Validation of Satellite-based Rainfall Estimates for Severe Storms (Hurricanes & Tornados)

Severe storms such as hurricanes and tornadoes cause devastating damages, almost every year, over a large area of the United States. More accurate forecasting of intensity and track of a heavy storm can help to reduce if not to prevent its damages to lives, infrastructure, and economy. Estimating accurate high resolution quantitative precipitation (QPE) for a hurricane is still a challenging problem because of its physical characteristics, especially when it is still over the ocean. Satellite imagery seems to be a valuable source of information for estimating and forecasting heavy precipitation and also floods, particularly over high mountains and oceans where the traditional ground-based gauge and radar techniques cannot be applied. To improve the capability of a rainfall retrieval algorithm to estimate and predict more accurate QPE of severe storms, the quality of the model output must be evaluated against the observations.

High (hourly 4-km x 4-km) resolution satellite infrared-based rainfall product, from the NESDIS Hydro-Estimator (HE) algorithm, has been examined against NEXRAD stage-IV measurements and rain gauge observations, for hurricanes, in this project. Three strong hurricanes: Charley, Jeanne, and Frances from the summer 2004 and also two very strong hurricanes Katrina and Wilma from the summer 2005, that caused devastating damages over Florida and New-Orleans, have been considered to be investigated. Preliminary results, demonstrate that: 1) for all hurricanes except for hurricane Charley, HE tends to underestimate rain rates especially when NEXRAD shows a heavy storm; 2) radar measurements show a few peaks of heavy rainfall for hurricanes Frances, Jeanne, and Wilma but HE shows a big underestimation at the time of the peaks; 3) HE tends to overestimate the amount of rainfall for the hurricane Charley, in general. HE rainfall product for the selected hurricanes will be validated against hourly rain gauge observations as well.