Space-Based Rainfall Capabilities in Hurricanes Offshore and Inland

An overview of satellite algorithms for estimating extreme precipitation events from Mesoscale Convective Systems (MCSs) and landfalling hurricanes is described. Operational and experimental techniques will be presented, including infrared and microwave techniques.

For over 20 years, NESDIS has provided National Weather Service (NWS) Forecast Offices and River Forecast Centers (RFCs) with satellite precipitation estimates (SPEs) for use in heavy precipitation and flash flood situations. From 1978-1999, the production of SPEs was a manual/interactive process that only allowed one storm at a time to be analyzed. Precipitation estimates for flash floods are GOES driven products due to their high spatial and time resolution characteristics. An automatic GOES 10.7 micron algorithm was recently implemented which allows more spatial and temporal coverage of SPEs and improves their timeliness. Automation will also enable the product to be used by the NWS/OH stage 3 product (gauge + WSR 88D + satellite), hydrological flow models that are run at local RFCs, numerical weather prediction models for improved Quantitative Precipitation Forecasts (QPE), and cloud models (that will in turn provide precipitation estimates). Collaboration has already started in many of these new applications. A GOES Multi-spectral Algorithm is also being developed to improve the screening of anvil debris cirrus and to make the algorithm more robust by including rainfall estimates from more of the warm top convection.

Microwave data from SSM/I and AMSU (much more physically related but less timely than GOES) are being used to estimate rainfall from hurricanes prior to landfall (called TraP=tropical rainfall potentials). A combined microwave and GOES algorithm (from Turk at NRL, Monterey, CA) is also being tested. A regression approach will be compared with the histogram methodology and probability matching that are currently used in the Turk algorithm. The regression method would determine the best relationship between GOES IR cloud top temperature and SSM/I and AMSU rainfall estimates when these two (GOES and microwave) are coincident in time and space.

Improvements are needed in ground truth measurements so that satellite algorithms can become more mature and reliable. Calibration and validation for different types of precipitation systems must be accomplished in order to make the algorithms more robust. The above would include regional calibration for different seasons using mesoscale networks (e.g. Oklahoma Mesonet) and TRMM observations over land and water. If TRMM observations could become more accessible in real-time, they could be used (in combination with GOES) in the heavy precipitation/flash flood forecasting process.

The ultimate in diagnosing extreme precipitation events is to integrate the GOES with the SSM/I and AMSU, WSR 88D and in-situ rain gauge measurements to provide a multi-spectral/multi-sensor algorithm for estimating precipitation from all types of precipitation systems (MCSs, hurricanes, etc.).