmol m s )-2 -1

day-time NEE, TM, if CT av.

day-time NEE, TMTM TM,

TC avTC

a 0.057 0.063 0.061

FGPP,sat 24.6 23.2 27.6

Rday 5.4 5.1 4.1

r2 0.66 0.59 0.58

Page 16: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower C

time series of CO2-fluxes with stationarity tests

-40

-30

-20

-10

0

10

20

FC

O(µ

mo

lms

)2

-2-1

1.5. 3.5. 5.5. 7.5. 9.5. 11.5. 13.5. 15.5. 17.5. 19.5.

Date & Time

FCO2 TM

FCO2 TC

FCO2 TM if stat2

FCO2 TC if stat2

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

0

100

200

300

400

500

600

700

800

900

1000

R(W

m)

G-2

9.5. 10.5. 10.5. 11.5. 11.5. 12.5. 12.5. 13.5. 13.5. 14.5. 14.5. 15.5.

Date

FCO2 TM

FCO2 TC

FCO2 TM if stat2

FCO2 TC if stat2

global radiation RG

May 8 – 14, 2006

Page 17: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower CApr 11 – Jun 19, 2006

When do instationaries occur?

0

5

10

15

num

ber

ofin

stat

ion

ary

dat

a(%

)

10 50 90 130

170

210

250

290

330

370

410

450

490

530

570

610

650

690

730

770

810

850

R (W m )G-2

TM

TC

Instationarities occur mainly atlow or zero radiation conditions

Page 18: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzstein, main tower and tower CApr 11 – Jun 19, 2006data gaps summary

step Main tower (TM)

Tower C (TC)

1 3.6% 4.8%

2(after pre-selection)

4.5% 30.2%rainy, moist conditions

3(after stationarity test 1)

9.8% 33.2%

Low radiation conditions

Page 19: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Do we have perfect data now?

Are these data reliable as input for gap filling procedures?

Still missing:advective processesnight flux treatment

reliability check

Page 20: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

HainichDrainage/advective fluxes

Data fromW. Kutsch

-20

-10

0

10

FC

O(µ

mo

lms

)2

-2-1 storage flux

turbulent flux

-10

-5

0

5

10

FC

O(µ

mol

ms

)2

-2-1 Horiz.Advection

running mean

0.0

1.0

qual

ityfla

gs

17.8. 18.8. 18.8. 19.8. 19.8. 20.8. 20.8.

Date & Time

physical/technical flag

stationarity flag

nighttime u* flag

12:00 12:00 12:00 12:00

Page 21: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Night-flux problem

• Weak turbulence• Instrumental problems, large footprints,

gravity waves• Turbulent flux is influenced by other

transport/storage processes→Site dependentsee eg: Lee, 1998

Aubinet et al, 2003, 2005Staebler and Fitzjarrald, 2004Feigenwinter et al, 2004

Page 22: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Night-flux corrections

Empirical:

Separate calm and turbulent periods, remove calm periods, fill the gap

u*-criterion mostly used

Page 23: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Aubinet et al. AER30, 2000

NEEnight versus u*

0.0

1.0

2.0

3.0

4.0

norm

aliz

ed

CO

-flu

x2

0

2

4

6

8

10

12

14

16

18

20

freq

uen

cy(%

)

0.0

250.

125

0.2

250.

325

0.4

250.

525

0.6

250.

725

0.8

250.

925

1.0

251.

125

1.2

251.

325

1.4

251.

525

1.6

251.

725

1.8

251.

925

u* (m s )-1

Wetzstein, n=7446

Page 24: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

-40

-30

-20

-10

0

10

20

30

FC

O(µ

mol

ms

)2

-2-1

15.9. 17.9. 19.9. 21.9. 23.9. 25.9. 27.9. 29.9.

Date

WetzsteinNEE 2005, unrealistic high night-time fluxes

Page 25: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzsteinwhen do high fluxes occur?

• u*>0.4m s-1• wind direction between 200° and 280° or 30° and 40°• neutral atmospheric conditions:

stability parameter: -0.0625<ζ<0.0625

(determined by M. Zeri)

→ turbulent upwind mixing from the valley

Page 26: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

WetzsteinNEE 2005

after application of MZ criteria

-40

-30

-20

-10

0

10

20

30

FC

O(µ

mol

ms

)2

-2-1

15.9. 17.9. 19.9. 21.9. 23.9. 25.9. 27.9. 29.9.

Date

NEE MT, after MZ criteria

NEE MT, before MZ criteria

for 2005:72% data available58% data available

Page 27: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

WetzsteinNEE 2005

after application of MZ criteria

0.0

1.0

2.0

3.0

4.0

norm

aliz

ed

CO

-flu

x2

0

2

4

6

8

10

12

14

16

18

20

freq

uen

cy(%

)

0.0

250.

125

0.2

250.

325

0.4

250.

525

0.6

250.

725

0.8

250.

925

1.0

251.

125

1.2

251.

325

1.4

251.

525

1.6

251.

725

1.8

251.

925

u* (m s )-1

Wetzstein, n=7446

0.0

1.0

2.0

3.0

4.0

norm

aliz

edC

O-f

lux

2

0

2

4

6

8

10

12

14

16

18

20

freq

uenc

y(%

)

0.02

50.

125

0.22

50.

325

0.42

50.

525

0.62

50.

725

0.82

50.

925

1.02

51.

125

1.22

51.

325

1.42

51.

525

1.62

51.

725

1.82

51.

925

u* (m s )-1

Wetzstein, n=4991

after selection criteria

Page 28: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Wetzsteinnight-time NEE 2005

after application of MZ criteria

-10

0

10

20

30

FC

O(µ

mo

lms

)2

-2-1

-20 -15 -10 -5 0 5 10 15 20 25 30

T (°C)air, 2m

NEE after pre-selection, stationary

NEE after pre-selection, stationary and MZ criteria

R10=3.9R10=3.0

Page 29: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

WetzsteinNEE comparison during advection experiment

after application of MZ criteria

-30

-20

-10

0

10

20

FC

O(µ

mol

ms

)2

-2-1

11.5. 12.5. 13.5. 14.5. 15.5. 16.5. 17.5. 18.5. 19.5. 20.5. 21.5.

Date in 2006

MT, left after MZ criteria

MT, discarded with MZ criteria

TC

TS

Page 30: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

SummaryAmount of data gaps strongly depending on:• site• type of quality check

(still no common agreement in CarboEurope-IP!)• type of analyser, weather pattern• threshold criteria for u*

(have to be objective, Gu et al. AFM128, 2005)

Derived dependencies on meteorological variables vary with data left after selection

→biased datasets

Reliability has to be tested against chamber and biometric measurements

Page 31: Gap Filling Comparison Workshop, September 18-20, 2006, Jena, Germany

Thanks for your attention!

Questions?