Improving Precipitation Type Detection from Satellite Passive Microwave Observations
Abstract
Since the 1970's Passive Microwave (PMW) satellite precipitation products have played an invaluable role in hydrometeorological and climate applications. Along with improvements in instrument channel selection, calibration and spatial/temporal coverage, PMW precipitation algorithms have evolved into state-of-the-art retrievals used by centers around the world to meet constantly growing demand for low-latency high-quality measurements. Recent analysis has shown that, at large-scales, operational PMW retrievals meet the latest accuracy requirements for both short- and long-term precipitation estimates with a constellation of orbiting radiometers. However, over regional scales, PMW retrievals often exhibit significant biases between different meteorological regimes. These regional biases are highly correlated with differences in precipitation type (i.e. convective versus stratiform) and are due to limitations the PMW algorithms have in distinguishing between the storms of similar radiometric signatures but different morphology. In this study, we develop a new method to help address this problem.
Using the Global Precipitation Measurement (GPM) mission Microwave Imager (GMI) and Dual-frequency Precipitation Radar (DPR) observations to establish a link between changes in PMW brightness temperatures and the convective fraction within a raining scene, a method is developed to classify precipitating scenes. For two precipitation type categories (convective and stratiform), the method correctly classifies 90% of precipitation, using only the brightness temperatures of the 13 GMI channels as input. To assess the potential of this technique to reduce regime-dependent precipitation biases, it is implemented into NASA's operational GPM PMW retrieval (GPROF) by adding the resulting precipitation type classification to the observation vector (i.e. the 13-channel Tb). Including this additional precipitation type classification results in a decrease in the corresponding precipitation rate bias of approximately 50% across both convective and stratiform scenes. While shown to be consistent across different surface types, seasons and regions, this new PMW precipitation type detection model exceeds the bias-reduction potential of current methods for convective/stratiform classification by more than a factor of 2. This result suggests that this improved technique for precipitation type detection may help to significantly reduce precipitation biases in PMW retrievals between meteorological regimes.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.H43F2490P
- Keywords:
-
- 3354 Precipitation;
- ATMOSPHERIC PROCESSESDE: 3360 Remote sensing;
- ATMOSPHERIC PROCESSESDE: 1817 Extreme events;
- HYDROLOGYDE: 1847 Modeling;
- HYDROLOGY