Abstract
Abstract Climate change is intensifying extreme phenomena, and the world is increasingly vulnerable to a variety of disasters whose impacts are considerable and varied over time, from one place to another and from one community to another. Due to its geographical location, Madagascar is the most cyclone-prone country in Africa and the ninth most vulnerable country in the world. Almost every year, Madagascar is hit by cyclones, causing loss of life and property for the population. In terms of prevention, Madagascar already has an early warning system to inform the population, but during a crisis, it still lacks a decision support system for rapid, real-time intervention to minimize damage. In this paper, we propose a real-time geo-decision support system based on real-time data integration, real-time ETL and real-time cube building. In the proposed architecture, continuous data ˚ow is required for real-time data integration. The proposed real-time ETL unit is composed of the capitalization of risk analysis experiments to ensure their reusability, as well as the insertion of processing parallelization to optimize the processing time of voluminous data. The real-time SOLAP unit consists of real-time cube formation using a spatial database that stores spatio-temporal data from a given point in time, with query optimization using materialized query technology. Our prototype uses NASA’s weather data streaming service via an API. The ETL is written in a Matlab script and loads the data into a spatial database in Postgresql after processing. A web mapping application queries the constitution of a cube and displays the result for visualization.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call