Assessing the Impact of Environment and Storm-Relative Uncertainty on Tropical Cyclone Intensity Changes: Katia and Ophelia (2011)

Although there is great interest in understanding tropical cyclone (TC) intensity change, there are relatively few studies that consider the problem in a realistic framework that simultaneously takes into account the TC vortex and its environment. The present study seeks to fill this gap by investigating TC intensity changes for two initially-weak tropical storms, Katia (2011) and Ophelia (2011), using high-resolution, full-physics, ensemble forecasts initialized from an ensemble Kalman filter data assimilation scheme. Here, the differences in the initial conditions between the ensemble members are consistent with analysis errors, yet the five-day intensity forecasts for these two cases showed large variability (i.e., depression to category 3 TC).

For each storm, two subsets of twelve members each were selected for comparison: one subset with members that predicted the most intense TCs (named strongest members) and another subset with members that predicted the least intense TCs (named weakest members). Comparing these two subsets indicates that the strongest members had a stronger circulation and more vigorous convection than the weakest members during the first 12 h, which led to more stretching and the spin-up of the low-level vortex. In the case of TC Katia, convection was enhanced via higher water vapor in the lower troposphere over a broad area surrounding the TC. By contrast, strongest members in the Ophelia forecasts were characterized by higher water vapor close to the TC center. Additionally, upward motion seemed to be enhanced via proximity to an upper level trough located near Ophelia.

These results suggest that observations from the wind and moisture field could potentially reduce initial condition uncertainty and the variability of TC intensity forecasts, thus increasing the confidence on the forecasts. This hypothesis is tested with sensitivity experiments whereby simulated observations of the wind and moisture fields are assimilated. The impact of these simulated observations on the initial conditions and the forecast evolution is been investigated to further elucidate how this data can produce TC intensity changes within this ensemble prediction system.