A Comparison of Tropical Cyclone Intensity Estimates Using the Dvorak Technique Applied to 2- and 8-km Resolution Geostationary Satellite Data

Abstract The Dvorak satellite intensity estimation technique (DSIET), a technique using enhanced infrared (EIR) satellite imagery to estimate the intensity of tropical cyclones (TCs), has been used by TC warning centers around the world since 1984. Although there are several subtechniques of the DSIET, the most reliable and objective is the “eye pattern” technique. However, because higher (∼2 km) resolution satellite imagery can resolve warmer brightness temperatures in some TC eyes than the imagery used to create the DSIET (∼8-km resolution), it can potentially produce artificially higher DSIET estimates. To test this postulate, satellite data during the 2020–24 North Atlantic and eastern North Pacific hurricane seasons were used to analyze 34 major hurricanes that possessed well-defined eye features. The DSIET estimates were then compared to reconnaissance aircraft flight-level observations. Of the 237 images analyzed, 69% exhibited the same current intensity (CI) number using 2- and 8-km resolution imagery, though a considerable portion of cases (25%) produced higher CI numbers using 2-km resolution. The Wilcoxon signed-rank test shows these differences to be statistically significant ( p < 0.01), suggesting that the higher resolution of current GOES satellite imagery yields higher intensity estimates for some hurricanes with well-developed eyes. Comparison with flight-level observations shows that estimates based on 2-km EIR imagery have slightly higher error than those based on 8-km imagery for hurricanes with intensities less than ∼63 m s −1 . This has implications for operational hurricane intensity estimates and TC postanalysis, which are often guided by subjective and objective estimates of TC intensity.