Precipitation Forecast with an Incorporation of Pollutants: Visual Analytic Approach

Weather forecasting is stochastic in nature since the parameters are influenced by several external factors such as pollution. Pollution is raised by human activity and potentially damages the health of all living beings. This research work aimed to reveal the relationship between the parameters of weather, pollution, and perform the medium-term precipitation (rainfall) forecast. The associations among the parameters are identified using intervening variable analysis and the most promising attributes with respect to precipitation are identified. Three forecasting models based on regression, ARIMA, and LSTM are constructed to perform medium-term forecasting. Visual analytics is the technique that enables the binding of visualization with the underlying forecast model, and the performance of the model is evaluated both in quantitative results and visual form. This research work has used weather and pollution observations of ‘Chennai’ city recorded from 2016 to 2018 are fetched from official websites.