CLM-CN update: Sensitivity to CO 2, temperature, and precipitation in C-only vs. C-N mode Peter Thornton, Jean-Francois Lamarque, Mariana Vertenstein,

CLM-CN update: Sensitivity to CO2, temperature, and

precipitation in C-only vs. C-N mode

Peter Thornton, Jean-Francois Lamarque, Mariana Vertenstein,

Nan Rosenbloom, Jeff Lee

CLM-CN: Summary Model Structure and Fluxes

Leaf

FineRoot

DeadStem

DeadCoarse Root

LiveStem

LiveCoarse Root

PreviousStorage

CurrentStorage

Wood Litter(CWD)

Litter 1(Labile)

Litter 2(Cellulose)

Litter 3(Lignin)

SOM 1(fast)

SOM 2(medium)

SOM 3(slow)

PlantPools

LitterPools

Soil OrganicMatter Pools

C flux

Legend

Tempsensitivity

N flux

AtmosphericCO2

VegetationBiomass

SoilOrganicMatter

AtmosphericN species

Coupled Carbon-Nitrogen dynamics• Strong feedback between decomposition and plant growth: soil mineral N is the primary source of N for plant growth.

• Can result in a shift from C source to C sink under warming.

P.E. Thornton, NCAR

CLM-CN spinup summary: global total C pools (CAM drivers)

using accelerated decomposition method of Thornton and Rosenbloom, Ecol Mod (2005)

offline CLM-CN(CAM drivers)

coupled(CAM – CLM-CN)

transient, control (transient-control)GPP(CO2+Ndep)

offline CLM-CN(CAM drivers)

coupled(CAM – CLM-CN)

transient, control (transient-control)NEE(CO2+Ndep)

offline CLM-CN(CAM drivers)

coupled(CAM – CLM-CN)

transient, control (transient-control)TotC(CO2+Ndep)

Total C uptake (PgC)

(land fraction)Mean NEE (PgC/y)

Expt 1850-2000 2000-2100 1980-2000 2080-2100

N dep 16 (6%) 50 (3%) -0.24 -0.73

CO2 fert 61 (22%) 220 (13%) -0.98 -2.56

CO2+Ndep 79 (29%) 301 (17%) -1.31 -4.13

CLM-C 223 (81%) 843 (49%) -3.80 -10.75

Cumulative land carbon uptake and net ecosystem exchange, 1850-2100with constant climate (25-yr cycle), prescribed [CO2]atm

Land fractions referenced against cumulative fossil fuel emissions of276 PgC for 1850-2000 and 1732 PgC for 2000-2100 (SRES A2)

CLM-C

CLM-CN (CO2,Nfix,dep)

CLM-CN (CO2,Nfix)

CLM-CN (CO2)

C4MIP models

C4MIP mean

Land biosphere sensitivity to increasing atmospheric CO2 (L)

Results from offline CLM-CN, driven with CAM climate, in carbon-only (CLM-C) and carbon-nitrogen (CLM-CN) mode, from present to 2100. Using SRES A2 scenario assumed CO2 concentrations.

CLM-CN (CO2,Nfix,dep)

CLM-CN (CO2,Nfix)

CLM-CN (CO2)

Land biosphere sensitivity to increasing atmospheric CO2 (L)

Evidence that increasing N-limitation under rising CO2 has an important effect on the transient behavior of L, and that consideration of anthropogenic N deposition reverses this trend by around 2060.

Tair Prcp

NE

E s

ens

itiv

ity

to T

air

(P

gC

/ K

)

0

1

2

3

4

5

NE

E s

ens

itiv

ity

to P

rcp

(P

gC

/ m

m d

-1)

-25

-20

-15

-10

-5

0

CLM-CCLM-CN

NEE sensitivity to Tair and Prcp (interannual variability)

Coupling C-N cycles buffers the interannual variability of NEE due to variation in temperature and precipitation (global means, control simulations).

NEE sensitivity to Tair and Prcp (CLM-CN vs CLM-C)

CLM-CN CLM-CN

CLM-CCLM-C

Tair Prcp

FIRE

HR

NPP

NEE

Components of NEE temperature response

NPP dominates NEE response to temperature in most regions. Exceptions include Pacific Northwest, Scandanavia.

Dissection of NPP temperature response

GPP

Soil ice

BtranNPP

Warmer temperatures lead to drying in warm soils (increased evaporative demand), and wetting in cold soils (less soil water held as ice).

FIRE

HR

NPP

NEE

Components of NEE precipitation response

NPP dominates NEE response to precipitation in tropics, midlatitudes, HR dominates in arctic and coldest climates.

Dissection of HR precipitation response

Snow depthNEE

HR

Tsoil

Higher Precip in arctic/cold climate produces deeper snowpack, warmer soils, increased HR.

Potential for complex climate feedbacks depending on the spatial patterns of changing temperature and precipitation.

NPP variability dominates the Tair and Prcp response in most locations, but HR dominates the Prcp response in cold climates, due to feedback between snowpack, soil warming, and enhanced HR.

Tair Prcp

% c

han

ge

fro

m c

on

tro

l

-40

-20

0

20

40

60

CLM-C: +CO2

CLM-CN: +CO2

CLM-CN: +CO2 +Nmin

NEE sensitivity to Tair and Prcp: effects of rising CO2 andanthropogenic N deposition

Carbon-only model has increased sensitivity to Tair and Prcp under rising CO2. CLM-CN has decreased sensitivity to both Tair and Prcp, due to increasing N-limitation.

Tair Prcp

NE

E s

en

sit

ivit

y to

Ta

ir (

Pg

C /

K)

0

1

2

3

4

5

NE

E s

en

sit

ivit

y to

Prc

p (

Pg

C /

mm

d-1

)

-25

-20

-15

-10

-5

0

CLM-CCLM-CN

Tair Prcp

% c

han

ge

fro

m c

on

tro

l

-40

-20

0

20

40

60

CLM-C: +CO2

CLM-CN: +CO2

CLM-CN: +CO2 +Nmin

CLM-CN Summary: C-cycle response to nitrogen coupling,CO2, temperature, and precipitation

1. Nitrogen coupling reduces sensitivity to CO2. This effect increases with increasing CO2.

2. Anthropogenic nitrogen deposition alleviates this effect.

3. Nitrogen coupling reduces global mean sensitivity to temperature and precipitation.

4. Complex spatial patterns of NEE response to T and P. These responses would be in tension under warmer-wetter climate.

5. CO2 increases T and P sensitivities in carbon-only model, decreases sensitivities in carbon-nitrogen model.

CLM-CN summary contd:Nitrogen cycle buffers land carbon-

climate feedbacks

Nitrogen coupling…• reduces CO2 fertilization (reduces a

negative feedback on climate system)• reduces T sensitivity (reduces a positive

feedback)• reduces P sensitivity (uncertain feedback

sign)• So, not a simple result with respect to total

carbon-climate system gain.

CLM-CN development path

• 3-pool vs 4-pool switch implemented and tested, with expected results.

• Landcover change effects (product pools) underway.• 13C now incorporated and tested, transient runs

underway, collaboration with CSU.• Collaboration with LLNL to add 14C on the same

framework.• Implement age-class distributions.• Introduce N speciation and direct effects of ozone on

physiology, in collaboration with CCSM Chem-Climate WG.

Speciation of land N emissions

Nitrification vs. denitrification depends on aerobic state of soil, probably at the microscopic scale.

Sophisticated models already exist, and it should be possible to adapt them for use in CLM-CN.

Agricultural emissions could be tied to new efforts with crop modeling.