GERFS1 Top-down approach to estimation of the regional carbon budget in West Siberia S. Maksyutov (1) T. Machida, K. Shimoyama, N.Kadygrov, A. Itoh (1)

GERFGERFS1S1

Top-down approach to estimation of the regional carbon budget in West Siberia

S. Maksyutov (1)

T. Machida, K. Shimoyama, N.Kadygrov, A. Itoh (1) P. Patra (2) M. Arshinov, O. Krasnov, B. Belan (3), N. Fedoseev (4)

(1) CGER/NIES, Tsukuba, Japan(2) FRCGC/JAMSTEC, Yokohama, Japan(3) IAO, Tomsk, Russia(4) Permafrost Research Institute, Yakutsk, Russia

Inverse modeling of regional CO2 fluxes using tower network in West Siberia

Transcom workshop, Purdue Univ., Apr 23-27, 2007

GERFGERFS1S1

Method: Inverse model of the atmospheric CO2 transport is applied to constrain surface CO2 fluxes by the observed patterns of the atmospheric CO2 (with seasonal cycles)

Components:

- Bottom-up estimate of C fluxes (long term), based on the terrestrial ecosystem NPP, respiration, biomass, biomass change.

1. Forward models: terrestrial ecosystem flux model (hourly to seasonal

scale): coupled to atmospheric transport model.

2. Inverse model of atmospheric transport, finding optimal corrections to the surface fluxes


GERFGERFS1S1

Q. (Back in 2002) Why do inverse modeling of carbon fluxes in West Siberia?

- Established collaboration programs on atmospheric and ground based observations

- Advantage for atmospheric transport modeling: simplified meteorology, flat terrain.

- Scientifically interesting region: strong impact by climate change.


GERFGERFS1S1


Integrated vegetation map (1x1 km)based on

GLC2000,forest inventory MODIS VCF vegetation continuous fields (500m)wetland typology map

GERFGERFS1S1

Bottom-up estimate of C fluxes (long term), based on the terrestrial ecosystem NPP, respiration, biomass, biomass change.

Inventory of the carbon content and its long-term changes using field observations and forest survey data.

- Forest inventory: provides observations of the wood stock (carbon stock) and annual change in forest area by category (felling, fire, etc)

- Soil carbon stock inventory (by agricultural soil profile observations)

- Wetlands (25% of the area, accumulators of peat deposits: need to know the area, and area fraction by landscape elements


GERFGERFS1S1

Empirical modeling of the forest carbon stock inventory and dynamicsForest state account (FSA): provides observations of the wood stock and annual

change in forest area by category (felling, fire, etc), Frequency: reporting 1 year, full reporting 5 years, unit survey about 15 years


FSA: regions and enterprises

FSA data for each unit: wood stock, area - by species, age class

Net Primary Production

0

100

200

300

400

500

600

700

800

0 100 200 300

Age, Year

I

II

III

IV

V

Va

Carbon, g/m2/yr

Input to empirical dynamic model

Cedar (pinus sibirica)

NPP by age and productivity class.

Based on yield tables.

GERFGERFS1S1


soil profiles1 -Soil carbon database

2- IIASA Russia CD

45

50

55

60

65

70

60 65 70 75 80 85 90

lon

lat lat

lat

1. Comparison of the soil carbon surveys in recent 30 years indicates soil carbon loss to erosion. (Titlyanova et al)

Problem: While forests and wetlands provide carbon sink, the agricultural lands show some loss of carbon

2. Long term observations of the soil carbon after conversion to arable lands

show stabilization of carbon content after ~50years. (Barsukov et al)

Siberian soil carbon profile database (Titlyanova et al)

soil map

Surgut1993 -

NovosibirskYakutsk

GERFGERFS1S1


Airborne observations: air sampling and analysis

GERFGERFS1S1

Top-down approach to estimation of the regional carbon budget in West SiberiaOBSERVATIONS

Igrim(IGR)（63o12’N, 64o24’E）47m, 24m

Demyanskoe(DEM)（59o47’N, 70o52’E）

63m, 45m

Parabel(PRB)（58o15’N, 82o24’E）67m, 35m

Berezorechka (BRZ)（56o09’N, 84o20’E）

80m, 40m, 20m, 5m

Yakutsk(YAK)（62o50’N, 129o21’E）70m, 11m

Noyabrsk(NOY)（63o26’N, 76o46’E）

43m, 21m

planned

NoyabrskIgrim

Beloretsk

Demyanskoe

Azovo

Parabel

Berezorechka

Savuushka

ZotinoYakutsk

Igrim(IGR)（63o12’N, 64o24’E）47m, 24m

Demyanskoe(DEM)（59o47’N, 70o52’E）

63m, 45m

Parabel(PRB)（58o15’N, 82o24’E）67m, 35m

Berezorechka (BRZ)（56o09’N, 84o20’E）

80m, 40m, 20m, 5m

Yakutsk(YAK)（62o50’N, 129o21’E）70m, 11m

Noyabrsk(NOY)（63o26’N, 76o46’E）

43m, 21m

Working

NoyabrskIgrim

Vaganovo

Demyanskoe

Azovo

Parabel

Berezorechka

Savuushka

ZotinoYakutsk

Observations: hourly data, -> selected afternoon data and fitted with Globalview type fits

350375400425

350375400425

CO

2 [

ppm

]

63m 45m

350375400425

350375400425

43m 21m

350375400425

350375400425

47m 24m

350375400425

350375400425

67m 35m

2002 2003 2004 2005 2006 2007350375400425

350375400425

BRZ

DEM

NOY

IGR

PRB

YEAR

70m 11m

YAK

350375400425

350375400425

80m 40m

GERFGERFS1S1


Inverse modeling: monthly fluxes from 66 regions

64 region map 66 region map (color map – by region number)

Inverse model used in this study is documented in: Patra, P.K., M. Ishizawa, S. Maksyutov, T. Nakazawa, and G. Inoue, (2005)

Global Biogeochem. Cycles, 19, GB3005, doi:10.1029/2004GB002258.

Seasonal variation of CO2 Flux with 66 region inversion and 1 year (2005) of tower data

-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2

CASA land44 CASA + TDI, land44

-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


W.Siberia South (top), North (bottom)

Central.Siberia South (bottom), North (top)

Seasonal variation of CO2 Flux with 66 region inversion and 1 year (2005) of tower data



-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2


Seasonal variation of CO2 Flux with 66 region inversion and WITHOUT 1 year (2005) of tower data



-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay

/m2

CASA CASA + TDI, land25

-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay/m

2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay/m

2


-4

-3

-2

-1

0

1

2

3

4

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006g/d

ay/m

2


land44

-1.3

-0.8

-0.3

0.3

0.8

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006

CO

2 F

lux

[tC

/ha/

Mo

nth

]

CASA TDI Sim-CYCLE

land43

-1.3

-0.8

-0.3

0.3

0.8

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006C

O2

Flu

x [

tC/h

a/M

on

th]

CASA TDI Sim-CYCLE

land27

-1.3

-0.8

-0.3

0.3

0.8

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006

CO

2 F

lux

[tC

/ha/

Mo

nth

]

CASA TDI Sim-CYCLE

land25

-1.3

-0.8

-0.3

0.3

0.8

2004.9 2005 2005.1 2005.2 2005.3 2005.4 2005.5 2005.6 2005.7 2005.8 2005.9 2006

CO

2 F

lux

[tC

/ha

/Mo

nth

]

CASA TDI Sim-CYCLE

Seasonal variation of CO2 Flux, Comparison with another biospheric model (Simcycle)

W.Siberia South (bottom), North (top)


GERFGERFS1S1


Summary and conclusions:

Tower observations in 2002-2006 were fitted using data filtering procedure (Nakazawa et al) at 5 W.S. sites to produce Globalview type fits.

Only daytime observations were used

Seasonal flux variability in well constrained regions shows same amplitude as CASA (prior) but earlier summer drawdown.

Seasonal variability in under-constrained ones is not that reasonable.

The inverted fluxes amplitude and seasonality look similar to those simulated by Simcycle – prognostic model, which may suggest that:

1) the inverse model fluxes are reasonable (biologically)2) regional scale inverse model might be used as a tool to observe(!) regional

scale seasonality of natural terrestrial carbon cycle

GERFS1 Top-down approach to estimation of the regional carbon budget in West Siberia S. Maksyutov (1) T. Machida, K. Shimoyama, N.Kadygrov, A. Itoh (1)

Documents

regional carbon budget

loss of carbon

modeling of carbon fluxes

forest carbon stock

carbon sink

carbon content

soil carbon stock inventory

west siberia slide