Incorporating NDVI-Derived LAI into the Climate Version of RAMS and Its Impact on Regional Climate

In this study, a climate version of the Regional Atmospheric Modeling System (ClimRAMS) was used to investigate the sensitivity of regional climate simulations to changes in vegetation distribution in the Great Plains and Rocky Mountain regions of the United States. The evolution of vegetation phenology was assimilated into the ClimRAMS in the form of estimates of the leaf area index (LAI) derived from the normalized difference vegetation index (NDVI). Initially, two model integrations were made. In the first, the NDVI-derived vegetation distribution was used, while the second integration used the model’s ‘‘default’’ description of vegetation. The simulated near-surface climate was drastically altered by the introduction of NDVI-derived LAI, especially in the growing season, with the run in which observed LAI was assimilated producing, in general, a wetter and colder near-surface climate than the default run. A third model experiment was then carried out in which the (comparatively more homogeneous) spatial distribution of the LAI remained the same as in the ‘‘default’’ run, but the overall, domain-averaged magnitude of the LAI was reduced to be consistent with that of NDVI-derived LAI. This third run simulated a drier and warmer near-surface climate compared to the default run. Taken together, these results indicate that regional climates are indeed sensitive to seasonal changes in vegetation phenology, and that they are especially sensitive to the land surface heterogeneity associated with vegetation cover. The need to realistically represent both the spatial and temporal distribution of vegetation in regional climate models is thus highlighted, and the value of assimilating remotely sensed measures of vegetation vigor in Four-Dimensional Data Assimilation (4DDA) systems is demonstrated.