Using MODIS Data to Characterize Climate Model Land Surface Processes and the Impacts of Land Use/Cover Change on Surface Hydrological Processes
Robert Dickinson (Principal Investigator), Georgia Institute of Technology
Improvements in CLM3 land surface air temperature due to the use of a new land surface dataset created from MODIS. The upper panel (a,b) shows initial temperature bias compared to observations in winter (DJF) and summer (JJA). The lower panel (c,d) shows temperature differences after the MODIS dataset was used. The red color represents positive change while blue color represents negative change. Stippling shows regions where the difference is statistically significant. Color contrasts between the upper and lower panels shows where improvements are made. For example, the model cold bias over northern high latitudes in summer and the warm bias over northern high latitude America in winter are significantly reduced.
This project is being undertaken by a multidisciplinary multi-institutional team with considerable experience in both climate modeling and remote sensing algorithms. This team is developing a quantitative basis for the assessment of consequences of land-use/cover change in terms of climate and hydrological changes. It is building on a climatology of spectral albedos related to the plant functional types (PFTs) of the Community Land Model (CLM3), a component of the Community Climate System Model (CCSM3), hosted at NCAR. The first part of this investigation is focused on establishing more realistic radiative models for these PFTs and using MODIS data to improve descriptions of land surface parameters such as leaf area index, fractional vegetation cover, fractional snow cover, and emissivity/skin temperature in CLM3. Relevant model processes are also reformulated to ensure their consistency with the MODIS observations. Major progress achieved so far includes (a) derivation of new land surface datasets from MODIS (more accurate and consistent) for use in climate models and (b) development of new land-surface parameterizations of snow fraction for the melting season, albedo for arid and semi-arid regions, and characterization of urban surfaces. These new datasets and schemes significantly improve land surface climate and energy balance simulations in the NCAR CLM3/CAM3 models. The second part of this research will investigate regions of change identified by MODIS land cover dynamic algorithms and will characterize such changes in terms of changes of spectral albedos and hence in terms of some combination of changes of PFTs or of the other properties connected to a particular PFT. This study will provide a basis for including year-to-year changes of land-use/cover in climate prediction models and for establishing the hydrological consequences of past and future land-cover change. More information on this project can be found at: http://climate.eas.gatech.edu/dickinson. Other LCLUC Climate Variability and Change Projects: Bounoua Lahouari, GSFC. Development and Validation of Process Algorithms of Urbanization for Water Cycle, Data Assimilation and Climate studies Loveland Tom, USGS. The Influence of Historical and Projected Land Use and Land-Cover Changes on Land Surface Hydrology and Regional Weather and Climate Variability Sokolik Irina, Georgia Institute of Technology. Understanding the role of changes in land-use/land-cover and atmospheric dust loading and their coupling on climate change in the NEESPI study domain drylands
Integrated Regional Climate Study with a Focus on the Land-Use Land-Cover Change and Associated Changes in Hydrological Cycles in the Southeastern United States
Roger A. Pielke (Principal Investigator), Colorado State University
Figure 1: RAMS-simulated total condensate and longwave radiation fields at 03Z, June 1st 2001 over the southeast U.S. a) Vertically integrated total condensate mixing ratio (g/kg, blue shaded) and downwelling longwave flux at surface (W/ m2, red contour). b) Vertical profile of longwave heating rate (K/day, color shaded) and total condensate mixing ratio (g/ kg, contour) in the pressure coordinate. Download higher resolution image.
The land and atmospheric modeling system for this study has been developed. The land model (GEMTM-LEAF2) was integrated with NCAR’s Community Land Model (CLM) 2, and an explicit sun/shaded big-leaf scheme was introduced to account for diffuse/direct solar radiation. This new Unified Land Model (ULM) is currently implemented in the NASA’s Land Information System (LIS). The ULM-LIS system allows regional/ global offline simulation of land-surface processes at 1 km~25 km grid spacing, and will be applied to examine the change in the energy and water budget due to anthropogenic land-cover/land-use (LCLU) perturbations (Figure 2).
Figure 2: ULM-simulated daily net surface radiation (W/ m2) on April 8th 2001 on a) global, and b) regional-scale (eastern U.S.) with grid spacing of 25 km in the Land Information System. Download higher resolution image.
Simultaneously, Terra/Aqua MODIS global surface albedo products, the LCLU map, and surface skin temperature were compiled for the calibration and initialization of the ULM. This research is testing the North American Regional Reanalysis (NARR) for the lateral boundary conditions of the Regional Atmospheric Modeling System (RAMS). The use of NARR allows for implementation of RAMS at 5 km grid spacing with the explicit microphysics, while turning off the deep cumulus parameterization.
