Recent grid verification highlights in the NWS Eastern Region

Traditionally, verification of NWS public forecasts focused on observation points. However, gridded verification approaches are now necessary given the advent of gridded forecast databases, such as the National Digital Forecast Database (NDFD). This poster will summarize recent grid verification activities in the NWS Eastern Region, with an emphasis on the application of grid verification tools and techniques through specific forecast examples.

The overall purpose of grid verification is to provide timely, flexible feedback to forecasters on the quality of gridded forecasts, both spatially and temporally, and to complement the existing national grid verification products. Given the wide variety of users and applications of NDFD products, quality improvement of gridded forecasts is important, and can be more easily accomplished with shorter latency and additional detail at the WFOs.

The Eastern Region grid verification implementation approach was phased in with a series of steps, which focused initially on forecaster training. Starting in the spring of 2008, all ER weather forecast offices (WFOs) were briefed on the need, philosophy, and techniques behind the verification of gridded weather forecast products. All ER WFOs use the AWIPS Graphical Forecast Editor program BOIVerify (Version 2.02, developed at WFO Boise) in combination with forecast grids and NCEP's Real Time Mesoscale Analysis as the software program and analysis grid, respectively, to generate statistical output for interpretation. Currently, the focus of the statistical output is mainly on the temperature, dewpoint, and wind speed element grids.

Though in the early stages, the verification of forecast grids has yielded interesting insights on forecast aspects not easily verified with points. For example, offices found that during the warm season, cloud cover from upstream convection affected temperature forecasts over large parts of the CWA. Some offices took a more “phenomenon-based” approach, and used the grid verification output to examine their forecasts associated with local sea breezes, frost/freeze warnings, or winds associated with tropical cyclone remnants or fire weather potential. Additional examples of current and planned applications of gridded verification will be presented.