Global Convective Environments: Multiple Reanalysis Climatology

A global climatology of severe thunderstorm-related environments has been developed from three reanalyses (the Modern Retrospective Reanalysis for Research and Analysis, the ERA-Interim Reanalysis, and the NCEP-NCAR Reanalysis) using a consistent framework for calculation of parameters. Thermodynamic and kinematic parameters, together with composite metrics are explored to assess how different indices can be used to describe the global climatology of severe storms. Typically, analysis of convective environments is performed using a single reanalysis dataset, and thus is subject to the limitations of that reanalysis, rather than providing a comprehensive picture. Each reanalysis is derived to provide an accurate rendition of the instantaneous climate state, rather than the continuity of observations that would be necessary to produce a consistent climate analysis record.As the available observations are not consistent throughout the historical record, it is not atypical to see break-points in derived parameters associated with these changes. Furthermore, the forecast models used to produce the various reanalyses can also have a marked influence on the resulting analyzed state, including sensitivities to convective parameterization and boundary layer schemes.

This work is motivated by the disproportionate share of research focusing on the United States (U.S.) for convectively favorable environments. This regional focus suggests it is necessary to explore the efficacy of parameters derived for the U.S. in representing severe convection, hail and tornadoes outside of the domain of their formulation. Key requirements for using reanalyses for deriving a global convective climatology include whether different reanalyses can adequately capture the varying convective environments between different continents, and should the ‘best’ reanalysis be used in a deterministic fashion to explore the global distribution of environments or should instead be considered a cheap multi-reanalysis ensemble. In this presentation, the capability of the reanalyses to render global convective environments will be explored. Preliminary results suggest regional differences in the parameters most effective at identifying convective events, as well as regional limitations between the various reanalysis products. These factors suggest that care must be taken to ensure that parameters, particularly those with calibration for a given region, are applied with the knowledge that they are unlikely to capture all favorable environments.