Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{194911,
        author = {Mr V. Siva Shankar and V. Karthik Reddy and T. Vidya and Ch. Monika and S. Sai},
        title = {CNN–LSTM Hybrid Model for Coastal Weather Forecasting in Visakhapatnam},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {8113-8118},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=194911},
        abstract = {Accurate short-term weather forecasting remains critical for coastal cities such as Visakhapatnam, where cyclonic disturbances and monsoon variability significantly impact agriculture, fisheries, and disaster preparedness systems. Traditional statistical models such as ARIMA struggle to capture nonlinear multivariate interactions among meteorological variables, while standalone LSTM networks, though effective for temporal sequence modeling, may not fully exploit short- term cross-variable feature dependencies. This study proposes a CNN–LSTM hybrid architecture for next-day precipitation prediction using meteorological observations from the Visakhapatnam IMD station, including temperature, humidity, atmospheric pressure, wind speed, wind direction, and precipitation. One- dimensional convolutional layers extract multivariate patterns from 30-day sliding input sequences, followed by LSTM layers that model longer temporal dependencies for predictive inference. The proposed model is evaluated against ARIMA, standalone LSTM, and CNN-only baselines using chronological train–test splits, with performance assessed through MAE, RMSE, and heavy rainfall classification metrics. By focusing on cyclone- prone coastal Andhra Pradesh, this region-specific framework aims to enhance hyperlocal forecasting reliability while maintaining computational feasibility for agricultural and disaster management decision-support applications.},
        keywords = {CNN-LSTM, Coastal Weather Forecasting, Precipitation Prediction, Multivariate Time Series, Deep Learning, Time Series Analysis, Monsoon Variability, Hyperlocal Forecasting, IMD Weather Data},
        month = {March},
        }