Climatological studies with Doppler Radar

A simple 3-dimensional analysis technique has been developed for large Doppler radar data sets, aimed at facilitating statistical (climatological) studies of radar returns and their relationships to other measurements at the site.

Hourly histograms of observables are kept in a horizontally coarse (8km range by 30 degrees azimuth) but vertically fine (500m) cylindrical grid. This facilitates nimble exploration of the whole data set, by greatly reducing its volume, while small-scale structure is sacrificed. Interpolation to a square grid is avoided: in essence, the radar is viewed as an eccentric but efficient sampling engine, rather than as a fundamentally flawed Cartesian mapping tool. Calibration adjustments can be performed without reprocessing the original data.

Cylindrical geometry facilitates velocity-azimuth display (VAD) estimation of horizontal wind divergence, indicative of the interplay of atmospheric dynamics and heating processes. This can frequently be done even in sparse echo fields, using echoes of opportunity. Sampling noise can be reduced with large data sets. Mean wind profile estimates can also be derived.

As a pilot study we have analyzed 6 months of shipborne data from special field deployments in the tropics. We hope to analyze much larger amounts of data, and intend to seek out additional data sets at the meeting. Analysis products will be offered to the community by Web.