Recent Statistical Analyses of the Advanced Dvorak Technique (ADT) Poster

The Advanced Dvorak Technique (ADT) is an objective method designed to estimate tropical cyclone (TC) intensity from IR satellite imagery, patterned after the technique developed by Vern Dvorak. The University of Wisconsin – Madison's Cooperative Institute for Meteorological Satellite Studies (UW-CIMSS) developed the ADT as an objective guidance tool to augment the occasionally subjective Dvorak Technique. To date, the ADT has undergone a number of iterations with discrete improvements. Comprehensive statistical analyses are critical at these stages for understanding how updated versions of the ADT perform at estimating TC intensities.

Recent analysis has focused on important new upgrades, including the incorporation of passive microwave (MW) information that can be used to adjust the intensity estimates in eye formation stages. ADT intensity analyses were performed on a large sample of cases that were coincident with reconnaissance aircraft observations of MSLP and Vmax (ground truth). The baseline ADT results were compared to advanced versions, with the statistical analysis based on a balance of estimate bias, root mean squared error, and absolute error of the final Current Intensity (CI) number. Based on this analysis, a new version (v8.1.1) of the ADT has been released, with upgrades that will be summarized in the poster.

Further investigations are focusing on additional improvements to the ADT based on findings in the statistical analysis. For example, there are still high ADT errors possible in the mature stage of a TC, especially when an eye scene is only partially resolved. We are also attempting to optimize the integration of the MW data into the ADT estimates and associated history file. These studies will be highlighted in the poster presentation.