Which is a Better Predictive Model in Weather Forecasting: ARIMA or Fourier Transformation?

Time series prediction is a critical area of research with applications in various domains such as finance, weather forecasting, and stock market analysis. In this paper, we present a comparative analysis of two popular methods for time series prediction [1]: Fourier Transformation and the Autoregressive Integrated Moving Average (ARIMA) model. We took a dataset representing climate data for the state of Madhya Pradesh for a period of 2000 days. We took three analytical metrics: pressure, humidity and temperature to evaluate the performance of each method. We independently processed each metric and plotted the graph for predicted data to do the analysis. Subsequently, we implement the ARIMA model on the same dataset. We used Root Mean Squared Error test to find the accuracy for both Fourier and ARIMA model and compared the performance of the two methods based on the results of RMSE test. Our results indicate that temperature and humidity is predicted better with ARIMA model while pressure is better predicted by Fourier transformation. This study contributes to the existing body of knowledge by providing insights into the effectiveness of these methods for time series prediction, and can assist researchers and practitioners in selecting the appropriate approach for their specific needs.