Evaluating the impact of satellite data density within an ensemble data assimilation approach

New satellite observations from GOES-R, CloudSat, GPM and other current and future satellite missions will have increased spatial, temporal, and spectral resolution. The issues of how to optimally utilize information from these satellite measurements are of fundamental importance. In particular, capabilities of the current data assimilation methods to effectively assimilate dense satellite observations have to be evaluated. Toward this goal we explore a general framework based on an ensemble data assimilation approach (Maximum Likelihood Ensemble Filter, MLEF) and information theory.

The MLEF framework will be applied to the Colorado State University (SCU) Regional Atmospheric Modeling System (RAMS). The impact of data density of simulated satellite observations will be evaluated. The eigenvalue spectrum of an information matrix defined in the ensemble subspace will also be examined and used to quantify information content measures (e.g., degrees of freedom for signal and entropy reduction) of satellite observations.

We will also discus possible applications of ensemble data assimilation and information theory to define strategies for combining information from multiple sensors of future satellite missions.