Quantitative Precipitation Estimation by X-Band Dual-Polarization Radars in Complex Terrain over the Bay Area in California

Although the San Francisco Bay Area has coverage from the Next-Generation Weather Radar (NEXRAD) operational radar network, topographic influences provide a unique challenge for Quantitative Precipitation Estimation (QPE) in this region. The poor performance of NEXRAD radars can be associated to several factors, such as radar beam blockage, radar beam overshooting, discontinuity in vertical reflectivity profile (VPR) and anomalous propagation etc. which limit the accuracy of rainfall measurement. Often times, storms developing in mountainous regions are affected by orographic enhancement which in turn affects the intensity and spatial variability of precipitation. This creates a challenging task for obtaining accurate QPE from radar observations. To address these issues, X-Band dual polarization radars have been deployed in the Bay Area to augment NEXRAD’s coverage and aid in better monitoring of precipitation events as well as to better understand precipitation microphysics. This paper presents the methodology for rainfall estimation and QPE comparisons against rain gauges and operational products demonstrating the added value of the X-Band radars to the existing operational system.

The main goal of this study is to determine the best QPE in this region and to further improve its accuracy. Several dual-pol radar QPE algorithms are analyzed, such as the CSU-DROPS and a Z-R relationship which uses differential reflectivity (ZDR) to discriminate bright band from non-bright band stratiform rain. The algorithms are evaluated and cross-validated against in-situ observations. The gauge-based Mountain Mapper and Gauge Corrected QPE products from the Multi Radar-Multi Sensor (MRMS) suite are used for cross validation. Rain gauges operated by the NOAA Physical Science Division (PSD) and the Sonoma County Water Agency (SCWA) operated gauges are also used for assessing the radar-based QPE product. These rain gauges are independent from the gauges used in generation of MRMS QPE. Several precipitation events from January 2019 to April 2019 are evaluated. Hourly rainfall accumulations are calculated at the pixels corresponding to the rain gauge locations and various statistical scores are presented. Results show that the performance of NEXRAD product and MRMS product are highly variable from case to case. Overall, the performance of the X-Band radar QPE is more robust and is able to better capture the precipitation variability both during a specific event and across different storms.