Improving Satellite-Based Rainfall Retrievals By Incorporating High-Resolution Ground Radar Network Observations (Invited Presentation)

Rainfall products derived based on satellite measurements has proven to be very useful for regional and/or global hydrologic modelling and climate studies. A number of precipitation products at multiple space and time scales have been developed using Infrared (IR) brightness temperature information observed by geostationary satellites and observations from passive microwave (PMW) sensors onboard low earth orbit satellites. However, the accuracy of satellite-based products is often limited due to the spatial and temporal sampling as well as the parametric retrieval algorithms, particularly for extreme events such as heavy rain or light rain. On the other hand, ground-based radar is more mature for quantitative precipitation estimation (QPE), especially after the implementation of dual-polarization technique and further enhanced by urban scale radar networks. Nowadays, ground radar systems form the cornerstones of national severe weather warning and forecasting infrastructure in many countries. The ground-based dual-polarization radars also serve as critical tool for validation of various space measurements and products. Since 2012, the center for Collaborative Adaptive Sensing of the Atmosphere (CASA) has been operating a high- resolution dense urban radar network in Dallas-Fort Worth (DFW) Metroplex (Chandrasekar et al. 2017). The CASA DFW QPE system is very robust, and its excellent performance has been demonstrated during several years of operation in a variety of precipitation events (Chen and Chandrasekar 2015).

In this paper, a machine learning system is introduced to improve satellite-based rainfall retrievals by incorporating the high resolution radar observations from the DFW network. In particular, the National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center has developed a morphing technique (i.e., CMORPH) is first applied to derive combined PMW-based retrievals and combined IR data from multiple satellites (Joyce et al. 2014; Xie et al. 2017). The combined PMW-based retrievals and IR data then serve as input to the proposed machine learning model. The high-quality DFW rainfall products are used as target to train the model. The model training with a large number of precipitation events is detailed. The trained model is evaluated using existing CMORPH products and surface rainfall measurements from gauge networks during independent (testing) precipitation events.

References

Chandrasekar, V., H. Chen, and B. Philips, 2017: Principles of High-Resolution Radar Network for Hazard Mitigation and Disaster Management in an Urban Environment. Journal of the Meteorological Society of Japan, (in press)

Chen, H., and V. Chandrasekar, 2015: The Quantitative Precipitation Estimation System for Dallas- Fort Worth (DFW) Urban Remote Sensing Network. Journal of Hydrology, 531, 259-271.

Joyce, R.J., J.E. Janowiak, P.A. Arkin, and P. Xie, 2004: CMORPH: A Method that Produces Global Precipitation Estimates from Passive Microwave and Infrared Data at High Spatial and Temporal Resolution. J. Hydrometeor., 5, 487-503.

Xie, P., R. Joyce, S. Wu, S. Yoo, Y. Yarosh, F. Sun, and R. Lin, 2017: Reprocessed, Bias-Corrected CMORPH Global High-Resolution Precipitation Estimates from 1998. J. Hydrometeor., 18, 1617-1641.