Satellite-Based inputs for Quantitative Precipitation Forecasting Over the United States Mid-Atlantic Region

Reliable Quantitative Precipitation Forecasts (QPF) are essential in extreme rainfall event for flood and river flow prediction, mitigation and monitoring. In recognition of this, the US Weather Research Program (USWRP) has identified improvement of the skill of QPF forecasts as a major goal.

Recently, QPF models with neural-network formulations have been shown to have skill competitive to traditional models. To build a forecast, these neural network QPF prediction systems require inputs from a standard set of inputs. The inputs are determined in the training phase of the prediction system.

In this paper, we will demonstrate the use of broadscale satellite-derived measures of precipitating systems as inputs into a QPF prediction system for hydrologic and flood forecasting over the Mid-Atlantic region of the United States.