Challenges in Detecting High-Arctic Shrub Expansion From Optical Remote Sensing: Implications for Albedo and Climate Forcing

Abstract

Climate change-induced shrub expansion in the Arctic feeds back on climate by reducing surface albedo. Vegetation dynamics are typically monitored by tracking the evolution of vegetation indices, such as normalized difference vegetation index (NDVI), derived from satellite imagery in processes known as greening or browning. However, detecting changes in vegetation type requires sufficient spectral variation. Here, we measured the spectral albedos (346–2,400 nm) of assemblages of prostrate vegetation and of the only erect shrub species Salix richardsonii on Bylot Island in the eastern Canadian high-Arctic to assess spectral differences among common vegetation types. The broadband albedo of S. richardsonii (0.132 ± 0.009) was lower than that of prostrate vegetation (0.166 ± 0.008). However, NDVI values showed no significant difference (0.598 ± 0.074 vs. 0.561 ± 0.021). Satellite remote sensing using NDVI with spatial resolutions from 0.5 to 30 m using Pléiades, Sentinel-2 and Landsat-8 failed to detect differences in reflectance and NDVI between prostrate vegetation and S. richardsonii. These findings suggest that long-term NDVI trend analysis may be insufficient to capture the structural vegetation shift in these climate-sensitive areas. Failure to detect erect shrub expansion in the high-Arctic may therefore omit a climate change effect which produces a surface albedo decrease of 0.03 and a local summer solar forcing of 5.8 W m⁻².