Quantifying streambank erosion: A comparison of physical surveys, aerial imagery and UAS LiDAR surveys

Excessive or accelerated streambank erosion can harm stream ecosystems and negatively impact water supply and infrastructure systems. Streambank erosion can be the most significant source of in-stream sediment loads and associated contaminants. Site-specific, low-cost and timesaving methods to quantify active erosion rates are needed to identify and prioritize locations for restoration actions focused on reducing sediment loads and improving stream functions. Therefore, we examined several methods for quantifying streambank erosion at smaller spatial and temporal scales that are typical of most restoration projects. Physical surveys, aerial imagery analysis and Unmanned Aircraft System (UAS) based light detection and ranging (LiDAR) surveys were conducted to document erosion at three streams in the Blue Ridge and Ridge and Valley regions of Virginia. Three methods were used to quantify bank retreat: 1) cross-section (XS) surveys, 2) top-of-bank (TOB) surveys and 3) aerial imagery analysis. The bank retreat rates were compared to rates estimated from aerial images collected during 2007 to 2019. All three methods were then paired with field-measured bank heights to estimate the volume of erosion. Reachwide erosion volumes were also estimated with digital elevation models (DEM) of Difference (DoD) using LiDAR data collected with an UAS. The estimated eroded sediment volumes varied widely across all methods but were of a similar magnitude. DoD produced the lowest estimated sediment loads, highest uncertainty and was statistically different from the average of the maximum erosion measured at all cross-sections. The volume of streambank erosion ranged from 0.18 to 1.26 m3/m/yr across all sites for all methods excluding DoD. Based on the small size of our study streams and the short monitoring period (one to two years), DoD was the least suitable method. Selecting the appropriate method for documenting and quantifying erosion depends on the time available, the purpose of estimating erosion, the resolution required and stream conditions.