Better tracking of tropical vortices

Tropical vortices are often identified using sea level pressure or lower-tropospheric vorticity. For strong vortices such as typhoons, algorithms based on these variables can easily provide an unambiguous vortex track. However, for dynamically weaker vortices such as monsoon lows and monsoon depressions, which can produce intense rainfall despite their weak winds, tracking algorithms based on these variables can often fail. This occurs because these systems typically have peak intensity several kilometers above the surface, so that surface pressure is a poor indicator of the dynamics happening at that altitude; vorticity, even when measured at the altitude of maximum wind, is a highly noisy variable that often cannot capture the smooth evolution of the broader circulation of these storms.

We developed a methodology based on the lower-tropospheric streamfunction of the horizontal wind. This captures the full rotational flow, like vorticity, but has the smoothness of pressure or geopotential height (while, unlike those variables, not making any assumption about the geostrophic nature of the flow). We track streamfunction extrema using the TempestExtremes algorithm, a computationally efficient Lagrangian feature tracker developed at the University of California Davis that can operate on the native grid of high-resolution forecast models. We optimized the criteria used in tracking to best match a hand-analyzed track dataset of monsoon lows and depressions. For more details, see our paper: Vishnu et al. 2020