Sensitivity analysis of backscatter amplitude at snow-buried corner reflectors using C-/L-band SARs

Abstract:

In Japanese alpine regions, spatial awareness of snow depth distribution is basic information to manage sustainable local society. Toward recent extremely heterogeneous snowfall events and ongoing climate changes, Japanese society will need higher spatial-resolution and temporally-dynamic snow-depth map, but ground observation sites for snow depth measurement is not adequate in the mountain side. Our study aims to grasp accurate spatial distribution of snow depth using satellite-based synthetic aperture radars (SARs).

In this research, multiple cubic corner reflectors were buried in multi-depth snow layers in Nagaoka, Japan from 2019 to 2020. Temporal variations of backscatter amplitude are monitored when a C-band SAR (Sentinel-1) and an L-band SAR (PALSAR-2) had their observations.

As a result of the analysis, the C-band SAR showed a large decrease (-15 dB) in the reflection intensity from no snow to a snow depth of 0.2 m, and then gradually decreased (-5 dB) to about 1.0 m. Although it is difficult to estimate snow depth under dense vegetation, our result suggests a possibility of quantitative evaluation of the amount of snow on the tree canopy. It might be a clue to study a SAR application for water resource management in mountain forests. On the other hand, it was suggested that the L-band SAR was less sensitive to shallow snow than the C-band.

Steel-made corner reflectors have effective heat conductivity, which melt surrounding snow layers. An automatic time-lapse camera is an important tool for bias correction of over-estimated regional snow depth based on the meteorological observation.