GEOMORPHIC AND CLIMATIC INFLUENCES ON SPATIAL SCALING OF FLOODS

Hydrologic models depend on data for calibration. However, with climate and land use changes altering the hydrologic cycle, past data may not be adequate for modeling the future. Instead, modeling approaches incorporating physical watershed processes across multiple spatial scales are needed. Spatial scaling relationships result from the physical processes governing discharge operating in self-similar drainage networks. Therefore, spatial scaling relationships can be used to improve flood risk models and hydrologic prediction methods. This study investigates how scaling relationships depend on the geomorphic and climatic setting of the watershed. The relationship of discharge (Q) to drainage basin area (A) can be expressed as a power function: Q = αA^θ. Scaling exponents (θ) and coefficients (α) are compared for floods of varying magnitude across a selection of major Atlantic Coast watersheds. Comparisons are made by normalizing flood discharges to a reference area bankfull discharge for each watershed. These watersheds capture the geologic and geomorphic transitions along the Atlantic Coast from narrow bedrock-dominated river valleys to wide coastal plain watersheds. Additionally, the dominance of tropical storms and hurricanes as the primary type of hydrometeorological event that cause major floods decreases from south to north. Scaling exponents and coefficients were determined for both flood quantile estimates (e.g. 1.5-, 10-, 100-year floods) and selected hydrometeorological events (e.g. snowmelt, frontal systems, and hurricanes). Initial results indicate that southern coastal plain watersheds have lower scaling exponents (θ) than northern watersheds, but that 100-year and other large floods have higher relative magnitudes in the coastal plain rivers. Also, mid-Atlantic watersheds (e.g. Potomac River Basin) have similar flood magnitudes as the southern coastal plain watersheds, but scaling exponents are similar to northern watersheds. These differences reflect variations in both geologic/geomorphic and climatic settings. Understanding these differences can be used to improve flood models and risk assessments.