Assimilation of satellite-derived cloud cover into the Regional Atmospheric Model System (RAMS) and its impacts on modeled surface fields

The goal of this study is to provide an improved, high resolution, regional diagnosis of three important surface variables on the land surface energy and water balance, namely the downward short-wave and downward long-wave surface radiation fluxes, and precipitation. Cloud cover is a key parameter linking and controlling these three terms. An automatic procedure was developed to derive high-resolution (4 km x 4 km) fields of fractional cloud cover from visible band, (GOES series) geostationary satellite data using a novel tracking procedure to determine the clear-sky composite image. In our initial data assimilation studies, the surface short-wave radiation fluxes calculated by RAMS were simply replaced by the equivalent estimated values obtained by applying this high-resolution satellite-derived cloud cover in the UMD GEWEX/SRB model. However, this initial study revealed problems associated with inconsistencies between the revised solar radiation fields and the RAMS-calculated incoming long-wave radiation and precipitation fields, because modeled cloud cover remained unchanged and, consequently, these other surface fields retained their low, clear-sky values. It was recognized that the UMD GEWEX/SRB model provides an important relationship between cloud albedo, cloud optical depth and cloud water/ice. Thus, exploration was made of feasibility of directly assimilating vertically integrated cloud water/ice fields to update modeled cloud cover. This approach will not only enhance the realism of radiation scheme in RAMS, but it may also dramatically increase the model's capability to predict the location of precipitation, thus enhancing the ability of such mesoscale modeling systems to make accurate short-term forecasts of precipitation. This, in turn, would benefit flood forecasting as an associate hydrologic response. In the method adopted, the assimilated image takes the horizontal distribution of cloud from the satellite image but it retains a vertical distribution which is the area-average simulated by RAMS across the modeled domain in the time step immediately prior to cloud assimilation. Cloud assimilation is made every minute, with linear interpolation applied to derive cloud images for each minute between two GOES samples. Comparisons were made between modeled and observed data taken from the AZMET weather station network for model runs with and without cloud assimilation to demonstrate the improvement in RAMS' ability to describe surface radiation and precipitation fields. Cloud assimilation was found to substantially improve the RAMS model's ability to capture both the temporal and spatial variations in surface fields associated with observed cloud cover. The sensitivity of these comparisons to model initiation was explored by making five ensemble runs starting from different initiation. In general, RAMS with cloud assimilation technique is not sensitive to realistic perturbation of initial conditions.