High Resolution Land Surface Modeling with the next generation Land Data Assimilation Systems

Knowledge of land surface processes is important to many real-world applications such as agricultural production, water resources management, and flood predication. The Air Force Weather Agency (AFWA) has provided the USDA and other customers global soil moisture and temperature data for the past 30 years using the agrometeorological data assimilation model (now called AGRMET), merging atmospheric data. Further, accurate initialization of land surface conditions has been shown to greatly influence and improve weather forecast model and seasonal-to-interannual climate predictions. The AFWA AGRMET model exploits real time precipitation observations and analyses, global forecast model and satellite data to generate global estimates of soil moisture, soil temperature and other land surface states at 48km spatial resolution. However, to truly address the land surface initialization and climate prediction problem, and to mitigate the errors introduced by the differences in spatial scales of models, representations of land surface conditions need to be developed at the same fine scales such as that of cloud resolving models. NASA's Goddard Space Flight Center has developed an offline land data assimilation system known as the Land Information System (LIS) capable of modeling land atmosphere interactions at spatial resolutions as fine as 1km. LIS provides a software architecture that integrates the use of the state of the art land surface models, data assimilation techniques, and high performance computing and data management tools. LIS also employs many high resolution surface parameters such as the NASA Earth Observing System (EOS)-era products. In this study we describe the development of a next generation high resolution land surface modeling and data assimilation system, combining the capabilities of LIS and AGRMET. We investigate the influence of high resolution land surface data and observations on the land surface conditions by comparing with the operational AGRMET outputs.