Severe Thunderstorms and Climate: The Roads to get Where we are and Where we Might be Going

Serious attempts to understand the relationship of severe thunderstorms and climate are perhaps a half century or so old, with proximity sounding analyses being a large part of those efforts. Although those efforts were initially focused on improving forecasting, they provided insight into the severe thunderstorm/climate connection, and provided a philosphical approach for early work on severe thunderstorms and climate change that could crudely be included under the umbrella of statistical downscaling. Investigations of the impacts of greeenhouse gas warming on severe thunderstorm occurrence are on the order of a decade old and much has happened in a relatively short time. Proximity analyses in various locations around the globe support the notion that supercell thunderstorms and their associated weather threats are much more likely in a relatively rare set of environmental conditions involving large convective instability and deep tropospheric wind shear. Determining the exact threats is somewhat more problematic, in large part because of limitations of the operational severe thunderstorm event databases.

The statistical downscaling approach used with historical environments has also been applied to reanalyses and global climate model output. The models are in general agreement that, on average, convective available potential energy (CAPE) will increase and 0-6 km wind shear will decrease, but that the increase in CAPE means that, most likely, on average there will be more envrionments supportive of severe thunderstorms in the future. However, both analyses of historical reports and environments, as well as model projections, indicate an increase in variability of severe convective storms over the US in the past and future. In addition to statistical downscaling, dynamic downscaling using high resolution models driven by climate models have also shown a likely increase in the mean with an increase in variability.

Looking at individual hazards is somewhat more challenging. There is some support for a change in tornado occurrence similar to the changes in overall severe thunderstorms (increase in mean and variability). Note that this is not necessarily a guarantee, given that the fraction of severe thunderstorms that are tornadic could change in time. It is also the most difficult to detect observationally, because of the relatively small number of reports.

I'll try to review where we've been and where I think we may be going and the roadblocks that could exist along the way, both in the way we deal with historical events as well as future events. I apologize here and in the talk for whatever misrepresentations of the work of other people I make, as well as lack of understanding of my own work. Caveat auditor!