Multi-Scale Statistical Properties of Rainfall for Extreme Hydrometeorological Events in Mountainous Regions

Hydrometeorological events that produce heavy rainfall and catastrophic flooding in mountainous regions 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. This study quantifies the multi-scale statistical properties of rainfall for extreme hydrometeorological events in mountainous regions across scales of 1~20 km. The Buffalo Creek flood of 1996, Fort Collins flood of 1997, and several other extreme hydrometerological events in the Appalachian region and Front Range of the Rocky Mountains are included in the analysis. The following questions will be investigated: (1) does spatial scaling exist as a common feature in convective rainfall events in mountainous regions?, (2) at what spatial scales do meteorological and topographic controls manifest themselves in the space-time variability of the rainfall fields?, and (3) how does meteorological forcings and geographic location impact trends in topographic influences on the multi-scale statistical properties of rainfall? 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 rainfall is investigated. The potential for developing a statistical downscaling model that dynamically accounts for the interplay between meteorological forcings and topographic influences on the scaling behavior of the rainfall fields will be assessed.