A Web based Weather & Geoinformatics System using Multiple Linear Regression: Data Analytics Approach
Edie Boy M. Dela Cruz
Abstract:
Introduction: Accurate and timely weather forecasting is crucial for disaster preparedness, especially in flood-prone urban areas like Valenzuela City, Philippines. This study aims to develop a localized web-based weather and geoinformatics system that predicts rainfall probabilities using a data analytics approach. Methods: The system employs multiple linear regression (MLR) algorithms to analyze various meteorological parameters, including wind velocity, atmospheric pressure, temperature, and humidity. Historical weather data were mined to identify patterns and correlations influencing rainfall events. A centralized database was designed to manage the collection, storage, and accessibility of weather-related and geospatial data for key stakeholders. Results: Initial implementation of the system demonstrates the potential to produce reliable and location-specific rainfall forecasts. It also provides essential real-time information such as projected floodwater levels and the availability of nearby evacuation centers, thereby enhancing situational awareness and community responsiveness. Discussion and Conclusion: The integration of MLR-based predictive modelling within a web-enabled geoinformatics platform offers a cost-effective and scalable solution for urban disaster risk reduction. The system is expected to support local government units and community members in making informed decisions during weather-related emergencies. Future work will focus on refining prediction accuracy through machine learning enhancements and expanding coverage to other high-risk urban areas.
Keywords:
weather forecasting, geoinformatics, machine learning, multiple linear regression, data mining, disaster risk management.
Citation: Edie Boy M. Dela Cruz (2025). A Web based Weather & Geoinformatics System using Multiple Linear Regression: Data Analytics Approach. Horizon J. Hum. Soc. Sci. Res. 7 (S), 69–86. https://doi.org/10.37534/bp.jhssr.2025.v7.nS.id1295.p69-86
Download Full PDF
