1. Addition of the CLM3 Land- Surface Model to WRF Jimy Dudhia (MMM/NCAR) Ruby Leung (PNNL) Tom Henderson (MMM/NCAR) Mariana Vertenstein (CGD/NCAR) Gordon Bonan (CGD/NCAR) Jimy Dudhia (MMM/NCAR) Ruby Leung (PNNL) Tom Henderson (MMM/NCAR) Mariana Vertenstein (CGD/NCAR) Gordon Bonan (CGD/NCAR)

2. Current LSMs in WRF  Noah LSM (NCEP/NCAR/AFWA)  RUC LSM (FSL)  5-layer soil model (NCAR)  Noah LSM (NCEP/NCAR/AFWA)  RUC LSM (FSL)  5-layer soil model (NCAR)

3. Why another LSM?  Beneficial in regional climate model applications driven by CCSM boundaries (downscaling climate scenarios)  Want to use the same physics as CCSM  CLM3 LSM  CAM3 radiation  Expertise with CLM3 is here at NCAR  Bonan, Vertenstein and others  Beneficial in regional climate model applications driven by CCSM boundaries (downscaling climate scenarios)  Want to use the same physics as CCSM  CLM3 LSM  CAM3 radiation  Expertise with CLM3 is here at NCAR  Bonan, Vertenstein and others

4. Community Land Model 3.0  Land component of the Community Climate System Model (CCSM)  Actively under development  Technical Description of the Community Land Model  Oleson, Dai et al. (May 2004)  NCAR/TN-461+STR (online PDF file)  Land component of the Community Climate System Model (CCSM)  Actively under development  Technical Description of the Community Land Model  Oleson, Dai et al. (May 2004)  NCAR/TN-461+STR (online PDF file)

5. Community Land Model What is the contribution of land surface processes to seasonal-to- interannual variability in climate and atmospheric CO 2 ? Snow Soil water Leaf phenology (the seasonal emergence and senescence of leaves) Photosynthesis and stomatal conductance What is the contribution of land surface processes to climate sensitivity (paleoclimates, future climate)? Hydrologic cycle Carbon and nitrogen cycles Mineral aerosols Vegetation dynamics Land use and land cover change Research tool is the Community Land Model Land model for Community Climate System Model Partnership among NCAR, universities, and government labs through the CCSM land model working group Note: The CLM is not designed specifically for coupling to WRF, but there is no fundamental difference between land models for climate models and NWP models. Both are 1-D models of the soil-plant-atmosphere system. They differ primarily in the complexity with which they represent meteorological, hydrological, and ecological processes and how they utilize satellite data

6. The model simulates a variety of ecological, biogeochemical, and hydrological processes as climate feedbacks that are traditionally considered when assessing the impact of climate change. The model blurs the distinction between climate feedbacks and climate impacts. Ecology and biogeochemistry Carbon and nitrogen cycles Vegetation dynamics Leaf phenology Fire Mineral aerosols Biogenic volatile organic compounds Land use and land cover change Agroecosystems Urbanization Soil degradation Historical and future land cover datasets driven by population change Hydrology Global Land-Atmosphere Coupling Experiment (GLACE) Watershed processes River flow and biogeochemistry Water isotopes High resolution CLM Subgrid orography Downscaling WRF Community support Ongoing Activities

7. CLM And WRF: Issues Inclusion of CLM in WRF provides an important scientific opportunity for NCAR and the atmospheric (global, regional) modeling communities: Same land model for use with both a global climate model (CCSM) and a regional model (WRF) New terrestrial science for WRF (e.g., carbon cycle, land use, BVOCs) But … Science Are there common experiments that should be done with CCSM and WRF? Can CLM meet the needs of data assimilation? Software engineering Surface datasets for WRF grids Initial datasets How to maintain compatibility of CCSM and WRF CLMs

8. Main Features of CLM3  Surface Heterogeneity  3 levels of grid-cell sub-division  Landunits (5 types currently)  Vegetated, Glacier, Lake, Wetland, Urban  Columns (1 currently)  multi-layer soil and snow column in Vegetated Landunit  Plant Functional Types (4 PFTs currently/column)  % of vegetation types in Vegetated Landunit  Surface Heterogeneity  3 levels of grid-cell sub-division  Landunits (5 types currently)  Vegetated, Glacier, Lake, Wetland, Urban  Columns (1 currently)  multi-layer soil and snow column in Vegetated Landunit  Plant Functional Types (4 PFTs currently/column)  % of vegetation types in Vegetated Landunit

9. Nested Hierarchy of Data Structures in CLM3

10. 15 Plant Functional Types  Needleleaf Evergreen Tree - Temperate  Needleleaf Evergreen Tree - Boreal  Needleleaf Deciduous Tree - Boreal  Broadleaf Evergreen Tree - Tropical  Broadleaf Evergreen Tree - Temperate  Broadleaf Deciduous Tree - Tropical  Broadleaf Deciduous Tree - Temperate  Broadleaf Deciduous Tree - Boreal  Broadleaf Evergreen Shrub - Temperate  Broadleaf Deciduous Shrub - Temperate  Broadleaf Deciduous Shrub - Boreal  C3 Arctic Grass  C3 Grass  C4 Grass  Crop1/Crop2  Needleleaf Evergreen Tree - Temperate  Needleleaf Evergreen Tree - Boreal  Needleleaf Deciduous Tree - Boreal  Broadleaf Evergreen Tree - Tropical  Broadleaf Evergreen Tree - Temperate  Broadleaf Deciduous Tree - Tropical  Broadleaf Deciduous Tree - Temperate  Broadleaf Deciduous Tree - Boreal  Broadleaf Evergreen Shrub - Temperate  Broadleaf Deciduous Shrub - Temperate  Broadleaf Deciduous Shrub - Boreal  C3 Arctic Grass  C3 Grass  C4 Grass  Crop1/Crop2

11. Soil Layers  10 soil layers  mid-points near 0.7, 2.8, 6.2, 11.9, 21.2, 36.6, 62.0, 104, 173, 286 cm  Up to 5 snow layers on top  10 soil layers  mid-points near 0.7, 2.8, 6.2, 11.9, 21.2, 36.6, 62.0, 104, 173, 286 cm  Up to 5 snow layers on top

12. Biogeophysical Processes  Vegetation composition, structure, phenology  Absorption, reflectance, and transmittance of solar radiation  Absorption and emission of longwave radiation  Momentum, sensible heat (ground and canopy), and latent heat (ground evaporation, canopy evaporation, transpiration) fluxes  Heat transfer in the soil and snow including phase changes  Vegetation composition, structure, phenology  Absorption, reflectance, and transmittance of solar radiation  Absorption and emission of longwave radiation  Momentum, sensible heat (ground and canopy), and latent heat (ground evaporation, canopy evaporation, transpiration) fluxes  Heat transfer in the soil and snow including phase changes

13. Biogeophysical Processes (cont’d)  Canopy hydrology (interception, throughfall and drip)  Snow hydrology (snow accumulation and melt, compaction, water transfer between snow layers)  Soil hydrology (surface runoff, infiltration, sub-surface drainage, redistribution of water within the columns)  Stomatal physiology and photosynthesis  Lake temperatures (multi-layer) and fluxes  Routing and runoff from rivers to ocean (not in WRF yet)  Biogenic volatile organic compounds (BVOCs) (could be coupled to WRF-Chem)  Canopy hydrology (interception, throughfall and drip)  Snow hydrology (snow accumulation and melt, compaction, water transfer between snow layers)  Soil hydrology (surface runoff, infiltration, sub-surface drainage, redistribution of water within the columns)  Stomatal physiology and photosynthesis  Lake temperatures (multi-layer) and fluxes  Routing and runoff from rivers to ocean (not in WRF yet)  Biogenic volatile organic compounds (BVOCs) (could be coupled to WRF-Chem)

14. Melt Transpiration Canopy Water Snow Hydrology Drainage Evaporation Interception Sublimation Throughfall Stemflow Infiltration Surface Runoff Evaporation Precipitation Soil Water Redistribution Direct Solar Radiation Absorbed Solar Radiation Diffuse Solar Radiation Longwave Radiation Reflected Solar Radiation Emitted Long- wave Radiation Sensible Heat Flux Latent Heat Flux uaua 0 Momentum Flux Wind Speed Soil Heat Flux Heat Transfer Photosynthesis Biogeophysics Community Land Model CLM simulates energy and moisture exchanges between land and atmosphere Energy exchanges include radiative transfer, turbulent fluxes, and heat storage in soil These are controlled in part by the hydrologic cycle CLM has a detailed representation of the hydrologic cycle including: interception of water by leaves; infiltration and runoff; multi-layer snow accumulation and melt; 10- layer soil water; and partitioning of latent heat into evaporation of intercepted water, soil evaporation, and transpiration Bonan (2002) Ecological Climatology (Cambridge University Press)

15. Coupling CLM to WRF  CLM modules are kept in tact  Software maintenance easier with single code  Code will be shared between WRF, CCSM, and offline CLM implementations  CLM will operate on distributed-memory processors geographically collocated with WRF DM patches  Unlike coupling to CCSM where CLM is independently distributed from CAM for load-balancing  CLM modules are kept in tact  Software maintenance easier with single code  Code will be shared between WRF, CCSM, and offline CLM implementations  CLM will operate on distributed-memory processors geographically collocated with WRF DM patches  Unlike coupling to CCSM where CLM is independently distributed from CAM for load-balancing

16. Initializing CLM  Use CLM global datasets for landunits, PFTs, soil textures, etc. (static data)  Later use WRF hi-res global vegetation/soil data  Interpolate soil temperature, soil moisture, snow from WRF input to CLM levels (dynamic data)  Use CLM global datasets for landunits, PFTs, soil textures, etc. (static data)  Later use WRF hi-res global vegetation/soil data  Interpolate soil temperature, soil moisture, snow from WRF input to CLM levels (dynamic data)

17. Run-time Coupling  WRF to CLM  Lowest level atmospheric wind, temperature, water vapor  Downward radiation  Precipitation  WRF to CLM  Lowest level atmospheric wind, temperature, water vapor  Downward radiation  Precipitation

18. Run-time Coupling  CLM to WRF  Wind stress  Sensible heat flux  Latent heat flux and water vapor flux  Albedo  Surface (skin) temperature (upward longwave)  2m T and q  CLM to WRF  Wind stress  Sensible heat flux  Latent heat flux and water vapor flux  Albedo  Surface (skin) temperature (upward longwave)  2m T and q

19. WRF-CLM Coupling WRFCLM WRF SI CLM preprocessor WRFCLM Lat/long, land mask InitializationRun Land-surface map Initial atmospheric state Declination angle Skin temp, albedo Lowest level height, wind components, theta, water vapor, pressure, temperature. Downward longwave, shortwave flux. Precipitation (rain/snow) Skin temp, albedo Wind stress, sensible and latent heat flux, water vapor flux. 2m T and q. Initial land state (arbitrary) current

20. WRF-CLM Coupling WRFCLM WRF SI WRFCLM InitializationRun Landuse/soil, etc. mapped to grid Initial atmospheric state Declination angle Initial land state: Snow, Soil temp, soil moisture, canopy water Skin temp, albedo Lowest level height, wind components, theta, water vapor, pressure, temperature. Downward longwave, shortwave flux. Precipitation (rain/snow) Skin temp, albedo, emissivity Wind stress, sensible and latent heat flux, water vapor flux. 2m T and q. Soil temp, soil moisture, canopy water, snow, snow depth, runoffs, ground flux (for WRF output) planned Soil T,q interpolated vertically in REAL

21. Conclusion  So far we have done tests with fixed initial land- state  Work is ongoing to provide real initial state to CLM  Work is planned to provide WRF hi-res USGS landuse and FAO soil to initialize CLM landunits and PFTs  CLM3 coupled to WRF will be released to the WRF community when this work is completed  So far we have done tests with fixed initial land- state  Work is ongoing to provide real initial state to CLM  Work is planned to provide WRF hi-res USGS landuse and FAO soil to initialize CLM landunits and PFTs  CLM3 coupled to WRF will be released to the WRF community when this work is completed