Drought Prediction Based on Artificial Neural Network and Support Vector Machine

Drought Prediction Based on Artificial Neural Network and Support Vector Machine

Drought has aggravated in the humid areas of South China due to climate warming.Drought prediction is of great significance for the optimal management of water resources and the alleviation of drought.Based on the standardized precipitation evapotranspiration index (SPEI) of different time scales fo...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHAO Guoyang, TU Xinjun, WANG Tian, XIE Yuting, MO Xiaomei
Format: Article
Language:zho
Published: Editorial Office of Pearl River 2021-01-01
Series:Renmin Zhujiang
Subjects:
drought prediction
ANN
SVR
EMD
wavelet denoising
Dongjiang River Basin
Online Access:http://www.renminzhujiang.cn/thesisDetails#10.3969/j.issn.1001-9235.2021.04.001
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Drought has aggravated in the humid areas of South China due to climate warming.Drought prediction is of great significance for the optimal management of water resources and the alleviation of drought.Based on the standardized precipitation evapotranspiration index (SPEI) of different time scales for drought evaluation,this paper constructs the artificial neural network (ANN) and support vector regression (SVR) models to predict droughts in the prediction periods of 1 to 3 months,and builds the EMD-ANN and EMD-SVR coupling models to increase the prediction precision for the SPEI1 with the scale of 1 month.The results showed that:The ANN and SVR models have good prediction precision for SPEI with the scales of 3 months.In addition,the prediction precision of the SVR model is slightly better than that of ANN model.The shorter the prediction period is,the higher the prediction precision is.The coefficient of determination of the ANN and SVR models for the drought prediction period of 1 month accounts for 0.834~0.911.The ANN and SVR models are not suitable for the prediction of the SPEI1 with scale of 1 month.After processing by EMD and wavelet denoising,the prediction precision of the SPEI1 by the EMD-ANN and EMD-SVR models is significantly increased.
ISSN:1001-9235

Similar Items