Many phenomena are difficult or impossible to characterize with existing slowly scanning radars, mobile or stationary. The limitations of traditional radars have long been recognizedwith respect to radar observations of severe weather and phenomena threatening to aviation such as microbursts. Storm evolution during radar scans can cause significant retrieval errors . Tornadoes, microbursts, dust-devils, hurricane boundary layer wind streaks, etc. evolve on time and space scales too short to be well observed with conventional radars. Data from the DOWs suggest that phenomena associated with tornadogenesis evolve on timescales that are unobservable with conventional scanning radars. Wind streaks in the hurricane BL evolve and translate quickly, requiring observations at least as frequently as 10 s.

We are in the process of completing a rapid-DOW using hybrid electronic/mechanical scanning with a multiple beam system to collect partial volumetric data in ~10 s. The Rapid-DOW incorporates a slotted waveguide array antenna specifically designed to produce frequency dependent beam steering. Energy from a single transmitter, producing 6 different frequencies nearly simultaneously, is emitted in 6 pencil-beams, each with a different elevation angle (the prototype has 6, we hope to expand to 12 later). Fast mechanical azimuthal scanning results in a 6-beam sweep of the sky in six seconds. A single elevation stagger permits an additional 6 elevation angles to be collected in the next six seconds, resulting in a 12 tilt volume scan in 12 seconds.

The antenna array, approximately 2.4 m on a side, will consist of 109 individual slotted waveguide elements. At any given frequency, the array will produce a 0.8° beam. The frequency of each individual beam will be independently controlled by individual synthesizers in each frequency module. By changing the frequencies of the transmitted energy, the elevation angle offsets of the individual beams can be modified. Thus, elevation angle offset sets can be tailored for boundary layer or deep convective studies. The elevation angle of any beam can be changed on a integration-time-by-integration-time basis, permitting saw-toothed or dithered scans. This would permit a better matching of horizontal and vertical observation scales.