Comparative Study of Wavelet-SARIMA and EMD-SARIMA for Forecasting Daily Temperature Series

This paper aims to find a forecasting model based on the comparative study of wavelet- ARIMA and EMD-ARIMA models residual-based selection technique for temperature time series. Time series analysis is essential in studying data investigating variation predicting future trend, which we can control changes make good decisions. And most important understand trend with time. applied hybridized wavelet transform empirical mode decomposition seasonal autoregressive integrated moving average (SARIMA), combines two get better accuracy, daily central region Eritrea, Asmara. Daily was collected 30 years, from January 1, 1991, December 31, 2020. The compares WT-SARIMA EMD-SARIMA well fit model. Model techniques are determine best fits our data. AIC BIC used methods selection. uses an additional method residual In estimating accurate parameters, structure sequence had lot connection, stationary depict estimation. From this perspective, nonstationarity measurement relative performance predictive capability sample forecasts assessed. results indicate that wavelet-SARIMA more effective than other models, MATLAB soft-wire analysis.