The Effects of Historical Changes in Global Agricultural Land on the Terrestrial Carbon Cycle Navin...

The Effects of Historical Changes in Global Agricultural Land on the

Terrestrial Carbon Cycle

The Effects of Historical Changes in Global Agricultural Land on the

Terrestrial Carbon Cycle

Navin Ramankutty[nramanku@facstaff.wisc.edu]Center for Sustainability and the Global Environment (SAGE)Institute for Environmental Studies, University of Wisconsin

Navin Ramankutty[nramanku@facstaff.wisc.edu]Center for Sustainability and the Global Environment (SAGE)Institute for Environmental Studies, University of Wisconsin

5.5 ± 0.5

Land Use

Atmospheric Pool 750

(stores 3.2 ± 0.2 yr–1)

Ocean 40,000

Units: Stocks - Gt-C Fluxes - Gt-C yr–1

NPP60

60

1.6 ± 1.0

Rh

92 90

Landplants 700

Soils 1550

The Global Carbon Cycle

Net =2.0 ± 0.8

(adapted from Schimel et al., 1995)

(status in the 1980s)

Land Use EmissionsLand Use Emissions

• Current estimates -- Houghton et al.

• Land use data over 9 continental-scale regions

• “Book-keeping” model

• Current estimates -- Houghton et al.

• Land use data over 9 continental-scale regions

• “Book-keeping” model

BackgroundBackground

0

1

2

3

4

5

6

7

1860 1880 1900 1920 1940 1960 1980

Carb

on E

mis

sions

(Gt-

C/y

r)

Fossil Fuel

Land Use(Houghton et al.)

0

1

2

3

4

5

6

7

1860 1880 1900 1920 1940 1960 1980

Carb

on E

mis

sions

(Gt-

C/y

r)

Fossil Fuel

Land Use(Houghton et al.)

Houghton (1999)Houghton (1999)

Over the 1850-1990 period,

– Cropland change = 68% of total net C flux

– Harvest of wood = 16%

– Pastures = 13%

– Shifting cultivation = 4%

– Plantations = -1%

Over the 1850-1990 period,

– Cropland change = 68% of total net C flux

– Harvest of wood = 16%

– Pastures = 13%

– Shifting cultivation = 4%

– Plantations = -1%

This StudyThis Study

• Geographically-explicit land use data, albeit restricted to croplands

• Using process-based ecosystem models

• Conducted as part of the Carbon Cycle Model Linkage Project (CCMLP), funded by the Electric Power Research Institute (EPRI)

• Geographically-explicit land use data, albeit restricted to croplands

• Using process-based ecosystem models

• Conducted as part of the Carbon Cycle Model Linkage Project (CCMLP), funded by the Electric Power Research Institute (EPRI)

• Grand Slam Expt. -- concurrent effects of historical CO2, climate and land use on terrestrial carbon cycle

4 Terrestrial Biosphere Models HRBM -- Esser et al., Giessen, Germany IBIS -- Foley et al., Kucharik et al., Univ. of Wisconsin, USA LPJ -- Sitch, Prentice et al., PIK & MPI-Jena, Germany TEM -- McGuire et al., Tian et al., MBL, USA

3 simulations from 1860 to 1992(analysis from 1920)

• Grand Slam Expt. -- concurrent effects of historical CO2, climate and land use on terrestrial carbon cycle

4 Terrestrial Biosphere Models HRBM -- Esser et al., Giessen, Germany IBIS -- Foley et al., Kucharik et al., Univ. of Wisconsin, USA LPJ -- Sitch, Prentice et al., PIK & MPI-Jena, Germany TEM -- McGuire et al., Tian et al., MBL, USA

3 simulations from 1860 to 1992(analysis from 1920)

The CCMLP StudyThe CCMLP Study

S1: CO2 onlyS2: CO2 + ClimateS3: CO2 + Climate + Land use

S1: CO2 onlyS2: CO2 + ClimateS3: CO2 + Climate + Land use

This TalkThis Talk

Effects of land use on the terrestrial carbon cycle

= Simulation S3 - S2

Effects of land use on the terrestrial carbon cycle

= Simulation S3 - S2

Driving DataDriving Data

• CO2: Ice-core record + Avg. of Mauna Loa & South Pole

• Climate: Temperature and Precipitation anomalies from Jones et al. (1994) plus Leemans & Cramer (1991); surrogate for pre-1900

• CO2: Ice-core record + Avg. of Mauna Loa & South Pole

• Climate: Temperature and Precipitation anomalies from Jones et al. (1994) plus Leemans & Cramer (1991); surrogate for pre-1900

Driving Data (continued)Driving Data (continued)

• Land Use: Boolean version of historical croplands data set of Ramankutty and Foley (1999)– Synthesis of the IGBP 1km global land cover

data set with historical cropland census data

• Land Use: Boolean version of historical croplands data set of Ramankutty and Foley (1999)– Synthesis of the IGBP 1km global land cover

data set with historical cropland census data

--> talk in parallel session B2, at 4:25 pm, Room EF --> talk in parallel session B2, at 4:25 pm, Room EF

Global Cropland Distributions Source: Ramankutty & Foley, 1999Global Cropland Distributions Source: Ramankutty & Foley, 1999

Center for Sustainability and the Global Environment, Institute for Environmental StudiesUniversity of Wisconsin, MadisonCenter for Sustainability and the Global Environment, Institute for Environmental StudiesUniversity of Wisconsin, Madison

Global Cropland Distributions Source: Ramankutty & Foley, 1999Global Cropland Distributions Source: Ramankutty & Foley, 1999

Center for Sustainability and the Global Environment, Institute for Environmental StudiesUniversity of Wisconsin, MadisonCenter for Sustainability and the Global Environment, Institute for Environmental StudiesUniversity of Wisconsin, Madison

Global Cropland Distributions Source: Ramankutty & Foley, 1999Global Cropland Distributions Source: Ramankutty & Foley, 1999

Center for Sustainability and the Global Environment, Institute for Environmental StudiesUniversity of Wisconsin, MadisonCenter for Sustainability and the Global Environment, Institute for Environmental StudiesUniversity of Wisconsin, Madison

MicrobialRespiration

Net Primaryproductivity

NaturalVegetation

Biomass

Harvest

Product decayflux

ProductPools

= 1, 10, & 100 yrs

AgriculturalProducts

Net Carbon Exchange, NCE(positive into the atmosphere)

Net PrimaryProductivity

CropBiomass

Harvest

NaturalTurnover

Litter &Soil Organic Matter

Slash CropResidue

Product decayflux

= 1 yr

The Carbon Flow in CCMLP Expts.

0

0,5

1

1,5

2

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

HRBM IBIS LPJ TEM

0

0,5

1

1,5

2

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

HRBM IBIS LPJ TEM

CCMLP Land Use FluxCCMLP Land Use Flux

0

0,5

1

1,5

2

2,5

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

CCMLPEnvelope

0

0,5

1

1,5

2

2,5

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

CCMLPEnvelope

CCMLP Flux in Comparison with HoughtonCCMLP Flux in Comparison with Houghton

0

0,5

1

1,5

2

2,5

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

Houghton (cropland)

CCMLPEnvelope

0

0,5

1

1,5

2

2,5

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

Houghton (cropland)

CCMLPEnvelope

CCMLP Flux in Comparison with HoughtonCCMLP Flux in Comparison with Houghton

0

0,5

1

1,5

2

2,5

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

Houghton (all landuse)

Houghton (cropland)

CCMLPEnvelope

0

0,5

1

1,5

2

2,5

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

Houghton (all landuse)

Houghton (cropland)

CCMLPEnvelope

CCMLP Flux in Comparison with HoughtonCCMLP Flux in Comparison with Houghton

SummarySummary

CCMLP Houghton

1920-1990(Pg-C)

55-90 57

1980s avg.(Pg-C/ yr)

0.6-1.0 1.3

CCMLP Houghton

1920-1990(Pg-C)

55-90 57

1980s avg.(Pg-C/ yr)

0.6-1.0 1.3

What does this imply?What does this imply?

In the 1980s,• Houghton (all landuse) = 2.0 Gt-C/yr• Houghton (croplands) = 1.3 Gt-C/yr

==> croplands = 65% of total land use flux

• CCMLP cropland flux = 0.6-1.0 Gt-C/yr– Scaled by 65%, ==> all landuse flux = 0.9-1.5 Gt-C/yr

In the 1980s,• Houghton (all landuse) = 2.0 Gt-C/yr• Houghton (croplands) = 1.3 Gt-C/yr

==> croplands = 65% of total land use flux

• CCMLP cropland flux = 0.6-1.0 Gt-C/yr– Scaled by 65%, ==> all landuse flux = 0.9-1.5 Gt-C/yr

Implications for Missing SinkImplications for Missing Sinkbased on

C-budget Houghton(Gt-C/yr)

CCMLP(Gt-C/yr)

Atmospheric Increase -3.3

Fossil-Fuel 5.5

Land Use 2.0

Oceanic Uptake -2.0

Missing Sink -2.2

based onC-budget Houghton

(Gt-C/yr)CCMLP(Gt-C/yr)

Atmospheric Increase -3.3

Fossil-Fuel 5.5

Land Use 2.0

Oceanic Uptake -2.0

Missing Sink -2.2

Implications for Missing SinkImplications for Missing Sinkbased on

C-budget Houghton(Gt-C/yr)

CCMLP(Gt-C/yr)

Atmospheric Increase -3.3 -3.3

Fossil-Fuel 5.5 5.5

Land Use 2.0 0.9-1.5

Oceanic Uptake -2.0 -2.0

Missing Sink -2.2 -1.1-1.7

based onC-budget Houghton

(Gt-C/yr)CCMLP(Gt-C/yr)

Atmospheric Increase -3.3 -3.3

Fossil-Fuel 5.5 5.5

Land Use 2.0 0.9-1.5

Oceanic Uptake -2.0 -2.0

Missing Sink -2.2 -1.1-1.7

Why are the land use emissions different?

Why are the land use emissions different?

• Differences in the land use data?• Differences in the land use data?

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

1850 1870 1890 1910 1930 1950 1970

Rate of Change (Ramankutty)Rate of Change (Houghton)

Ra

te o

f C

ha

nge

of C

ropla

nd

(M

illio

n k

m2

/yr)

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

1850 1870 1890 1910 1930 1950 1970

Rate of Change (Ramankutty)Rate of Change (Houghton)

Ra

te o

f C

ha

nge

of C

ropla

nd

(M

illio

n k

m2

/yr)

Cropland Conversion RatesCropland Conversion Rates

Why are the land use emissions different?

Why are the land use emissions different?

• Differences in the land use data?

• Differences in process representation?– regrowth, soil turnover, product fluxes, ...

• Differences in the land use data?

• Differences in process representation?– regrowth, soil turnover, product fluxes, ...

0.04

0.06

0.08

0.1

0.12

0.14

0.16

200

400

600

800

1000

1200

1850 1870 1890 1910 1930 1950 1970

Rate of Change of CroplandCarbon Emissions

Rate

of

Change o

f C

ropla

nd (

Million k

m2/y

r)C

arb

on E

missio

ns (T

gC

)

0.04

0.06

0.08

0.1

0.12

0.14

0.16

200

400

600

800

1000

1200

1850 1870 1890 1910 1930 1950 1970

Rate of Change of CroplandCarbon Emissions

Rate

of

Change o

f C

ropla

nd (

Million k

m2/y

r)C

arb

on E

missio

ns (T

gC

)Houghton conversion rates and emissionsHoughton conversion rates and emissions

CCMLP conversion rates and emissionsCCMLP conversion rates and emissions

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0

0.5

1

1.5

2

2.5

1920 1930 1940 1950 1960 1970 1980

Rate of Change of Cropland

Carbon Emissions CCMLP Envelope

Rate

of

Change o

f C

ropla

nd (

Million k

m2/y

r)C

arb

on E

missio

ns (T

gC

)

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0

0.5

1

1.5

2

2.5

1920 1930 1940 1950 1960 1970 1980

Rate of Change of Cropland

Carbon Emissions CCMLP Envelope

Rate

of

Change o

f C

ropla

nd (

Million k

m2/y

r)C

arb

on E

missio

ns (T

gC

)

Why are the land use emissions different?

Why are the land use emissions different?

• Differences in the land use data?

• Differences in process representation– regrowth, soil turnover, ...

• Different vegetation types are cleared?

• Differences in the land use data?

• Differences in process representation– regrowth, soil turnover, ...

• Different vegetation types are cleared?

0

0,5

1

1,5

2

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

HRBM IBIS LPJ TEM

0

0,5

1

1,5

2

1920 1930 1940 1950 1960 1970 1980 1990

Car

bon

Em

issi

ons

(Pg-

C/y

r)

HRBM IBIS LPJ TEM

CCMLP Land Use FluxCCMLP Land Use Flux

ConclusionsConclusions

• There are now two different estimates of carbon emissions due to land use

• The disagreement between the estimates is related to:– Differences in land conversion rates– Differences in process representation

• Inverse estimates of the missing carbon sink are critically dependent on estimates of land use carbon emissions

• There are now two different estimates of carbon emissions due to land use

• The disagreement between the estimates is related to:– Differences in land conversion rates– Differences in process representation

• Inverse estimates of the missing carbon sink are critically dependent on estimates of land use carbon emissions