A spatial analysis of channel migration in relation to channel-spanning log jams and Riparain forest cover

Abstract

Log jams play an important role in Pacific Northwest river systems, providing habitat for aquatic species, influencing sediment and organic material dynamics, and shaping channel geomorphology. Riparian forests provide a source of largewood, generally recruited through lateral channel migration that erodes the streambank. The cycle of wood recruitment, log jam formation, and channel migration is a part of a process that drives further changes to channel geomorphology while also creating erosion-resistant hard points in the floodplain that are refugia for riparian forest vegetation that further drive this cycle and river system dynamics. In the Pacific Northwest, development within floodplains since the arrival of European settlers has interfered with the floodplain largewood cycle. My study examines the influence that channelspanning log jams and riparian forest cover have on channel migration over time to better inform management of floodplain development and facilitate restoration of healthy, dynamic river systems. Satellite imagery, remote sensing data, and GIS software are used to analyze channel migration in the Deschutes River (Thurston County, Washington) in relation to channel-spanning log jam presence and riparian forest cover. Using Google Earth and USGS imagery, I completed tracings of the channel, islands, and channel-spanning log jams for incremental years between 2003 and 2021. I then compared sequential channel year overlays using a QGIS model that calculates the area of channel gain and loss, total area of both merged channels, mean forest cover, and log jam area within a set of 75-m radius study cells. A Normalized Channel Migration Index (NCMI) was then calculated using the area of gain, loss, and total area of both merged channels, and indicated the level of channel migration that occurred within the study cell between the two years of comparison. I found that NCMI values are higher in study cells where log jams are present, and in study cells where log jams are present with varying riparian forest cover. Additionally, NCMI values are highest for study cells with mean forest cover between 0% and 25% that also have a log jam present as compared to all other study groups considering these variables. The higher NCMI values suggest that channel migration is greater in reaches where log jams are present and greater still where there is little to no intact riparian forest. The results of this analysis highlight the need for adequate riparian forest cover and the channel migration capacity of log jams.