Hybrid Xception-LSTM Model for Remote Sensing: Advanced Urban Heat Island and Land Use Analysis

Urban heat islands (UHI) and land use changes are critical environmental issues affecting urban areas, contributing to increased temperatures, reduced air quality, and adverse health effects. This study proposes a Hybrid Xception-LSTM Model to address these challenges by leveraging both spatial and temporal data for accurate UHI and land use analysis using remote sensing imagery. The Xception model, pre-trained on ImageNet, is fine-tuned for spatial feature extraction, focusing on land surface reflectance, vegetation indices, and surface temperature data. The LSTM layer captures long-term dependencies in temporal sequences to enhance prediction accuracy. This hybrid model was evaluated against nine existing models, including CNN, ResNet50, and EfficientNet. The proposed model achieved an accuracy of 94.38%, outperforming DenseNet121 (90.22%) and EfficientNet (90.12%). In addition, the model demonstrated superior precision (0.93), recall (0.91), and F1-score (0.92) compared to other models while maintaining competitive computational efficiency with a training time of 110.6 s and inference time of 3.0 s. These results highlight the effectiveness of combining spatial and temporal analysis for UHI prediction and land use classification. The proposed approach can aid urban planners in mitigating UHI effects and making informed land use decisions.