Using dynamic and thermodynamic profiler retrievals to improve the forecast of the May 24, 2011 outbreak: an observing systems simulation experiment

Description

In 2009, the National Research Council issued a report recommending a nationwide mesoscale network of ground-based instruments for measuring both momentum and thermodynamic variables within the atmospheric boundary layer. On May 24, 2011, during the Midlatitude Continental Convective Clouds Experiment, a line of severe storms passed through northern Texas, Oklahoma, and southern Kansas. A unique network of four atmospheric emitted radiance interferometers (AERIs) in the MC3E domain collected a high time resolution set of thermodynamic profiles during this event. With these observations, a unique opportunity exists to examine the impact a network of the thermodynamic instruments might have assisted forecasters in predicting the location and timing of convection initiation on May 24, as well as the storm motions that ensued. In this study, we perform an Observing Systems Simulation Experiment using a Weather Research and Forecasting (WRF) Model simulation at 1-kilometer resolution to serve as truth, and from which we generate observations to be assimilated into a coarser 4-kilometer resolution experimental run using a three dimensional variational (3DVar) framework, with different physics parameterization schemes to avoid the ``perfect model" scenario. The experiments performed vary observation densities and lead times before moving from analysis into forecasting mode. We compare the impacts of these factors on the location and timing of model-forecasted convection initiation against the true convection initiation, as well as differences in general storm motion.