Accurate rainfall prediction is crucial for space missions and agricultural activities, as it directly impacts mission success and food security. This study seeks to improve prediction accuracy by utilizing advanced Artificial Intelligence (AI) techniques, specifically deep learning (DL) models, to forecast periods of intense rainfall. These events are often associated with surface phenomena such as particle deposition, dust storms, and other environmental factors influenced by space weather. The dataset used includes critical features like crater slope, depth, boulder distribution, and shadow patterns, which provide valuable insights into the environmental conditions of targeted areas. While previous research has applied traditional Machine Learning (ML) techniques, including Random Forest, AdaBoost, Support Vector Machines (SVMs), and k-Nearest Neighbors (KNNs), this study explores the use of more sophisticated AI models to achieve higher prediction accuracy. Feature selection techniques are engaged to identify the most influential predictors, enhancing model performance by focusing on relevant data. Cross-validation is implemented to assess the robustness and generalizability of AI models. The results demonstrate that AI-driven techniques, particularly DL, outperform traditional ML models in capturing complex, nonlinear relationships within the data. This research highlights the transformative potential of AI-based approaches in advancing predictive analytics for extraterrestrial environments, where environmental factors are dynamic and intricate.