Adjusting measured peak discharges from an urbanizing watershed to reflect a stationary land use signal

A procedure to adjust gauged streamflow data from watersheds urbanized during or after their gauging period is presented. The procedure adjusts streamflow to be representative of a fixed land use condition, which may reflect current or future development conditions. Our intent is to determine what an event resulting in a peak discharge in, for example, 1950 (i.e., before urbanization) would produce on the current urban watershed. While past approaches assumed uniform spatial and temporal changes in urbanization, this study focuses on the use of geographic information systems (GIS) based methodologies for precisely locating in space and time where land use change has occurred. This information is incorporated into a hydrologic model to simulate the change in discharge as a result of changing land use conditions. In this paper, we use historical aerial photographs, GIS linked tax‐map data, and recent land use/land cover data to recreate the spatial development history of eight gauged watersheds in the Baltimore‐Washington, D. C., metropolitan area. Using our procedure to determine discharge series representative of the current urban watersheds, we found that the increase of the adjusted 2‐year discharge ranged from 16 to 70 percent compared with the measured annual maximum discharge series. For the 100‐year discharge the adjusted values ranged from 0 to 47 percent greater than the measured values. Additionally, relationships between the increase in flood flows and four measures of urbanization (increase in urban land, decrease in forested land, increase in high‐density development, and the spatial development pattern) are investigated for predicting the increase in flood flows for ungauged watersheds. Watersheds with the largest increases in flood flows typically had more extensive development in the areas far removed from the outlet. In contrast, watersheds with development located nearer to the outlet typically had the smallest increases in peak discharge.