Remote sensing of erosion and shallow water bathymetry to aid river navigation on the Colville River, Nuiqsut AK

Abstract

The Colville is the longest river (~600 km) in Arctic Alaska. Nuiqsut is an established Alaska Native community of ~400 people on the Colville River. Its residents rely heavily on the Colville for subsistence needs, however, changing river dynamics caused by accelerated bank erosion, river siltation, low water, and shifting and drying channels are causing concern and making boat travel increasingly difficult and dangerous. Recently, local residents have reported increased erosion at bluff sites along the Colville, which threatens existing infrastructure. Also reported are unexpected shallow water sections along the main channel of the Colville, limiting their access to subsistence food sources. Residents have expressed a need for monitoring erosional rates on the Colville as well as a map product that could aid in river navigation. These concerns shaped the main goals of this Thesis: 1) To use remote sensing techniques to map and quantify erosion rates and the volume of land loss at selected bluff sites along the main channel of the Colville, and to assess the suitability of automated methods of regional erosion monitoring. 2) To use optical satellite images for mapping river bathymetry and generate GIS map products that show potential shallow water sections (<2m) and poor channel connections, and to assess the feasibility of future monitoring based off our methods that rely on extracting relative water depth values from publicly available optical remote sensing images. For our erosional study we used orthomosaics from high resolution aerial photos acquired in 1955 and 1979/1982, as well as high resolution WorldView-2 images from 2015 to quantify long-term erosion rates and the cubic volume of erosion. We found that, at the selected sites, erosion rates averaged 1 to 3.5 m per year. The erosion rate remained the same at one site and increased from 1955 to 2015 at two of the four sites. We estimated the volume of land loss to be in the magnitude of 166,000 m³ to 2.5 million m³ at our largest site. We also found that estimates of erosion were comparable for manual hand-digitized and automated methods, suggesting our automated method was effective and can be extended to monitor erosion at other sites along river systems that are bordered by bluffs. For our bathymetry study we used summer 2017 scenes from three optical sensors (PlanetScope 3m, Sentinel 2 10m, and Landsat 30m) along with field measurements on the river to map shallow water bathymetry along a 45 km stretch of the Colville. We found a strong correlation (R²=0.89) between field-measured water depths and image-derived reflectance quantity (natural log ratio of green over red bands). We analyzed the two essential criteria for suitable bathymetry mapping from optical images: clear weather and clear water conditions. We expect several days (≈16) of suitable conditions during the ice-free season to facilitate reliable bathymetry mapping and remote monitoring of shallow water sites. We also discuss a relative depth mapping technique which is useful for boat navigation in the absence of ground truth measurements. We deliberately employed simple and robust empirical techniques that could serve as a basis for a fully developed river monitoring project in the near future led by local community residents. An implementation of our methods by the community, in order to develop a river depth monitoring program, would be an important step forward for the advancement of community-based science and the co-production of knowledge. Our technique may help address emerging environmental and societal issues in other regions where sufficient river navigation fosters local livelihoods.