A framework for the statistical analysis of large radar and lightning datasets

Computational capabilities, along with objective analysis algorithms, have progressed to the point that robust objective statistical analysis of polarimetric Doppler radar and three-dimensional total lightning information is now possible. To that end, a framework for performing statistical analysis has been developed and tested on a number of different radar and lightning datasets. This paper describes the key components of the framework, as well as a detailed demonstration and evaluation of the framework using datasets from several different field projects.

The linchpin to the framework is the processing and association of three-dimensional lightning information, environmental parameters, and aerosol observations with defined radar features (e.g., individual cells or mesoscale systems) in individual Cartesian-gridded NetCDF-format radar volumes. The processing components of the framework currently work in the most recent version of the Interactive Data Language (IDL) software, although the end-result NetCDF volumes can be used by most standard analysis software packages.

The framework is entirely modular. Support exists for several different possible lightning datasets (e.g., Los Alamos Sferics Array, U.S. Precision Lightning Network, Lightning Detection and Ranging, etc.), in addition to the standard National Lightning Detection Network ground-strike data coupled with three-dimensional Lightning Mapping Array observations. The lightning data can be objectively processed or hand analyzed. Radar data can be from a single radar or a mosaic of several radars, and can be polarimetric, Doppler (including multiple Doppler syntheses), or reflectivity only. Several algorithm options exist for the objective identification of radar features, including Thunderstorm Identification Tracking Analysis and Nowcasting (TITAN), Storm Cell Identification and Tracking (SCIT), the Nesbitt/Zipser Precipitation Feature algorithm, and an in-house TITAN/SCIT hybrid algorithm. Features even can be identified and tracked by hand. Environmental data can come from model or other objective analyses, or from a single sounding or surface station. Finally, support exists to take aerosol data from single stations, aircraft, or satellite observations.

The framework, and its capabilities in examining many outstanding research problems, will be demonstrated using several different datasets.