Development of the Temperate Shrub Submodel for the Community Land Model-Dynamic Global Vegetation Model (CLM-DGVM)

Xubin ZengXiaodong ZengMike Barlage

Department of Atmospheric Sciences

University of ArizonaTucson, AZ 85721

xubin@atmo.arizona.edu

Motivation

• Dry regions represent a large fraction of the global land;

• Most of the existing Dynamic Global Vegetation Model (DGVMs) do not include shrubs or do not effectively distinguish shrubs from grasses;

• Exclusion of DGVMs and associated carbon cycle is recognized as one of the main deficiencies of IPCC AR4 model simulations

Without

Without shrub components, the NCAR CLM-DGVM is deficient in simulating the global distribution of tree-grass-shrub distributionscompared with the MODIS data

Figure 1

Default DGVMphotosynthesisover SW U.S.

MODIS-basedphotosynthesis(from Runninget al.)

Shrubs can not be established in the default DGVM due to too smallphotosynthesis

Figure 2

BDT: broadleaf deciduous treeNET: needleleaf evergreen tree

Shrub Submodel

• drought-tolerance in the photosynthesis computation – use of different soil moisture stress function for shrubs

• appropriate phenology type – raingreen for shrubs; no air temperature limitation for establishment

• appropriate morphology parameters• consistent treatment of fractional vegetation coverage

[in default DGVM, photosynthesis over plant crown area (PCA) while plant maintenance respiration over foliar projective cover (FPC) are used; FPC < PCA]

• tree/grass/shrub hierarchy for light competition

solid line: newdotted: control

Shrubs can exist whengrasses or treescannot in the default DGVM

Shrubs occupiesthe bare area and slightly reducesgrass area indefault DGVM Figure 3

solid line: with shrubdotted line: DGVM

Shrubs occupies shrubs do not existthe bare area

Figure 4400-Yr Simulation using DGVM with shrub submodel

Figure 5

New -- ControlFigure 6

Panels (a), (b) and (d): surprising that tree/grass competition and soil moisture are affected over NH high latitudesPanel (c): shrubs cover part of the bare over arid regions

Figure 7

Shrubs coverthe correct regions but quantitativecomparisonsare difficult.

Why? (see Fig. 8)

Figure 8

Both MODIS land coverand fractional vegetationcover (FVC) data are needed for DGVM evaluations

NEWOLD

MODISLand cover+ FVC

MODISLand coveronly

Figure 9

Summary• Developed a shrub submodel for the DGVM for the global competition of trees, grass, and shrubs • Shrubs grow primarily by reducing the bare soil coverage and to a lesser degree, by decreasing the grass coverage

• Shrub coverage reaches its peak around annual precipitation (Pann) of 300 mm, the grass coverage reaches its peak over a

broad range of Pann (from 400-1100 mm), and the tree coverage reaches its peak for Pann = 1500 mm or higher (Fig. 9)

• Use of MODIS land cover data alone is not sufficient for the DGVM model evaluation (particularly for shrubs) (Fig. 9)

Figure 10

Remaining issues: a) MODIS shows a significant boreal shrub coverage which is not covered in this work; b) as mentioned in Fig. 6, high latitude tree/grass competition is affected

Current work: develop the boreal shrub submodel

Vegetation Pattern and DiversityVegetation Pattern and Diversity

(1) (2) (3) (4) (5) (6) (7)

Figure 11 Response of Ecosystem to Perturbations

Moisture index = 0.25 (very dry)

can not fullyrecover afterremoval

Moisture index= 0.26 (dry)

can largely recover after removal

Figure 12 Additional conclusions from Figs. 11 & 12

• Developed a 3-variable ecosystem model for dry regions to simulate the bifurcation and spatial patterns and study the

effect of grazing and climate variability on ecosystem

• When spatial interactions are included, vegetation can exist

even under the environmental condition in which uniform vegetation cannot exist

• None of the current DGVMs or land models considers spatial

interactions

• These modeling results need to be confirmed using high-resolution

satellite and insitu data

Relevant publications

• X.D. Zeng, X. Zeng, and M. Barlage, 2008: Growing temperate

shrubs over arid and semiarid regions in the NCAR Dynamic

Global Vegetation Model (CLM-DGVM). Global Biogeochemical

Cycles, in press.

• X.D. Zeng, and X. Zeng, 2007: Transition and pattern diversity

in arid and semiarid grassland: A modeling study. J. Geophys.

Res.-Biogeosciences, 112, G04008, doi:10.1029/2007JG000411.

• Miller, J., M. Barlage, X. Zeng, H. Wei, K. Mitchell, and D.

Tarpley, 2006: Sensitivity of the NCEP Noah land model to the

MODIS green vegetation fraction dataset. Geophys. Res. Lett.,

33, L13404, doi:10.1029/2006GL026636.  

1 2

3 4 5 6

7 8 9 10

11 12 13 14

15 16 17 18