The most recent version of the rain-type classification algorithm for Cartesian gridded datasets is here.

The most recent version of the algorithm for use on native polar coordinate data is here.

For a detailed description, see the JTECH article introducing and motiviating the algorithm, and review the README files available at the download location. The new algorithm takes the place of the convective/stratiform classification of Steiner et al. (1995), although the code for Cartesian gridded data is very similar to that of the Steiner classification. Both versions add support for identifying isolated, often shallow convection. They also classify ambiguous echoes around convective cores as "mixed" regions based on heating profiles from WRF is such areas that suggest echoes near convective cores contain features of both convective and stratiform echoes in their heating profiles.

Many users will probably want to use the version for Cartesian gridded datasets because working with a regular grid is easier and faster. However, the polar coordinate code offers at least a few advantages. First, it takes advantage of high spatial resolution along the radial. Second, it can identify very shallow convection that may not even be identified by reflectivity fields at 2 or 2.5km above the surface using the Cartesian code. Third, the polar coordinate algorithm produces a classification of the echoes nearest the ground that were observed. Analysis of low-elevation echoes may yield more accurate estimates of surface rain rate.

An example of the rain-type categorization during a rainy day at Addu Atoll.