A Study on Rainfall Prediction based on Meteorological Time Series

Abstract

This study aims to present the results of the research and development project on the urban inundation prediction technology during the heavy rain period. In this study, the results of rainfall prediction using heterogeneous weather data and machine learning are presented. In the predictive analysis of univariate time series data, it was confirmed that the CNN-LSTM model showed the best performance among several deep neural network models. In the predictive analysis of multivariate time series data, it was confirmed that the ConvLSTM model showed the best performance among several deep neural network models.