Sensitivity of storm-scale predictions to initialization with simulated Doppler radar data

With the ongoing implementation of a nationwide network of WSR-88D Doppler radars, efforts to find suitable methods for initializing cloud-scale models with Doppler radar-derived fields have intensified. Before conducting extensive experiments with real data, it is important to gather as much information as possible from Observing System Simulation Experiments (OSSE's) in which model-generated ″observations″ are used in place of actual radar data. Although OSSE's provide overly optimistic results, they allow us to systematically evaluate the sensitivity of a given assimilation scheme to a variety of controllable parameters. In this paper, we present the results of a number of simple OSSE's designed to test the sensitivity of a highly dealized assimilation scheme to 1) the strom type, 2) the period in the storm evolution and 3) the error characteristics of the initialization data.