Investigation of the Optimal Fusion Days on Cloud-Free Snow Observations over the Tibetan Plateau from the Combination of Landsat-8/9, Sentinel-2, MODIS and VIIRS

Accurate estimation of snow cover is vital for understanding climate dynamics and hydrological processes in cold regions. Recent advances in data fusion techniques have enabled the possibility of cloud-free, high-frequency snow observations by integrating data with high spatial and temporal resolution from multiple sensors. This study focuses on determining the optimal fusion days essential for accurately capturing the rapidly changing surface conditions of snow cover. We conducted a daily assessment of sensor coverage capabilities across various time-span windows, taking into account seasonal and spatial variations, and the impact of mixing high- and low-resolution sensor data on fusion accuracy and consistency. This analysis, spanning from April 2013 to September 2022, utilized extensive data from Landsat 9 OLI-2, Landsat 8 OLI, Sentinel-2A/B MSI, Terra/Aqua MODIS, and S-NPP VIIRS across the Qinghai-Tibet Plateau. Our findings indicate that a 4-day window is optimal, achieving over 90% data availability, where high-resolution data surpasses the availability of low-resolution data. Notably, optimal fusion days exhibit seasonal variations, necessitating longer fusion windows, up to 6 days or more, during summer months. Additionally, this study examined the geospatial factors influencing the optimal fusion days. This research offers a robust framework for data fusion in monitoring rapidly changing snow cover, providing valuable insights for related environmental and climatic studies.