Tracking winter extra-tropical cyclones based on their relative vorticity evolution and sensitivity to prior data filtering (cycloTRACK v1.0)

Abstract. In this study we present a new cyclone identification and tracking algorithm. Identification is based on a recognition pattern of enclosed contours of 850 hPa filtered relative vorticity values, while tracking is based on the minimization of a cost function. In particular, for each tracked cyclone our algorithm builds all possible tracks and finally chooses the one which presents the least differences of relative vorticity between consecutive track points. In parallel, for each track point the algorithm provides a cyclone area within which different physical diagnostics are calculated (such as pressure and wind speed). The area size is a function of the cyclone relative vorticity. To validate our approach we apply the algorithm on the Northern Hemisphere for the winters of 1989–2009. Three integrations of the algorithm were performed, each by using different filtering strengths. Using the three integrations, we assess the algorithm sensitivity to prior filtering the relative vorticity field. We show that filtering the input relative vorticity fields has an impact only on the weak cyclones, while in their majority the strong cyclones are independently detected and tracked.