Orographic events that are characteristic of organized thunderstorms and convection that produce heavy rainfall and catastrophic flooding in complex terrain present a great challenge for forecasters. Accurate predictions of flooding resulting from this type of storm require high resolution rainfall data. In a forecast mode, output from Numerical Weather Prediction (NWP) models must be used to drive the hydrologic models. Although much progress has been made in the past decade, the output from NWP models remains at a coarser resolution than what is needed for hydrologic predictions. Bridging the scale gap between precipitation forecasts from NWP models and the resolution needs of hydrologic models for streamflow prediction requires alternative methods such as statistical downscaling of the rainfall fields. The overall goal of this research is to improve streamflow prediction in alpine watersheds through better understanding of the orographic influences on the space-time scaling of heavy convective rainfall in mountainous terrain. In this paper, multi-scale statistical analysis is used to characterize the spatial organization of several heavy convective rainfall events in the Appalachian region and Front Range of the Rocky Mountains. Focus is placed on linking changes in the multi-scale statistical properties with orographic influences on the rainfall and developing predictive relationships between multi-scale parameters and meteorological and topographic forcings. Differences in geographic region and predominant orographic controls (e.g., windward versus leeward forcing) on trends in multi-scale properties of precipitation is investigated. A hydrological model is applied to explore the potential effects of using statistical downscaled rainfall versus coarser NWP model-predicted rainfall on streamflow predictions in alpine watersheds.