Wet and Dry Snow Detection Using Sentinel-1 SAR Data for Mountainous Areas with a Machine Learning Technique

Abstract

Traditional studies on mapping wet snow cover extent (SCE) often feature limitations, especially in vegetated and mountainous areas. The aim of this study is to propose a new total and wet SCE mapping strategy based on freely accessible spaceborne synthetic aperture radar (SAR) data. The approach is transferable on a global scale as well as for different land cover types (including densely vegetated forest and agricultural regions), and is based on the use of backscattering coefficient, interferometric SAR coherence, and polarimetric parameters. Furthermore, four topographical factors were included in the simple tuning of random forest-based land cover type-dependent classification strategy. Results showed the classification accuracy was above 0.75, with an F-measure higher than 0.70, in all five selected regions of interest located around globally distributed mountain ranges. Whilst excluding forest-type land cover classes, the accuracy and F-measure increases to 0.80 and 0.75. In cross-location model set, the accuracy can also be maintained at 0.80 with non-forest accuracy up to 0.85. It has been found that the elevation and polarimetric parameters are the most critical factors, and that the quality of land cover information would also affect the subsequent mapping reliability. In conclusion, through comprehensive validation using optical satellite and in-situ data, our land cover-dependent total SCE mapping approach has been confirmed to be robustly applicable, and the holistic SCE map for different months were eventually derived.

A Combination of PROBA-V/MODIS-Based Products with Sentinel-1 SAR Data for Detecting Wet and Dry Snow Cover in Mountainous Areas

“…In addition to the temperature information, vegetation indexes might also help to detect SCE especially in forested regions. In a previous study, we found [18] that the densely vegetated regions tend to show lower SCE mapping accuracy when compared to, e.g., bare area, sparse vegetation, and grassland. Moreover, previous studies also suggested that SAR backscatter, InSAR coherence, and PolSAR parameters are all related to vegetation types and conditions [23,[32][33][34].…”

“…For PolSAR parameters, the polarimetric matrix was first constructed from the calibrated and debursted SLC images and the derived eigenvalues and eigenvectors were then used for calculating H/A/α parameters [20]. For a more detailed description of the SAR pre-processing, we refer to Tsai et al (2019) [18].…”

“…The former was employed to calculate the topographical factors (elevation, slope, aspect, and curvature). The CCI land cover product was reclassified from 37 to 16 classes to reduce redundancy [18]. The daily and global SCE derived from DLR's Global SnowPack (GSP) [35] was employed as ground truth for training the model, which has been validated using in situ and higher resolution reference data for an accuracy of approximately 80% globally [35].…”

“…Dry SCE, however, cannot be detected by this approach alone. Because information about the total (wet + dry) SCE is critical for further applications (e.g., hydrological modeling), interferometric SAR (InSAR) as well as polarimetric SAR (PolSAR) have been investigated [18]. InSAR coherence enables total SCE mapping based on the fact that the snow-covered areas would show lower coherence due to a change in the SAR signal penetration depth [19], and PolSAR decomposed parameters enable dry/wet/total SCE mapping as they can reveal the scattering mechanism of different ground features [20].…”

“…In a previous study [18], we proposed a globally transferable, freely accessible data-only, all land cover class-applicable total and wet SCE mapping strategy utilizing only SAR-based observations (backscattering, InSAR coherence, and PolSAR parameters) as well as topographical factors. Based on a comprehensive validation approach, an overall accuracy of approximately 80% was confirmed.…”