A Survey on Weather Prediction using Big Data and Machine Learning Techniques

From the different studies of recent papers using Google Scholar, the study examined in 2018 the most relevant scientific articles published which are 500 that dealt with the field of numerical weather prediction and climate with machine learning methods. In the abstracts, the most common subjects of interest were accepted, and in between them, a few were investigated further: wind energy and photovoltaic, physics of atmospheric and development of numerical weather prediction; parameterizations, severe occurrences, and change of climate in climate research. The precipitation, pressure, radiation, temperature, and wind, like meteorological fields, Random Forests, Artificial Neural Networks, Deep Learning, XGBoost-like methods, and Support Vector Machine were discussed. Further, Australia, India, China, United States, and Germany like countries were there to extract data using the created database on these topics. To predict the future research by authors in the direction of these domains by conducting critical evaluations of the literature, machine learning with a major conclusion being approached will be a prominent element in future forecasting of the weather. This study also offers a way for doing a systematic review of the literature for big data analytical techniques in the forecasting of weather.