A Systematic Review on the Advancements in Deep Learning for Land Use and Land Cover Change Detection and Prediction

The rapid transformation of land use and land cover (LULC) due to anthropogenic activities necessitates effective monitoring and prediction methods to inform sustainable management practices. This paper reviews the application of deep learning techniques in detecting and predicting LULC changes using remote sensing data. A systematic literature review was conducted, focusing on peer-reviewed studies published between 2015 and October 2024. The review identifies key challenges in the field, including data scarcity, computational demands, and the transferability of models across different geographical contexts. Key findings reveal that deep learning models, particularly convolutional neural networks (CNNs), significantly enhance classification accuracy and change detection performance compared to traditional methods. Innovative approaches, such as the integration of attention mechanisms and generative adversarial networks (GANs), have shown promise in addressing data limitations and improving model robustness. The review highlights the need for further research into explainable AI methods to enhance model interpretability and the development of real-time monitoring systems utilizing streaming remote sensing data. Recommendations for future research include exploring multi-source data integration, advancing transfer learning techniques, and prioritizing the development of accessible deep learning tools for practitioners. This study contributes to the growing body of knowledge on deep learning applications in LULC monitoring, offering insights for researchers and policymakers aimed at fostering sustainable land use practices.