Characterizing tornadoes in multiple-Doppler radar data using a low-order model

Our goal is to detect and characterize tornadoes in multiple-Doppler radar data by fitting a low-order analytical model of a tornado to the data. The low order tornado model is a superposition of four idealized analytical flow types: a uniform flow, a linear shear flow, a linear divergence flow, and a Rankine vortex. The latter three flows are allowed to translate at different speeds and directions. There are a total of 19 free parameters including vortex location and propagation speed. A cost-functional J, designed to account for the discrepancy between model and observations, is defined by projecting the model wind in the direction of the radar(s) to obtain the measured radial wind, which is then compared to actual radial wind. This cost-functional is evaluated over both space and time so that observations can be used at the time they were acquired, thus bypassing the need for time interpolation, moving reference frames or other ad-hoc procedures. The parameters in this low order model are determined by minimizing J.

We are initially testing this model using an ARPS (Advanced Regional Prediction System) dataset of a tornado as our set of observations. The model will subsequently be tested using real data.