Fine Spatial and Temporal Ice/Snow Surface Temperature Generation: Evaluation Spatiotemporal Fusion Methods in Greenland Ice Sheet

Monitoring ice/snow surface temperature (IST) variations with high spatial and temporal resolution data from satellites are essential for research on the mass balance of the Greenland ice sheet (GrIS). However, the tradeoff between satellite sensors' bandwidth and re-entry cycle, coupled with the influence of cloudy weather, limits their ability to fine-monitor IST. Spatiotemporal data fusion is a way of producing high spatiotemporal datasets. This article uses four spatiotemporal fusion algorithms to fuse the Landsat 8 IST data and the Moderate Resolution Imaging Spectrometer IST to generate fine spatial-temporal IST in the GrIS regions. The quantitative evaluation of the different fusion data shows that the R² are all above 0.9. The spatial and temporal nonlocal filter based fusion model (STNLFFM) dual-temporal algorithm provided the highest accuracy with a root mean square error of 2.427 K, followed by the STNLFFM mono-temporal algorithm, the spatial and temporal adaptive reflectance fusion model (STARFM), the flexible spatiotemporal data fusion model, and enhanced STARFM. From the results, the fusion data are accurate and detailed in different regions. That is, the spatiotemporal fusion technique has the potential to generate IST datasets that possess high spatial and temporal resolutions for Greenland.