1
Bernard Longdoz1, D. Epron1, S. Goffin2, A. Granier1, E. Granier1, P. Gross1, N. Heydt1, J. Ngao3, C. Plain1, N. Marron1, F. Parent1, M. Schrumpf4
Combination of different techniques and multi
Combination of different techniques and multi
-
-
scale
scale
approach to understand CO
approach to understand CO
22budget in a temperate
budget in a temperate
beech forest
beech forest
2 Liège University, Gembloux Agro-Bio Tech, Belgium 3 Paris-Sud University, CNRS, UMR 8079, Orsay, France
1 INRA, Centre INRA de Nancy, UMR1137, 54280 Champenoux, France
Nancy University, INRA, UMR 1137, Vandoeuvre les Nancy, France
4 MPI Biogeochemistry, Jena, Germany
Context
Quantification C storage Understanding C processes
European Beech (Fagus Sylvatica), French forest
N E W
S
Nancy
Nancy HesseHesse
Hesse site
48°40’N, 7°05’E
48°40’N, 7°05’E
Mean annual air Temp. : 10°C10°C
Mean annual Precip. : 950 mm950 mm Temperate
Temperate
65 km from Nancy (East)
Climate
3 200 m Hesse 1 N 90% beech 42 years old Height 19 m
Soil type: stagnic luvisolstagnic luvisol
5 horizons : A, S1, Sg2, II Cg, III C Csoil 9.5 kg m-2 LAI : 7.5 -5 Density ≈ 3000 n ha-1 Aerial biomass ≈ 100 t ha-1
Hesse site
Automatic measurements :
• Net Ecosystem Exchange NEE (Eddy Cov., 30 min) • Micro-climate (T°, radiation,
humidity, precipitations) • Soil T° and water content • Trunk circumferences
(dendrometers) ⇨ C biomass
Measurement campaigns :
• Aerial (Stems, leaves, fruits,…) and below ground (roots) Biomass
• Soil composition & characteristics (density, C & N contents…
• δ13C of sampled materials (IRMS) and gaz (TDLS + IRMS) • Soil Respiration (Rs) • LAI Tower 22m Sonic anemometer IRGA (CO2)
Material
5 NEE GPP TER -30 -20 -10 0 10 20 14/7 15/7 16/7 17/7 Time NEE (µ m ol m -2 s -1 ) GPP GPP+TER+TER TER TER
NEE measurements
FootprintGPP
GPP & TERTER have difference in response to changes in environnemental conditions
1st approach:
1. TERTER (t°, SWC) determination from night and leafless data
2. TERTER (t°, SWC) extrapolation to leafy daytime
3. Daytime : GPP
GPP = NEENEE (data) – TERTER (extrapolation)
2nd approach: Combining 2 equations
1. NEENEE = GPPGPP + TERTER
2. NEENEE *δ13C
NEE = GPPGPP *δ13Catm + TERTER *δ13CTER
NEE partitioning
7 Improvement Rs model a 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 180 240 300 360 420 480 540 600 660 720 780 Jour (2003-2004) RS (µmo lCO 2 m -2 s -1 ) H1A H1B H1C H1D H1E H1F H1G H2B H2I Footprint fluctuations (WD, WS, u*)
+
TER (Rs) spatial variability
Low turbulence (u*<<) periods
Sous-estimation des flux de nuits
0 1 2 3 4 5 6 7 8 9 0.0 0.2 0.4 0.6 0.8 1.0 1.2 u* (ms-1) FL UX ( µ molm-2s-1) R10 ETE 1997 R10 ETE 1998 R10 ETE 1999 R10 ETE 2003 EC anomalies TER (Ts10, SWC10) validity 0 1 2 3 4 5 6 7 8 9 10 9 10 11 12 13 14 15 16 17 18 19 20 21 Ts10 (°C) T ER (µmo l m -2 s -1 )
1
stApproach Uncertainties
Large uncertainty on TER (t°, SWC) TER (Ts10, SWC10): R² < 0.1
(Longdoz et al., 2008)
N
E
W
S
100 m H1A H1A H1B H1B H1C H1C H1D H1DH1EH1E H1F H1F H1G H1G HydromorphicHydromorphic levellevel
BOR BO Bre
Hesse1
Hesse1
Li-6252Rs spatial variability
Temporal variability ( (ρρSS))AAsoil bulk density of the A layer (C/N) (C/N)AA ratio of carbon to nitrogen content of A layer &
Above & belowground characteristics
(
)
(
exp
a
b
SWC
)
exp
Q
R
R
10 10 T 10 10 s 10⋅
−
−
⋅
⋅
=
−9 (R2 = 0.87) R2 = 0.6602 0.0 0.5 1.0 1.5 2.0 2.5 0.4 0.6 0.8 1.0 1.2 Densité (g cm-3) R1 0 (µ m o lCO 2 m -2 s -1 ) R2 = 0.7258 0.0 0.5 1.0 1.5 2.0 2.5 8 9 10 11 12 C/N A0 R1 0 (µ m o lCO 2 m -2 s -1 ) (C/N)A (ρS)A (g/cm3)
( )
( )
A A S 101
.
47
0
.
19
C
N
R
=
−
⋅
ρ
+
⋅
R10 mapSoil sampling for ((ρρSS))AAand (C/N)(C/N)AA mapping Soil sampling points (>100) in footprint area
Rs spatial variability
Rs spatial variability
( )
( )
A A S 101
.
47
0
.
19
C
N
R
=
−
⋅
ρ
+
⋅
11 Rs spatial variability
Confirmed by first analyse of EC-TER data
ANOVA with 45° Sectors significant difference (p=0.029) “Hot area” at West-South-West
“Cold area” at South-West
T°s & SWC R10 map Rs map Rs spatial variability Rs contributions to TER Estimations of TER temporal fluctuations due to Rs spatial variability N N sparse vegetation mature deciduous young deciduous coniferous agriculture sparse vegetation mature deciduous young deciduous coniferous agriculture
Distance East - West [m]
-600 -300 0 300 600 900 D is ta nce N orth - So uth [m] -900 -600 -300 0 300 600 + 5 10 20 N
Distance East - West [m]
-600 -300 0 300 600 900 D is ta nce N orth - So uth [m] -900 -600 -300 0 300 600 + 5 10 20 N N Footprint model
13 a 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 180 240 300 360 420 480 540 600 660 720 780 Jour (2003-2004) RS (µmo lCO 2 m -2 s -1 ) H1A H1B H1C H1D H1E H1F H1G H2B H2I Improvement Rs model
Large uncertainty on TER (t°, SWC) TER (Ts10, SWC10): R² < 0.1 (Longdoz et al., 2008) EC anomalies Low turbulence (u*<<) periods TER (Ts10, SWC10) validity
Sous-estimation des flux de nuits
0 1 2 3 4 5 6 7 8 9 0.0 0.2 0.4 0.6 0.8 1.0 1.2 u* (ms-1) FL UX ( µ molm-2s-1) R10 ETE 1997 R10 ETE 1998 R10 ETE 1999 R10 ETE 2003 0 1 2 3 4 5 6 7 8 9 10 9 10 11 12 13 14 15 16 17 18 19 20 21 Ts10 (°C) T ER (µmo l m -2 s -1 )
1
stApproach Uncertainties
Footprint fluctuations (WD, WS, u*) +
TER (Rs) spatial variability
Improvement of Rs model
Differentiate
Soil CO2 efflux
Production & Transport
Better description of the Transport processes
Not enough information
Important vertical
profile
Need for CO
Need for CO22 concentration concentration
measurements in the different
measurements in the different
layers
layers
∆ [CO2]
∆ [CO2]
∆ [CO2]
Better description of the Production processes
Differentiate sources and their behaviour response to environmental
conditions Also necessary to have better understanding
15 Improvement of Rs model
To be continued in the Caroline
To be continued in the Caroline
Plain Talk Plain Talk
+ CO
+ CO
22Transport
Transport
model
model
IRGA IRGA (CO2) http://www.foxitsoftware.com For evaluation only.2nd approach: Combining 2 equations 1. NEENEE = GPPGPP + TERTER
2. NEENEE *δ13C
NEE = GPPGPP *δ13Catm + TERTER *δ13CTER
2
ndApproach Uncertainties
Footprint fluctuations +
δ13C
TER (δ13CRs ) spatial variability
δ13C
TER (δ13CRs ) temporal variability
Large uncertainty on δ13C
17
Soil Keeling Plot
Soil Keeling Plot
LICOR Li LICOR Li- -6200 6200 Sampling chamber Sampling chamber IRGA IRGA Li Li--62006200 Respiration Respiration chamber chamber Mass spectrometer Mass spectrometer Keeling plot Keeling plot
δ
13C
Rsspatio-temporal variability
δ13C Rsspatial variabilityAir
Soil
Sample Pump
Pump Reference gas (10% CO2)
Laser Sample cell Sample detector Reference detector Dewar δ13C temporal variability Ref : Daniel
Ref : Daniel EpronEprontalk, talk, MarronMarronet et al. 2009 al. 2009 TDLS (δ C of gas) Soil chambers (δ13C out - δ13C) ⇒δ13C Rs
19 Marron et al., 2009 Subplots Seasonal variation ≈3‰ Daily variation ≈2‰ Small scale spatial variation ≈2‰ Large scale spatial variation > 1‰ δ
δ13C Rs Spatio-temporal variations from –25‰ to -30 ‰ TER TER uncertainty of 20% Complex climatic determinism (SWC 4 to 5 days before) Use of δ13C as tracer : C transfer,
allocation-source partitioning (Daniel Epron talk)
More knowledge about 13C
discriminations
(δ13C Production & δ13CTransport)
Understanding
δ13C Production &
δ13CTransport determinism
Improvement of δ13C
21
Improvement of
δ
13C
Rsmodel
Better description of the δ13C Production
processes
Differentiate sources and their behaviour Differentiate
Important vertical
profile
Need for
Need for δδCO2 concentration CO2 concentration measurements in the different
measurements in the different
layers
layers
Soil δCO2 efflux
δ13C Production & δ13C Transport
response to environmental
conditions
Not enough information Not enough information
Better description of the δTransport processes
∆ [δCO2]
∆ [δCO2]
∆ [δCO2]
Improvement of δ13C model
To be continued in the Caroline
To be continued in the Caroline
Plain Talk Plain Talk
+
+
δ
δ
CO
CO
22Transport
Transport
model
model
Sample PumpPump Reference gas (10% CO2)
Laser Sample cell Sample detector Reference detector Dewar
23
あり
がと
う
Rauto = 69% de Rs
Corrections: - trenched roots decomposition
- higher SWC in trenched plot
NPP = NEE – Rs hetero
4 m 4 mR
R
HeteroHeteroR
R
AutoAutoR
R
HeteroHeteroRs autotrophic-heterotrophic
partitioning: Trenched plots
NPP from Eddy covariance (NPPec)
Li-6252
Rs : soil chambers
25 Rauto 52.0% tree growth 29.8% leaves 11.0% fine roots 4.8% fruits 0.2% mortality2.2%
Pourcentage from GPP = 1404 gC m
-2 NPPbioNPP from biomass measurements (NPPbio)
0 200 400 600 800 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 N P P (g C m -2 y r -1 ) NPPec NPPbio