A Proposed Adjustment for the Advanced Dvorak Technique during Extratropical Transition

During extratropical transition, the empirical relationships between cloud patterns and tropical cyclone intensity that underlie the Dvorak and Advanced Dvorak Techniques degrade as the energy source responsible for cyclone maintenance and intensification transitions from the underlying ocean surface to the background vertically-sheared flow. Consequently, tropical cyclone intensity estimates derived using the Dvorak and Advanced Dvorak Techniques become less reliable during extratropical transition.

Recent research indicates that Advanced Dvorak Technique-derived estimates of maximum sustained surface wind are, on average, weak-biased during extratropical transition. The magnitude of this weak bias is directly proportional to cyclone intensity and is maximized shortly after the onset of extratropical transition. This is primarily the result of the rapid evolution relative to the rate of cyclone weakening in scene type, to “curved band” and/or “shear,” that occurs as the transitioning tropical cyclone's cloud pattern is first significantly disrupted by the effects of vertical wind shear.

Informed by these findings, we seek to develop an adjustment to the Advanced Dvorak Technique that is applicable only during extratropical transition. This adjustment first determines whether extratropical transition has begun and, if so, applies an empirically-determined intensity adjustment.

The onset of extratropical transition is evaluated utilizing the “B,” or 900-600 hPa cross-cyclone thermal asymmetry, parameter of the cyclone phase space, as derived using NCEP Global Forecast System data. Further criteria utilized to determine extratropical transition onset include a latitude threshold informed by the global climatology of extratropical transition (poleward of 20°N or 15°S) and inferred scene types within the most recent six hours (>50% shear or >50% shear or curved band).

Once extratropical transition has begun, the Advanced Dvorak Technique-derived current intensity number, from which the tropical cyclone intensity estimate is derived, is increased by one T-number. In this fashion, the extratropical transition adjustment is applied after all other intensity constraints have been applied. The adjustment applied at the onset of extratropical transition is linearly blended backward in time by twelve hours so as to avoid a non-physical jump in the estimated intensity.

The performance of each adjustment algorithm, evaluated relative to scatterometer-derived estimates of maximum sustained surface wind, is demonstrated for a subset of 2015 Northern Hemisphere tropical cyclones. Preliminary results indicate a reduced weak bias relative to the operational Advanced Dvorak Technique during extratropical transition, albeit with substantial case-to-case variability. These will be discussed in the presentation along with future plans for the extratropical transition adjustment.