Drought Forecasting Using Gaussian Process Regression (GPR) and Empirical Wavelet Transform (EWT)-GPR in Gua Musang
Drought forecasting is important in preparing for drought and its mitigation plan. This study focuses on the investigating the performance of Gaussian Process Regression (GPR) and Empirical Wavelet Transform-Gaussian Process Regression (EWT-GPR) in forecasting drought using Standard Precipitation Index (SPI). Daily rainfall data from Gua Musang, Kelantan from 1975 to 2008 was used in this study. SPI data of 3, 6, 9, and 12 months were then calculated using the rainfall data. EWT was employed to decompose the time series into several finite modes. EWT was used to create Intrinsic Mode Functions (IMF) which are used to create ARIMA models. The objective of this study is to compare the effectiveness of the proposed method against GPR model in the accuracy of forecasting drought in Arau, Malaysia. It was found that the proposed model performed better compared to GPR model in all SPI studied.
KeywordsForecasting Drought Gaussian Process Regression (GPR) Empirical Wavelet Transform (EWT)
The authors would like to express their deepest gratitude to Research Management Center (RMC), Universiti Teknologi Malaysia (UTM), Ministry of Higher Education (MOHE) and Ministry of Science, Technology and Innovation (MOSTI) for their financial support under Grant Vot 4F875.
- 1.Belayneh, A., Adamowski, J., Khalil, B., Ozga-Zielinski, B.: Long-term SPI drought forecasting in the Awash River Basin in Ethiopia using wavelet neural network and wavelet support vector regression models. J. Hydrol. 508, 418–429 (2014). https://doi.org/10.1016/j.jhydrol.2013.10.052CrossRefGoogle Scholar
- 5.Rasmussen, C.E.: Gaussian processes for machine learning. Presented at the (2006)Google Scholar
- 6.Guang, Y.Y.U., Hu, Z.Y.U., Liu, X.S.I.: A novel strategy for wind speed prediction in wind farm. TELKOMNIKA Indones. J. Electr. Eng. 11, 7007–7013 (2013)Google Scholar
- 8.Liu, Y.P., Wang, Y., Wang, Z.: RBF prediction model based on EMD for forecasting GPS precipitable water vapor and annual precipitation. Adv. Mater. Res. 765–767, 2830–2834 (2013). https://doi.org/10.4028/www.scientific.net/AMR.765-767.2830CrossRefGoogle Scholar
- 9.Shabri, A.: A modified EMD-ARIMA based on clustering analysis for fishery landing forecasting 10, 1719–1729 (2016). https://doi.org/10.12988/ams.2016.6389
- 14.Shaaban, A.J., Low, K.S.: Droughts in Malaysia: a look at its characteristics, impacts, related policies and management strategies. Presented at the Water and Drainage 2003 Conference (2003)Google Scholar