Ian WoodwardIan Woodward

NERC Centre for Terrestrial Carbon DynamicsNERC Centre for Terrestrial Carbon Dynamics

Department of Animal & Plant SciencesDepartment of Animal & Plant Sciences

University of SheffieldUniversity of Sheffield

QUERCCQUERCC: : QUQUantifying antifying EEcosystem cosystem RRoles in the oles in the CCarbon arbon CCycleycle

Overall objective: – NERC-speak.

To quantify the contemporary terrestrial carbon cycle using new combinations of data and models.

This will be achieved through 4 work packages

Overall objective: – reality?.

Change/replace TRIFFID components to include a nitrogen cycle, a wider range of functional types and more ecologically realistic sub-grid scale dynamics.

TRIFFID

Schematic showing TRIFFID carbon flows for each vegetation type. Processes above the dotted line are fluxes calculated in the MOSES2 land surface scheme every atmospheric model time step (≈ 30 minutes).

In dynamic mode, TRIFFID updates the vegetation and soil carbon every 10 days using time-averages of these fluxes.

TRIFFID and GCM coupling.

Changes in the distribution and structure of five functional types feedback to climate via two routes.

1. Vegetation determines the biophysical parameters which affect fluxes of heat, water and momentum.

2. Changes in the carbon stored in vegetation and soil ( NEP) also change atmospheric CO2 and climate.

Nitrogen deposition is also shown as a driver for vegetation change, but this version of TRIFFID does not include an interactive nitrogen cycle.

QUERCCQUERCC: : QUQUantifying antifying EEcosystem cosystem RRoles in the oles in the CCarbon arbon CCycleycle

WP1. Develop new calibrated models of soil chemistry and nutrients that influence the carbon cycle, with particular emphasis on the N cycle.

WP2. Develop, expand and validate descriptions of plant function.

WP3. Create model(s) to capture sub-grid scale dynamics of vegetation behaviour.

WP4. Combine the above models into the JULES structure and apply globally.

Work package objectives

QUERCCQUERCC: : QUQUantifying antifying EEcosystem cosystem RRoles in the oles in the CCarbon arbon CCycleycle

Observational data base

Simulations data base

WP2Functional types (FTs)

Leeds

WP1Soil N,P

Aberdeen

WP3Sub-grid cell dynamics

Sheffield

WP4Input to JULES

CEH/Exeter

Work Package 1Work Package 1

How to represent nutrient availability in How to represent nutrient availability in modelsmodels

1. Temporal compartmentalisation of soil C cycling.

2. Identify the role of soil nutrient availability (N and P) on rapid and slow cycling.

Work Package 1Work Package 1

HypothesisHypothesis ––Organic matter pool turnover regulated by nutrient driven feedbacks in the short and long term

Connect the soil and Connect the soil and vegetation modelsvegetation models

Wardle et al. (2004) Science

Atmosphere

Aerobic conditions

Anaerobic conditions

Organic matter

Detritalorganic matter

NH4+

NH4+

NO2NO3

N2

N2O

N2O

N2N2O

N2

Photosynthesis

NitrificationBacterialdegradation

Bacterialdegradation

Denitrification

N2 fixation

After Wollast (1981)

The Nitrogen cycleWP1

Work Package 2Work Package 2GLC 2000 vegetation map

Key regions

Wright et al.: Major axis of leaf variation

1001,000

10,000

10

• Major axis of variation from fastto slow living plants.

• Also correlated: – N content, – P content– Assimilation

capacity, – Dark respiration

Change in leaf longevity of dominant vegetation (months)

• Black = Outside Climatic range

Sub-grid scale activityWork Package 3Work Package 3

Work Package 3Work Package 3TRIFFID/MOSES land surface scheme

Bare Gd Shrub

C4 grass

C3 grassNLeaftree

BLeaftree

• Grid cell divided into tiles.• Each tile is one PFT• Size of tile determined by

empirical dominance hierarchy• Tendency for a single

vegetation type to dominate

• E.g. LPJ (Smith et al 2001)N

PP

Existing JULES model

NL

BL

BL

NL

Work Package 3Work Package 3Ecosystem Demography Model (ED)

Moorcroft et al. (2001)“Size and age structured

approximation of an individual based model”

0 y.o. 5 y.o.

94 y.o.

28 y.o.61 y.o.

14 y.o.

Bare Gd Shrub

C4 grass

C3 grassNLeaftree

BLeaftree

TRIFFID ED= PFT1= PFT2= PFT3= PFT42 years

5 years

30 years

150 years

Area Area

Simulating succession in EDLeaf cost vs. mortality trade off

Cheap leavesHigh mortality

Expensive leavesLow mortality

Work Package 4Work Package 4

soil

Canopy conductancephenology

Dynamic vegetation

radiation

snow

hydrology

Surface exchange

WP1: N cycle and P response

WP2: Improved characterisation of PFTs

WP3: Sub-grid cell dynamics

WP4: Combine new knowledge into JULES structure

“Newer” processes for land surface modelling

Soil Nutrient regulated Ecosystem Carbon Cycle

Nutrient availability (N:P content) Fast C Slow C

VegetationC

CO2

Can N and P availability be used as a predictor of Carbon cycling?

Slide from Nick Ostle, CEH Lancaster

Waterlevel

Moist

ure Texture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Soillevel

INPUTSMax.Water

levelRain,PET

INPUTSAir Temp

INPUTSSoil

Parameters

BIO HUM IOM

CH4 CH4

MethaneOxidation

Meth.Oxid.

Decomposition

CO2 CO2INPUTS

NPPLU Type

DPMRPM

DOC

Waterlevel

Moist

ure Texture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Moist

ure Texture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Moist

ure Texture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Soillevel

INPUTSMax.Water

levelRain,PET

INPUTSAir Temp

INPUTSSoil

Parameters

INPUTSMax.Water

levelRain,PET

INPUTSAir Temp

INPUTSSoil

Parameters

BIO HUM IOMBIO HUM IOM

CH4 CH4

MethaneOxidation

Meth.Oxid.

CH4 CH4

MethaneOxidation

Meth.Oxid.

MethaneOxidation

Meth.Oxid.

Decomposition

CO2 CO2

Decomposition

CO2 CO2

DecompositionDecomposition

CO2 CO2CO2 CO2INPUTS

NPPLU Type

DPMRPM

DOCDOCDOC

Slide from Peter Smith, Aberdeen

Work Package 3Work Package 3

Leaf life span change from fixed per FT to Wright Leaf life span change from fixed per FT to Wright et alet alvariable relationship with temperature and precipitationvariable relationship with temperature and precipitation

MonthsMonths

WP1 Soil model evaluation and improved process description

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10Year

% C

Rem

aini

ng

a)

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10Year

% C

Rem

aini

ng

a)

RothCCANDY

DNDCCENTURY

DAISYSOMM

ITEVerberne

NCSOIL0

5

10

15

20

25

RMSE

aa a a a a

b bb

RothCCANDY

DNDCCENTURY

DAISYSOMM

ITEVerberne

NCSOIL0

5

10

15

20

25

RMSE

aa a a a a

b bb

Testing models

Waterlevel

Moistu

reTexture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Soillevel

INPUTSMax.Water

levelRain,PET

INPUTSAir Temp

INPUTSSoil

Parameters

BIO HUM IOM

CH4 CH4

MethaneOxidation

Meth.Oxid.

Decomposition

CO2 CO2INPUTS

NPPLU Type

DPMRPM

DOC

Waterlevel

Moistu

reTexture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Moistu

reTexture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Moistu

reTexture

Oxy

gen

Temperature

pH

DecompositionDrivers

WaterModule

Temperature

Module

pH Module

TextureModule

Oxy

gen

Mod

ule

Soillevel

INPUTSMax.Water

levelRain,PET

INPUTSAir Temp

INPUTSSoil

Parameters

INPUTSMax.Water

levelRain,PET

INPUTSAir Temp

INPUTSSoil

Parameters

BIO HUM IOMBIO HUM IOM

CH4 CH4

MethaneOxidation

Meth.Oxid.

CH4 CH4

MethaneOxidation

Meth.Oxid.

MethaneOxidation

Meth.Oxid.

Decomposition

CO2 CO2

Decomposition

CO2 CO2

DecompositionDecomposition

CO2 CO2CO2 CO2INPUTS

NPPLU Type

DPMRPM

DOCDOCDOC

Model improvement

Predicted changes in LL/SLA ratio

C4

Needleleaf

Evergreen Bleaf

C3

Dec B’leaf

Dec N’leaf

Wright et al 2004

Work Package 3Work Package 3

Wright et al 2004

Work Package 4Work Package 4

JULES modular structure

Ocean andsea ice

Unified Model IMOGEN Other driving data

soil

Canopy conductancephenology

Dynamic vegetation

radiation

snow

hydrology

Surface exchange

Wright et al.: Major axis of leaf variationChange in leaf longevity of dominant vegetation (months)TRIFFID/MOSES land surface schemeEcosystem Demography Model (ED)Moorcroft et al. (2001)TRIFFIDEDSimulating succession in EDLeaf cost vs. mortality trade offModulation of leaf traits by climate: Wright et al. 2005. Predicted changes in LL/SLA ratio