As much effort as has gone into evaluating space-borne precipitation estimates, it is the utility of these estimates for a particular application that determines their value. The speaker will describe a new approach to assessing the usefulness of satellite rainfall for flood forecasting. As river basins are natural “integrators” of water flow processes, it is important to understand how the variability and the uncertainty are affected by spatial and temporal scales. The new approach to addressing this problem is based on the empirical observation that peak flows follow power laws. These power laws have their origin in the self-similarity of river drainage network. The speaker will present results of simulations based on landscape conceptualization that preserves the self-similar structure of real river networks and allows studying the effects of rainfall variability and uncertainty, runoff generating mechanisms, and channel routing of the flood waves. He will include examples from the recent Iowa floods.