41EIPT Establishing the Airborne Phased Array Radar (APAR) Observing Simulator as a Proving Ground for Radar Scan Analysis and Algorithm Development

Airborne weather radars provide critical data for basic research, for operational reporting and analysis, and for improving future forecasts of high-impact weather events. Historically, data collected by the U. S. National Science Foundation (NSF) NCAR ELDORA radar and NOAA’s Tail Doppler Radar (TDR) have significantly advanced the science and understanding of severe, mesoscale weather systems. However, the landscape of airborne Doppler weather radars is in a state of change due to retirement of ELDORA and NOAA’s upcoming transition from a WP-3D to C-130J aircraft. The design of the C-130 aircraft prevents installation of a mechanically scanning Doppler radar on the tail of the aircraft. With this design constraint in mind, using a flat panel, phased array weather radar (PAR) would make sense for current and future C-130 platforms. PARs are becoming more common because of this tool’s unique, high resolution, and targeted scanning capabilities compared to traditional mechanically scanning radars. The technology that is required to construct such a PAR is also becoming more prevalent and more affordable. To take advantage of this unique radar technology, NSF and NCAR are in the process of developing the first airborne PAR, known as the Airborne Phased Array Radar (APAR). This new radar is designed to fly on the NSF NCAR C-130 aircraft and to provide both dual-Doppler and dual-polarization capabilities using agile scanning methods. While APAR is still in development, a tool was created at NSF NCAR that allows scientists to test various uses of APAR in a controlled setting. The APAR Observing Simulator (AOS) uses model output and user-defined radar scan and flight parameters to provide realistic radar output that would be expected with APAR’s given technical specifications.

Previous experiments with the AOS output investigated some of the APAR’s technical limitations related to flight and scan strategies as well as different applications of dual-Doppler and microphysics analyses. Evaluation of the dual-Doppler winds and their comparison with known storm characteristics confirmed that APAR could provide this information with a high level of accuracy. These experiments show that at a minimum, APAR can support objectives that were designed for mechanically scanning tail Doppler radars. Previously presented dual-polarimetric analyses highlight new information that can be combined with the kinematic information to provide a more complete picture of storm structure in terms of precipitation, microphysics, and winds.

Recent updates to the AOS have positioned this tool to serve as a testing platform for future scanning considerations and algorithm development. For this presented work, a first set of these developments are provided to showcase the utility of the AOS and to highlight enhanced capabilities of APAR. The first area of investigation centers on APAR’s representative physical beamwidth. It is known with PAR that beamwidth will increase when scanning away from the broadside of the panel. Unlike with mechanically scanning radars, this increase in beamwidth has implications for range gate volume size and possible smoothing of the radar moments. Several experiments were performed comparing a constant beamwidth scan, similar to the TDR, with the variable beamwidth condition of APAR. On a scan-by-scan evaluation level, differences for radar reflectivity are on average < 1 dB at range > 50 km or elevation > 35 degrees. Results also indicate a near zero difference at closer range and lower elevation. For radial velocity, mean differences < 0.5 m/s are found on the scan edges. When using these data in post-processing scripts, such as a 3-D wind analysis, the analysis results are nearly identical to the constant beamwidth scenario. Therefore, it is apparent that the variable beamwidth characteristic of APAR will have a low impact on the relevant data and associated analyses. A second area of investigation focuses on using the PAR agile scan design to obtain more direct measurements of vertical wind speed. Along with the discrimination of hydrometeor classifications through dual-polarization variables, there is an opportunity to quantify error reduction in vertical wind speed estimates. Several possible algorithm variations are presented with statistical details on improvements over current methods of estimation from airborne radar. Results suggest that using information from the dual-polarization variables confines the terminal fall speed estimate, and applying this information through updated algorithms improves estimates of vertical wind speed by as much as 30-40%. While more investigation is planned, this initial study of improved vertical wind speed estimates serves as an example of the utility of the AOS in support of APAR's development.