Development of the Temperate Shrub Submodel for the Community Land Model-Dynamic Global Vegetation Model (CLM-DGVM) Xubin Zeng Xiaodong Zeng Mike Barlage Department of Atmospheric Sciences University of Arizona Tucson, AZ 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 Without shrub components, the NCAR CLM-DGVM is deficient in simulating the global distribution of tree-grass-shrub distributions compared with the MODIS data Figure 1 Default DGVM photosynthesis over SW U.S. MODIS-based photosynthesis (from Running et al.) Shrubs can not be established in the default DGVM due to too small photosynthesis Figure 2 BDT: broadleaf deciduous tree NET: 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: new dotted: control Shrubs can exist when grasses or trees cannot in the default DGVM Shrubs occupies the bare area and slightly reduces grass area in default DGVM Figure 3 solid line: with shrub dotted line: DGVM Shrubs occupies shrubs do not exist the bare area Figure 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 latitudes Panel (c): shrubs cover part of the bare over arid regions Figure 7 Shrubs cover the correct regions but quantitative comparisons are difficult. Why? (see Fig. 8) Figure 8 Both MODIS land cover and fractional vegetation cover (FVC) data are needed for DGVM evaluations NEW OLD MODIS Land cover + FVC MODIS Land cover only 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 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 Diversity (1) (2) (3) (4) (5) (6) (7) Figure 11 Response of Ecosystem to Perturbations Moisture index = 0.25 (very dry) can not fully recover after removal 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: /2007JG 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: /2006GL