Land Use and Land Cover Change Detection in the Gaza Strip Using Sentinel-2: Integrating Enhanced Lee Filter with Artificial Neural Networks

This study employs Remote Sensing (RS) and Geographic Information Systems (GIS) technology to analyse the transformations that occurred from September 2023 to May 2025. Despite the limited time available for observation, research indicates that there have been substantial alterations in land cover. The identification of change patterns will be facilitated by Sentinel-2 images. We have implemented supervised classification techniques to divide the land into five distinct categories: water, built-up, barren, vegetation, and rubble. For the classification procedure, we incorporate the Enhanced Lee Filter and the Multi-Layer Perceptron (MLP) neural network. Lee filters are typically employed with SAR data; however, this study will implement the Enhanced Lee filter on optical datasets. This approach provides valuable insights into the environmental repercussions of the encounter and allows for the quantification of the extent of these modifications. The quantitative accuracy evaluation of this research was supplemented with damage estimates from the United Nations Satellite Centre (UNOSAT) as a result of the challenges associated with obtaining real-world data in a combat zone. The results indicate a substantial increase in desolate land, with a 57.5% increase. Conversely, the vegetation and built-up areas experienced a 59% and 71.5% decline, respectively. The detection of minor damage in areas that were covered by rubble proved to be a difficult task. The spatial resolution of the Sentinel-2 satellite was the primary factor contributing to this challenge. Nevertheless, a vast expanse of debris zones has been identified, with a total area of 68 km² that was completely demolished.