the CN–DGVM coupling

Sam LevisPeter Thornton & Gordon BonanTerrestrial Sciences Section, NCAR

CLM-CN

• Lead: Peter Thornton• Since: CLM3.5• Based on:

Biome-BGCof Running et al.

CLM-DGVM

• Lead: Sam Levis• Since: CLM3.0• Based on:

LPJ-DGVMof Sitch et al.

Goal: Merge two efforts

Simulates:

• C and N in veg & soil• Vegetation structure

Simulates:

• C in veg & soil• Vegetation structure

and distribution

CLM-CN CLM-DGVM

CN + DGVM →

CNDV

Goal: Merge two efforts

Subroutine CN DGVM CNDVNDep/Fixation Hourly - Hourly

MRespiration Hourly Hourly Hourly

Decomp/Allocation Hourly Hourly/Annual Hourly

Phenology Hourly Hourly Hourly

GRespiration Hourly Hourly Hourly

GapMortality Hourly Annual H: with mods

FireArea/Fluxes Hourly Hourly/Annual H: with mods

NLeaching Hourly - Hourly

VegStructUpdate Hourly Annual H: with mods

LightCompetition - Annual Annual

Bioclimatology Satellite: Annual Annual Annual

Establishment New C: Annual C →

%pft: Annual Annual

Beyond the Coupling

• Let prescribed crops coexist with simulated natural vegetation

• Introduce shrubs (Xubin’s group, other)• Increase DGVM sophistication• Introduce crop model

CN →

CNDV• Subroutine driver1, every hour:

– Call CNEcosystemDyn• Call …• Call GapMortality ←

constant% vs growth efficiency• Call Fire ←• Call NLeaching• Call VegStructUpdate ←

CN’s d, h, tlai, tsai vs LPJ’s

• Subroutine driver2, every year:– Call dv

• Call Bioclimatology & LightCompetition• Call Establishment ←

NOW WITHOUT SAPLINGS

– Call dvreset & resetWeights

max and annleafold

wood

newwood

oldnew C

CCPP =

CN →

CNDV Establishment

LPJ: Existing pfts may establish saplings every year Allometric eqns with laisapl =1.5 calculate csapl

c + csapl and P + Pestabl → d, ca, lai → FPC

CNDV: New/Expand. pfts get Cleaf =10 gC m-2 pft area Assume FPCinit to convert units and calculate

c and P + Pestabl → d, ca, lai → FPC

– Initial Vegetation →– Year-long Bioclimatology accumulators– End-of-Year Establishment →

CNDV year 1

0 , , 222 ≥jjj FPCPC

[ ]npftjFPCPC jjj ,1 ,0111 ∈===

end of

year 1

lake wetland

natural vegetation

glacier urban

0.11 =bareFPC

crop

lake wetland

natural vegetation

glacier urban

999.02 ≈bareFPC

crop

CNDV year 2

– Hourly biogeophysics and biogeochemistryAlloc., Mortality, Fire, VegStructUpdate →

– End-of-year processes• First guess →• Competition for Light →• Establishment →• VegStructUpdate →

botj

topjjjold

woodj

newwoodj

oldjnew

jfirej

mortj

allocj hhSAILAI

CCP

PCCC ,,,,,,

,=⇒Δ+ΔΔ

333, jjj FPCPC ⇒

jexcess

grasstreeexcesstrees PFPCFPC Δ⇒>+> 195.0 ,

333 ,, jjjj FPCPPC ⇒Δbotj

topjjj hhSAILAI ,,,

First guess

End-of-Year y Processes

( )( )

( )( )

11

1,

1,

111

,,

,,

++

+

+

+++

=⇒

⎪⎪

⎭

⎪⎪

⎬

⎫

=

=

=

=

yjjj

yj

yleafjjjj

jj

ywoodjj

yj

yj

yj

yj

PfpccaFPC

ccaslaffpc

dfca

cfd

FPCPCfc

1+yjFPC

End-of-Year y Processes

( )( )

entEstablishm andLight in updated 1

1

1

+

+

+

Δ+Δ+Δ+Δ+=

Δ+Δ+Δ+=

yj

estabj

lightj

firej

mortj

yj

yj

firej

mortj

allocj

yj

yj

FPC

PPPPPP

CCCCC

Light and Establishment

CN →

CNDV

• Results from the first attempt (2006)

CLM-DGVM

CNDV with CN Mortality LPJ d

CNDV with CN Mortality CN d

CLM-DGVM

CNDV with CN Mortality LPJ d

CNDV with CN Mortality CN d

Follow-up Work

• Repeat with clm3.5

• Interp. P and FPC to hourly (Peter: done?)

• Add other LPJ concepts to CNDV? For example,1. 100% mortality for NPPann ≤

0?

2. Boreal-pft heatstress?

3. Age and size classes by extending “avg individual”

• Repeat CNDV work with CASA?