A Comprehensive Review of Environmental Weather Monitoring and Prediction System Using IoT and Multi-Model Progressive Dense Self-Attention

Abstract

Environmental weather monitoring and prediction systems are increasingly vital for applications such as agriculture, disaster management, smart cities, and climate analysis. Traditional approaches based on numerical and statistical models often struggle to capture complex spatiotemporal dependencies in environmental data. The integration of Internet of Things (IoT) and Artificial Intelligence (AI) has enabled real-time data collection and improved forecasting accuracy. This paper reviews IoT-based weather monitoring systems using advanced deep learning and self-attention mechanisms, focusing on the Multi-Model Progressive Dense Self-Attention Network. IoT sensors collect large-scale environmental data such as temperature, humidity, pressure, rainfall, and wind speed, requiring models capable of handling both spatial and temporal relationships. Deep learning techniques like CNNs, LSTMs, and hybrid models enhance prediction accuracy, while transformer-based architectures effectively capture long-range dependencies. Multimodal approaches integrating sensor, satellite, and meteorological data further improve performance. Despite these advancements, challenges such as computational complexity, data heterogeneity, scalability, and real-time implementation remain. This review highlights recent developments and identifies future research directions for efficient and intelligent weather prediction systems.