Fluvial geomorphometry of two large Arctic river networks mapped from Sentinel-2 satellite imagery and high-resolution ArcticDEM data

The Arctic river networks transport and store large quantities of alluvial material, and affect climate and ecosystems via biogeochemical cycles. However, the unique fluvial geomorphometry of Arctic river networks and their spatiotemporal dynamics, is poorly quantified. High-resolution satellite observations have raised prospects for studying river networks in those remote areas. We present a new automated method to map two large Arctic river networks (Colville and Anabar River Basins, with study areas of ~60,000 km² and 90,000 km², respectively), by merging Sentinel-2 multispectral satellite imagery with high resolution ArcticDEM digital elevation data. First, preliminary river networks are generated from enhanced Sentinel-2 imagery. Second, non-water features are eliminated and river gaps along remotely sensed small river channels are filled using the ArcticDEM-modeled drainage networks. Finally, continuous small rivers as narrow as 10 m are successfully delineated. Moreover, we quantify several key fluvial geomorphometric characteristics (stream order, distance upstream, river width, sinuosity, meander wavelength, slope, catchment area, bifurcation ratio, etc.) of these two Arctic river networks and reveal their unique patterns from a basin perspective. The results will raise prospects for tracking dynamic Arctic surface water processes with high spatial resolution and revealing fluvial geomorphometry of Arctic river networks.