MAPPING FLUVIAL LANDFORMS AND CHANNEL CHANGE FROM HISTORICAL AERIAL PHOTOGRAPHS

In planning restoration projects, understanding the river's lateral migration processes and rates; changes in sediment storage, channel width, sinuosity, and riparian vegetation; and effects of land use change and natural disturbances is essential for ecologically effective restoration. The U.S. archive of historical aerial photographs provides a critical information base on these processes over the last 50 to 70 years, a highly relevant time scale. We developed methods for mapping fluvial landforms and boundaries in GIS from historical aerial photos and applied those methods to long reaches of three rivers in Oregon. Photos are scanned and georectified, and fluvial features are digitized. We developed a map unit classification system appropriate for black-and-white aerial photos, including the active channel boundary, sub-channel features (bars, secondary channels, etc.), and floodplain features (vegetation types, off-channel water features, human-built structures, etc.), and a geodatabase structure for this classification. A number of issues complicate interpretation and mapping from aerial photos, including photo quality, spatial accuracy, variability of flow stage, and landform identification, and we developed some partial solutions to these challenges. We present results from the Umatilla and Applegate Rivers (active gravel-bed rivers), and the Sprague River (a low-energy sand bed river), showing how the results of mapping can be used to quantify channel change and understand processes of change.