DSD retrieval and comparison on 3 co-located Wind Profilers in Adelaide, South Australia

Radar measurements of rainfall are an integral component of modern meteorology. The rainfall drop size distribution (DSD), in particular the evolution of the DSD, both temporally and in the descent from cloud to ground, impacts returned reflectivity and hence rainfall measurements. Increasing understanding of DSDs and their evolution is therefore very important. Vertically pointing wind profilers, primarily designed to measure horizontal winds, can be used to estimate DSDs, and are therefore useful tools in precipitation studies.

Fundamentally, any profiler capable of measuring a vertical Doppler spectrum sensitive to precipitation is a candidate for use in rainfall studies. Measurement of clear-air motions is of prime importance, as these motions must be corrected for in the measured precipitation echo. Upward (downward) vertical air motion left un-corrected results in an underestimation (overestimation) of the DSD. At lower VHF, Bragg scatter from the clear-air and Rayleigh scatter from precipitation with rain-rates greater than ~5 mm hr-1 are returned with roughly equal magnitude, and a single system can be used for DSD studies. Lower VHF retrievals are unable to resolve drops with diameters smaller than 1 mm, as the clear-air signal obscures the precipitation echo at the fall speeds which correspond to small drops. This is a significant problem, as these small drops represent a large portion of the precipitation spectrum, and therefore bias the resulting DSD. In contrast, UHF profilers have a higher sensitivity to Rayleigh scattering from precipitation, and so can retrieve drops smaller than 1 mm. However, in all but the lightest rainfall, the precipitation signal overwhelms the clear-air signal, and thus a second instrument must be used for clear-air information if DSDs are desired in moderate to heavy conditions.

The University of Adelaide, in partnership with ATRAD Pty Ltd, an Adelaide based radar design and manufacturing company, operates two 55 MHz wind profiling radars at the University field site, Buckland Park (BP). BP is located approximately 40 km north of the city of Adelaide, with no significant topography. The first profiler is a 12 kW Boundary Layer Profiler (BLP), sampling from 300 m to 8 km, utilizing the Spaced Antenna (SA) Full Correlation Analysis (FCA) technique to measure horizontal and vertical winds. The system consists of 27 Yagi antennas, arranged in 3 groups of 9 antennas on a square grid, with the centroid of each square on the vertices of a triangle. The system uses the entire array to transmit, and receives radar returns separately on the 3 sub-groups. The beamwidth is approximately 30o on receive, and narrower on transmit. The second radar is a 40 kW Stratospheric Tropospheric Profiler (STP), sampling from 500 m to 15 km, utilizing the Doppler Beam Steering technique to measure horizontal and vertical winds. This system is composed of 144 Yagi antennas, arranged on a 12 by 12 square grid. The entire array is used for both transmission and reception, with a beam width of approximately 7°. Electronic phase delays are used to steer the beam 15° from vertical in any of the cardinal directions. Wind profilers of the same build standard as the BP systems have recently been purchased by the Australian Government Bureau of Meteorology, and have been deployed across Australia as part of a new operational network. Both BP radars are capable of retrieving DSDs, and so provide an excellent basis for comparative studies of precipitation. In general, the STP is capable of retrieving smaller drops than the BLP, due to its narrower beamwidth.

Recently, ATRAD has developed a prototype 449 MHz wind profiling radar, which has been deployed at BP for testing. This test system represents the first stage in the development of a UHF profiler, scalable in both power and antenna aperture to suit user requirements. The prototype system operates at approximately 900 W, sampling from near the ground to 1 km. The system has been operated using antennas in both SA FCA and Doppler configurations. Due to its higher frequency the prototype UHF system is more sensitive to rainfall retrievals, and due to its co-location with VHF systems, provides an excellent basis for objective rainfall retrieval studies.

Rainfall drop size distribution retrievals have been collected using two VHF profilers with different beamwidths, and a UHF profiler. As these systems are co-located, they provide an excellent opportunity for comparative studies of rainfall. Retrievals on all three systems will be discussed, compared and contrasted.