Study on Rainfall Distribution in Yunnan Province Based on ARIMA Model by Big Data Computation
Kaimin Li · Academic Journal of Computing & Information Science · 2023
Taking the monthly rainfall data of 16 prefectures and cities in Yunnan Province from 2012 to 2020 as the research object, an ARIMA time series model is established based on big data calculation. First, we use the cluster analysis method to divide all regions into three categories according to the distribution characteristics of rainfall and select a representative city from each category to study the rainfall distribution. Then, according to the results of the ADF test, the method using phase and seasonal differences is used to eliminate the non-stationary and seasonal trends of the series. Finally, the ARIMA prediction model of rainfall distribution in Kunming, Dali, and Pu'er is obtained by combining the sequence autocorrelation and partial autocorrelation analysis diagram to determine the values of various parameters in the ARIMA model. The prediction accuracy of the model is high, and the residual sequence is a white noise sequence, which has a good fitting effect. It effectively predicts the fluctuation law of rainfall and provides an early warning mechanism for drought, flood, debris flow, and other disasters.