Climate change has intensified the frequency and severity of natural calamities such as floods, cloudbursts, and landslides, making accurate, real-time weather prediction and early-warning systems an urgent public-safety priority. This paper presents the AI-Based Climate Forecast Prediction System, a comprehensive Android application that integrates Machine Learning (ML), Artificial Intelligence (AI), and cloud-based services to deliver precise live weather forecasts and intelligent disaster alerts. The proposed system is developed using Java and XML on the Android platform and leverages the OpenWeather API for real-time meteorological data acquisition (temperature, humidity, atmospheric pressure, wind speed, and precipitation) along with Firebase Realtime Database for low-latency synchronization across devices. A hybrid predictive engine combining Long ShortTerm Memory (LSTM) networks for time-series forecasting and Random Forest / XGBoost classifiers for disaster classification is trained on historical climate datasets and continuously refined through adaptive learning. Comparative experiments against six conventional models demonstrate that the proposed hybrid LSTMRandom Forest pipeline achieves an overall accuracy of 93.8 %, recall of 94.2 %, and an ROC-AUC of 0.948, significantly outperforming baseline approaches. To enhance disaster preparedness, a multi-tier notification protocol delivers geo-fenced push alerts to users and rescue teams via Firebase Cloud Messaging, with a measured mean endto-end latency of approximately 3.4 seconds. An NLP-driven AI chatbot, built on Dialogflow, supplements the platform by providing personalized safety tips and conversational emergency guidance. The system is designed to be scalable, energy-efficient, and deployable across diverse geographic regions, contributing to a measurable reduction in casualty risk through timely, actionable intelligence.

The AI-Based Climate Forecast Prediction System is an end-to-end Android-native development composite which helps in forecasting climate with accurate weather prediction as well as disaster prediction with the help of hybrid LSTM and Random Forest. With a four-tier modular architecture and a well-designed data pipeline, our system achieves an overall classification accuracy of 93.8 % and a ROC-AUC of 0.948. These results far exceed the published accuracy threshold for production-grade early-warning systems of 85 %. The average time taken from the event occurrence to alert generation is 3.4 seconds, well within the 10 second design budget. This shows that a smartphone-resident solution can be accurate and operable. Future work has several potential directions. The input feature set can be expanded to include satellite radar imagery and IoT soil-moisture telemetry. This will help in improving landslide and cloudburst predictions in hilly terrain. Moreover, substituting the Dialog flow chatbot with a transformer-based model for on-device use, such as a quantised variant of LLa MA or Phi, would enable more sophisticated conversational guidance in case of loss of-connectivity. Moreover, you can extend the system to iOS and progressive web platforms. Additionally, creating a strong connection to the existing national disaster-response infrastructure would help to establish the link between citizen alerts and action by institutions. Last but not the least, federated learning across devices would allow predictive models to keep on improving with time without compromising on users’ privacy.

[1]C. O. de Burgh-Day and T. Leeuwenburg, “Machine Learning for Numerical Weather and Climate Modelling: A Review,” EGU Preprints, 2023. [2]T. Nguyen, J. Brown, and M. Patel, “Climat eLearn: Benchmarking Machine Learning for Weather and Climate Modeling,” arXiv, 2023. [3]A. Elsayed, H. Hassan, and M. Khalil, “Leveraging Machine Learning to Enhance Climate Models: A Review,” arXiv, 2023. [4]W. Jaharabi, S. Kim, and R. Lee, “Predicting Temperature of Major Cities Using Machine Learning and Deep Learning,” arXiv, 2023. [5]X. Li, “A Comparative Study of Statistical and Machine Learning Models on Near-Real-Time Emissions Prediction,” arXiv, 2023. [6]H. M. A. Iram, S. Khan, and A. Ahmed, “An Innovative Machine Learning Technique for Weather Prediction,” IEEE Access, vol. 11, 2023. [7]S. Sharma, R. Gupta, and P. Singh, “AI and Machine Learning in Climate Change: Predictive Models,” JETIR, 2025. [8]R. Verma, A. Joshi, and N. Shah, “AI & ML Models to Predict Climate Change,” IJPREMS, 2025. [9]K. Patel, S. Mehta, and D. Shah, “Weather Prediction and Climate Analysis Using Machine Learning Algorithms,” IJERT, 2024. [10]P. Bauer, P. Dueben, and M. Chantry, “Deep Learning in Weather and Climate Prediction,” Nature Reviews Earth & Environment, 2023. [11]T. Schneider, “Harnessing AI and Computing to Advance Climate Modelling,” Nature Climate Change, 2023. [12]O. Mudele, J. Carter, and L. Smith, “High-Resolution Climate-Health Vulnerability Mapping Using Machine Learning,” Nature Communications, 2024. [13]Y. Zhang, L. Chen, and X. Wang, “AI-Based Weather Forecasting Using Deep Learning Models,” IEEE Access, 2024. [14] R. Kumar, A. Singh, and V. Mishra, “Real-Time Weather Prediction System Using Machine Learning,” IEEE Xplore, 2023. [15]S. Patel and A. Shah, “Flood Prediction Using Machine Learning Techniques,” IEEE International Conference, 2023. [16]J. Singh, R. Kaur, and M. Verma, “Landslide Prediction Using AI and IoT Integration,” IEEE Sensors Journal, 2024. [17]M. Gupta, S. Agarwal, and N. Jain, “Cloudburst Prediction Using Deep Learning Models,” IEEE Access, 2024. [18]K. Verma, R. Tiwari, and S. Dubey, “Smart Disaster Management System Using AI,” IEEE Conference, 2023. [19]A. Roy, S. Das, and P. Banerjee, “Real-Time Weather Monitoring Using IoT and AI,” IEEE Xplore, 2023. [20]D. Sharma, V. Arora, and K. Bansal, “Hybrid Machine Learning Model for Climate Forecasting,” IEEE Access, 2024. [21]L. Wang, H. Zhao, and Y. Liu, “Deep Learning-Based Weather Forecasting System,” IEEE Transactions on Geoscience, 2023. [22]P. Mehta, A. Desai, and K. Trivedi, “AI-Based Early Warning System for Natural Disasters,” IEEE Conference, 2024. [23]S. Reddy, P. Nair, and K. Iyer, “NLP-Based Chatbot for Disaster Management,” IEEE Access, 2023. [24]V. Kumar, R. Sinha, and A. Pandey, “Big Data Analytics for Climate Prediction,” IEEE Big Data Conference, 2023. [25]N. Joshi, S. Kulkarni, and P. Patil, “Predictive Analytics for Environmental Monitoring Using AI,” IEEE Access, 2024.

© 2025 International Journal of Engineering and Techniques (IJET).

Digital Object Identifier (DOI)

IJET assigns a unique DOI to every accepted article via Zenodo for persistent linking and citation. Your article’s DOI: https://doi.org/{{doi}}

Open DOI About Zenodo DOI

Close