Shallow Precipitation Detection and Classification Using Multifrequency Radar Observations and Model Simulations

Abstract

AbstractDetection of shallow warm rainfall remains a critical source of uncertainty in remote sensing of precipitation, especially in regions of complex topographic and radiometric transitions, such as mountains and coastlines. To address this problem, a new algorithm to detect and classify shallow rainfall based on space–time dual-frequency correlation (DFC) of concurrent W- and Ka-band radar reflectivity profiles is demonstrated using ground-based observations from the Integrated Precipitation and Hydrology Experiment (IPHEx) in the Appalachian Mountains (MV), United States, and the Biogenic Aerosols–Effects on Clouds and Climate (BAECC) in Hyytiala (TMP), Finland. Detection is successful with false alarm errors of 2.64% and 4.45% for MV and TMP, respectively, corresponding to one order of magnitude improvement over the skill of operational satellite-based radar algorithms in similar conditions. Shallow rainfall is misclassified 12.5% of the time at MV, but all instances of low-level reverse orographic enhancement are detected and classified correctly. The classification errors are 8% and 17% for deep and shallow rainfall, respectively, in TMP; the latter is linked to reflectivity profiles with dark band but insufficient radar sensitivity to light rainfall ( mm h−1) remains the major source of error. The potential utility of the algorithm for satellite-based observations in mountainous regions is explored using an observing system simulation (OSS) of concurrent CloudSat Cloud Profiling Radar (CPR) and GPM Dual-Frequency Precipitation Radar (DPR) during IPHEx, and concurrent satellite observations over Borneo. The results suggest that integration of the methodology in existing regime-based classification algorithms is straightforward, and can lead to significant improvements in the detection and identification of shallow precipitation.