The Smart City (SC) framework, renowned for enhancing lives and bolstering public safety, encounters challenges due to its reliance on IoT devices, contributing to electronic waste and resource consumption. Integrating weather research and a weather-smart grid into the SC framework becomes crucial to address these issues and safeguard the environment and residents’ well-being. This study proposes a novel approach, EcoSense: A Revolution in Urban Air Quality Forecasting for Smart Cities, which incorporates Bi-directional Stacked LSTM with a Weather-Smart Grid (BlaSt). BlaSt utilizes temporal aggregation and lagged features to capture temporal dependencies and trends in air quality data, considering air pollutants and meteorological factors for future air pollutant concentration prediction. The model is designed for 1-hour prediction intervals and outperforms traditional methods by leveraging feature attribute values, enhancing accuracy, and reducing computational complexity. Experimental results demonstrate the improved accuracy and computational efficiency of the BlaSt model compared to conventional models. Additionally, it effectively handles extensive air quality data and exhibits promising predictive capabilities for future data.