A narrow zone of strong horizontal moisture gradient, known as a dryline, is frequently observed over portions of the Southern Great Plains of the United States. The dryline acts as a focus for severe convective storms, and often leads to flooding and tornadoes. Although most storms initiate at or near the dryline, the exact processes by which convection is triggered and the preferred location for convection along the dryline are not well understood. Because the underlying processes are highly nonlinear, past numerical weather prediction models show poor skill in their ability to accurately forecast these events.

In this research a non-convective dryline case over the Oklahoma and Texas panhandle on 22 May 2002 was considered. Using high spatial and temporal resolution observational data from the International H2O Project (IHOP), a field campaign in 2002, and the National Center for Atmospheric Research (NCAR) Weather Forecasting and Research (WRF) model, moisture evolution and variability in the boundary layer is thoroughly analyzed and investigated. Performance of the model and the possible reason why the anticipated dryline on 22 May 2002 did not trigger convective storm over Homestead – OK area are discussed. The simulation does appear to recreate the dryline event well. Results of the observational analysis indicate that abundant moisture did not sustain over Homestead – OK area during 22 May 2002. Moreover, observed vertical structure of water vapor mixing ratio indicate that moisture was not deep enough for vertically moving air parcels due to low-level convergence provide the necessary destabilization effect to support deep convection initiation during this period.