Flood and Heat Wave Prediction using Weighted Moving Average, Anomaly Detection and K-Nearest Neighbours for the City Of Mangalore

Natural calamity is a naturally occurring event in nature that usually is out of the control of human beings. Little to nothing can be done to avoid it. Natural disasters cause damage to living beings and other structures and property. Floods and heat waves are relatively common natural calamities that affect lives. Flood is an abnormal overflow of water and heat wave is an abnormal increase in temperature.Flood and heat wave are natural calamities that claim many lives every year in India. This paper describes a method to forecast the occurrence of flood and heat wave using weather data for the city of Mangalore. In this method, a combination of weighted moving average and K-Nearest Neighbours is used on the weather data to predict floods and anomaly detection is used to detect heat wave. The prediction of flood is based on temperature, humidity and pressure parameters obtained from weather stations. The combination of K-Nearest Neighbours and weighted moving algorithms is used to forecast rain and then infer the prediction of flood. Heat wave is detected by finding anomalous temperature behaviour in the weather readings considering the temperature readings for an extended period of time.