Open AccessDrought forecasting using new machine learning methods / Prognozowanie suszy z wykorzystaniem automatycznych samouczących się metod
Author(s) -
Anteneh Belayneh,
Jan Adamowski
Publication year - 2013
Publication title -
journal of water and land development
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.377
H-Index - 16
eISSN - 2083-4535
pISSN - 1429-7426
DOI - 10.2478/jwld-2013-0001
Subject(s) - artificial neural network , wavelet , support vector machine , precipitation , extreme learning machine , structural basin , artificial intelligence , machine learning , computer science , meteorology , geography , geology , paleontology
In order to have effective agricultural production the impacts of drought must be mitigated. An important aspect of mitigating the impacts of drought is an effective method of forecasting future drought events. In this study, three methods of forecasting short-term drought for short lead times are explored in the Awash River Basin of Ethiopia. The Standardized Precipitation Index (SPI) was the drought index chosen to represent drought in the basin. The following machine learning techniques were explored in this study: artificial neural networks (ANNs), support vector regression (SVR), and coupled wavelet-ANNs, which pre-process input data using wavelet analysis (WA). The forecast results of all three methods were compared using two performance measures (RMSE and R 2 ). The forecast results of this study indicate that the coupled wavelet neural network (WA-ANN) models were the most accurate models for forecasting SPI 3 (3-month SPI) and SPI 6 (6-month SPI) values over lead times of 1 and 3 months in the Awash River Basin in Ethiopia.