USING REMOTE SENSING AND HIGH-RESOLUTION DIGITAL ELEVATION MODELS TO IDENTIFY POTENTIAL EROSIONAL HOTSPOTS ALONG RIVER CHANNELS DURING HIGH DISCHARGE STORM EVENTS

As global warming intensifies, large precipitation events and associated floods are becoming increasingly common. Channel adjustments during floods can occur by both erosion and deposition of sediment, often damaging infrastructure in the process. There is thus a need for predictive models that can help managers identify river reaches that are most prone to adjustment during storms. Because Vermont rivers flow over a mixture of bedrock and alluvial substrates, the identification of bedrock vs. alluvial channel reaches is an important first step in predicting vulnerability to channel adjustment during flood events, especially because bedrock channels are unlikely to adjust significantly, even during floods. This study develops a semi-automated approach to predicting channel substrate using a high-resolution LiDAR-derived digital elevation model (DEM). The study area is the Middlebury River in Middlebury, VT--a well-studied watershed with a wide variety of channel substrates, including reaches with documented channel adjustments during recent flooding events. Multiple metrics were considered for reference—such as channel width and drainage area—but the study utilized channel slope as a key parameter for identifying morphological variations within the Middlebury River. Using data extracted from the DEM, a power law was fit to selected slope and drainage area values for each branch in order to model idealized slope-drainage area relationships, which were then compared with measured slope-drainage area relationships. Differences in measured slope minus predicted slope (called delta-slope) are shown to predict river channel substrate. Compared with field observations, higher delta-slope values correlate with more stable, boulder rich channels or bedrock gorges; conversely the lowest delta-slope values correlate with flat, sediment rich alluvial channels. The delta-slope metric thus serves as a reliable first-order predictor of channel substrate in the Middlebury River, which in turn can be used to help identify local reaches that are most vulnerable to channel adjustment during large flood events.