Abstract:

There is currently a need for robust, high-resolution monitoring techniques to assess and quantify ecosystem dynamics within surface water bodies and their riparian ecosystems. This study presents a cost effective, user-friendly technique for examining the ecohydrology of stream and river corridors through the use of digital imagery. Using a simple digital camera, we captured hourly images of a small portion of a headwater Appalachian stream and adjacent floodplain. Then, we used pixel classification techniques to evaluate ecohydrologic parameters (e.g., inundation surface area, floodplain wetness, and vegetation dynamics) in each image. Results highlight the episodic nature of river floodplain connectivity, variation in surface wetness across the gradient from river to upland ecosystems, and the seasonal variability of vegetation density and health. To validate the accuracy of image-based measurements, we then compared inundation area estimates to an existing inundation model and found a high level of agreement (R2 = 0.94; NRMSE = 7.96%). Our study highlights the use of time-lapse imagery as a robust, cost-effective method to capture the dynamics of river corridors and associated ecosystem services.

Link to code: coming soon.