An Evaluation of Advanced Dvorak Technique-Derived Intensity Estimate Errors and Biases During the Extratropical Transition of Tropical Cyclones

Owing primarily to a lack of in situ observations, both real-time and historical tropical cyclone intensity estimates during extratropical transition (ET) are largely derived using satellite-based methods such as the Dvorak Technique (DT) and Advanced Dvorak Technique (ADT). A preliminary analysis of twelve landfalling ET cases between 2005 and 2008, however, suggests that ADT-derived intensity estimates during ET are significantly weak-biased, with biases on the order of one-half to one Saffir-Simpson Hurricane Wind Scale category in magnitude. The degradation in intensity estimate quality arises primarily because the empirical relationships developed between cloud patterns and tropical cyclone intensity that underlie the DT and ADT are primarily tropical in nature and are therefore less robust during and after ET.

However, observations of maximum intensity during ET, whether over land or over water, are virtually non-existent. Consequently, while DT and ADT intensity estimates appear to be less reliable and weak-biased during ET, the precise magnitude(s) and temporal evolution(s) of these biases are unknown. To address this issue and motivate research into means by which the ADT's performance during ET may be improved, we utilize synthetic satellite imagery obtained from numerical simulations of thirty representative North Atlantic Ocean ET events between 2000 and 2012 to quantify ADT-derived intensity estimate biases during ET.

Hourly time series of minimum sea level pressure (hPa) and maximum sustained surface wind speed (kt) obtained from each numerical simulation serve as the "observed" intensity estimate for each event. ADT intensity estimates are obtained from synthetic satellite imagery. This approach allows for an internally-consistent comparison between simulated ("truth") and ADT-estimated intensity to be obtained. The performance of the ADT is evaluated by comparing the ADT-derived, synthetic satellite imagery-based intensity estimate to the model-derived maximum sustained surface wind and minimum sea level pressure. Results of this evaluation, including preliminary insight into the sensitivity of research findings to the specific numerical model configuration utilized, will be presented. Extensions of the results to the development of a new Advanced Dvorak Technique "scene type" specific to ET will be discussed.