Real-Time Detection of Weather-based Disasters

Every year a large number of people and organizations are affected by weather-based disasters like cyclones, hurricanes, heatwaves, etc resulting in adverse effects on life and property. In order to alleviate the repercussions and foster better preparedness towards these disasters, it is imperative for a system to be developed that accurately predicts the aberrant nature of weather and helps to portend the onset of weather-based disasters. The users will be able to ubiquitously track the weather conditions and the probability of a weather-based disaster at any given place. The probability of formation, intensity & impact prediction is carried out using a multimodal ensemble system applying optimized models for each segment. The users can get an insight into the possible impact that disaster will have in real-time at a macro and micro level, keeping the users updated on the severity of weather conditions at all times. A real-time track prediction will also be made along with a plethora of plots and graphs in order to give the customers a better visual insight. The main objective of the proposed work is to provide real-time updates, and best performance results to give the users the best shot at tackling any impending disasters.