Impact of Missing Wind Observations on the Simulation of a Severe Storm Environment

Abstract

The sensitivity of a numerical simulation of a severe storm environment in the southwestern United States to a missing wind sounding is investigated. The case is the AVE-SESAME '79 storm of 10 April 1979. On that day a major outbreak of severe local storms took place in Oklahoma and Texas. Two 24 h fine-mesh forecasts wore conducted using the Drexel Limited Area Mesoscale Prediction System (LAMPS). The initial wind fields of thew two forecast were significantly different in the speed and structure of the upper-level jet around the base of a sharp trough over northwestern Mexico. This difference was caused by the deletion of one wind sounding located at the base of the trough from the initial observations used by the analysis scheme. Numerical results show profound impact of such changes on the 24 h simulations. Detailed comparisons between the experiments based on the simulation motion fields are made to provide physical understanding of the impact. The numerical experiments underline the seriousness of uncertainty in wind analyses. In particular, due to missing or inadequate observations, the consequential errors in portraying the jet streak can result in the generation of spurious vorticity in the upper troposphere. This spurious vorticity, as it moves toward warm, moist air over the Gulf States, can cause erroneous prediction of the circulation and of the organized convection. This case shows clearly that sensitivity to sparse data is strong function of the gradients that exist at observation time and other factors such as the associated topographic features and moisture patterns.