Incorporating Smoke into the Sea Clutter and Chaff Classes for an Enhanced WSR-88D Hydrometeor Classification Algorithm

The dual-polarimetric Weather Surveillance Radar–1988 Doppler (WSR-88D) network is capable of determining hydrometeor types via the Hydrometeor Classification Algorithm (HCA). While there are ten possible classes, the only non-hydrometeorological options are Ground Clutter and Biologicals. Several common targets do not fall into these classes, such as sea clutter, chaff, smoke, and radio frequency interference. Recently, a new class known as Inanimate was developed based on the traditional HCA fuzzy logic technique, and is capable of isolating these four target types. However, more-advanced methods are necessary to separate these targets from each other due to significant overlap in their polarimetric statistics. It has been shown that Inanimate classifications can be separated into sea clutter, chaff, and “other” target types using a Support Vector Machine (SVM) approach coupled with a series of image processing steps. The original SVM training, however, did not include distinctly separated cases of smoke and combustion debris. This study presents a preliminary evaluation of SVM performance when incorporating smoke/combustion cases. Analysis is discussed in two respects: (1) how the original sea clutter/chaff SVM model performance changes with smoke included in the “other” target type; and (2) how a fourth target type of smoke performs as part of an expanded multi-class SVM. Results are quantified in the form of various metric evaluations, including common skill scores, precision-recall and ROC curves, and holdout/cross-fold validations.