Recent Improvements to the GOES Microburst Products

Abstract

The downburst is defined as a strong downdraft produced by a deep convective storm that induces strong or damaging winds on or near the earth's surface. Because of the intense wind shear they produce, downbursts are a hazard to aircraft in flight, especially during takeoff and landing phases. Retrieved profiles of temperature and moisture obtained from the Geostationary Operational Environmental Satellite (GOES) sounders have been shown to be useful in assessing the potential for convective downbursts. Sounder-derived parameters examined in this paper include the wind index (WINDEX), used to estimate maximum wind gusts; a dry microburst index (DMI), used to estimate dry microburst potential; and maximum theta-e deficit (TeD), used to estimate wet microburst potential. Currently under development is a new wet microburst index that will summarize the physical processes of convective storm development and downburst generation to quantify the potential severity of convective wind gusts. The experimental indices are plotted on regional GOES images (visible, infrared, or water vapor) and are made available on the GOES microburst products Web page. This paper briefly reviews the development of each of the GOES microburst products, describes recent improvements, provides updated validation data and a case study, and discusses future plans. Recent improvements in the processing of sounding data to generate the microburst products include a change in the first-guess numerical model, use of single field-of-view retrievals, a filter for removing high DMI values where convection is unlikely, and a change in the calculation of nighttime WINDEX values to reduce a nighttime low bias. Improvements to the display of the microburst products include animation of many sectorized products, color coding of TeD and WINDEX ranges for data plots, plotting of numerical values of WINDEX instead of color-coded boxes, and the plotting of Storm Prediction Center (SPC) severe weather reports. Validation continues by comparing product output values to preliminary severe weather reports from the SPC as well as surface observations. Mean absolute error was <2 kt (1 m s?1) for 43 daytime events during summer 2002, a significant improvement over a mean absolute error of 3 kt (1.5 m s?1) for the 2001 convective season. A marked reduction in mean error for nighttime events was noted, improving from>6 kt (3 m s?1) for the 2001 convective season to 4 kt (2 m s?1) for summer 2002. A case study is presented that discusses the improved performance of the WINDEX during a nighttime convection event in the central plains.