Comparing CI-number/Pressure Relationships in the Western Pacific using HURSAT-ADT and Historical Reconnaissance

Most of the world's tropical cyclone intensity data are derived from variants of the Dvorak satellite analysis technique. Wide-ranging modifications to this technique have contributed to significant uncertainty and heterogeneity in the historical best track data. The western Pacific is particularly problematic because two of the warning agencies use different tables for converting Dvorak-based “Current Intensity” (CI) numbers into wind and pressure. The Joint Typhoon Warning Center uses the Dvorak table whereas the Japanese Meteorological Agency uses one developed by Koba et al. These two tables diverge for the most intense storms. The Dvorak table for an T-8.0 storm is 30 hPa more intense than Koba for pressure and at least 30 kt for winds. Aircraft reconnaissance in the western Pacific was terminated in 1987, which limits the possibilities to reconcile the Dvorak and Koba tables using modern measurement techniques. However, reconnaissance-based pressure data are available prior to that date. In this study, those data are matched with CI-numbers from the Automated Dvorak Technique applied to historical HURSAT data (HURSAT-ADT). The initial results suggest that the Koba table for pressures is more representative of the observed storms. Winds are more difficult to validate due to the lack of direct observations, even with reconnaissance. However, we will use the established Knaff–Zehr wind pressure relationship to explore these differences. We will also explore similar analysis in the Atlantic where reconnaissance is still routinely available. The results may have important implications for the analysis of tropical cyclone intensities in the western Pacific.