High spatiotemporal-resolution mapping for a seasonal erosion flooding inundation using time-series Landsat and MODIS images

Abstract

AbstractSeasonal erosion flooding events present a significant challenge for effective disaster monitoring and land degradation studies. This research addresses this challenge by harnessing the combined capabilities of time-series Landsat and MODIS images to achieve high spatiotemporal-resolution mapping of flooding during such events. The study underscores the critical importance of precise flood monitoring for disaster mitigation and informed land management. To overcome the limitations posed by the trade-off between spatial and temporal resolution in current satellite sensors, we emplyedand theflexible spatiotemporal data fusion (FSDAF) methods to produce synthetic flood images with enhanced spatiotemporal resolutions for mapping by using MODIS and Landsat data from August 29 to September 3, 2016. A comparison was made between flood maps from several post-disaster forecasts based on ground-obtained time-series images of the Tumen River flood in China. According to the FSDAF approach, the input Landsat image of March 25, 2016, and the fused results had a root mean square error (RMSE) of 0.0301, average difference of 0.001, r of 0.941, and structure similarity indexof 0.939, indicating that temporal variation data had been effectively incorporated into a forecast on August 16, 2016. Results also indicated that the FSDAF forecast values are lower than those from the actual Landsat image. The results of the study also showed that the generated images could be effectively used for flood mapping. By using our newly developed simulation model, we were able to produce a comprehensive map of the inundated areas during the event from August 29 to September 3, 2016. This shows that FSDAF holds great potential for flood prediction and study and has the potential to benefit further disaster-related land degradation by combining multi-source images to provide high temporal and spatial resolution remote sensing information.