Climatological trends in severe-thunderstorm occurrences from high-resolution WRF simulations

Historical trends in hazardous convective weather such as tornadoes and damaging winds have been difficult to assess, due to the fact that the hazard observations are derived largely from eyewitness reports and/or damage. Such reports are convolved with the growth of population, changes in report procedure, development of education and awareness programs, etc. To supplement, or even as an alternative to these observations, we explore the use of high-resolution dynamical downscaling for trend reconstruction. Our basic approach involves a multi-decadal sequence of daily integrations of the Weather Research and Forecasting model, with global reanalysis data as initial and boundary conditions. Horizontal grid lengths of 4.25 km allow for explicit representation of deep convective storms and hence a compilation of their occurrence statistics over a large portion of the conterminous United States.

Our analyses exploit quantitative measures such as storm-scale updraft rotation to discriminate between severe and non-severe thunderstorm events, and thus far show that the downscaling technique is skillful in representing the spatial distribution of severe-thunderstorm occurrence. Consistency in changes in storm occurrence with modes of internal climate variability is also found. Finally, comparisons of the explicit modeling results with regionally averaged storm forcing offer additional credibility to the modeled trends. The implications of this study on identifying severe-thunderstorm trends in an anthropogenically enhanced climate are discussed.