{"title":"基于水文因子深度学习集合建模的径流区间预测方法研究","authors":"Jinghan Huang, Zhaocai Wang, Jinghan Dong, Junhao Wu","doi":"10.1007/s00477-024-02780-6","DOIUrl":null,"url":null,"abstract":"<p>Precise prediction of runoff is not only conducive to the prevention of floods and droughts but also to the rational use of water resources. Due to the frequency of weather extremes and the complexity of runoff variability, achieving accurate runoff predictions is challenging. This research develops a deep-learning ensemble model for interval prediction based on meteorological and hydrological factors. The model can be divided into four stages: feature extraction, decomposition, point prediction, and interval prediction. First, Pearson's correlation coefficient filters out key driving variables affecting runoff. Next, the original data are decomposed by variational modal decomposition (VMD) to intrinsic modal function (IMF); Then, each IMF is decomposed by complementary ensemble empirical modal decomposition (CEEMD) to capture more data details. Following, the runoff point prediction portion is realized by the attention mechanism fusion gated recurrent unit (AM-GRU). In this study, data from Dunhuang and Panjiazhuang stations, located in the upper and lower reaches of the Shule River in China, were used to validate and analyze the VMD-CEEMD-ISSA-AM-GRU (VCIAG) model. The results show that (1) the VCIAG model has the best fitting effect which the NSE values of Dunhuang and Panjiazhuang stations are 0.97 and 0.96, respectively. (2) In the multi-period prediction in advance, the highest prediction accuracy is achieved when the prediction period is 1 day and the accuracy of the prediction decreases gradually as the prediction period becomes longer. (3) In flood early warning, the VCIAG performs well at both stations, which suggests that the proposed model can take precautionary measures in advance before the floods come. (4) In terms of interval prediction, the VCIAG model has the narrowest prediction interval width and the highest prediction accuracy, which enhances the application value of the model.</p>","PeriodicalId":21987,"journal":{"name":"Stochastic Environmental Research and Risk Assessment","volume":"38 1","pages":""},"PeriodicalIF":3.9000,"publicationDate":"2024-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Research on runoff interval prediction method based on deep learning ensemble modeling with hydrological factors\",\"authors\":\"Jinghan Huang, Zhaocai Wang, Jinghan Dong, Junhao Wu\",\"doi\":\"10.1007/s00477-024-02780-6\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Precise prediction of runoff is not only conducive to the prevention of floods and droughts but also to the rational use of water resources. Due to the frequency of weather extremes and the complexity of runoff variability, achieving accurate runoff predictions is challenging. This research develops a deep-learning ensemble model for interval prediction based on meteorological and hydrological factors. The model can be divided into four stages: feature extraction, decomposition, point prediction, and interval prediction. First, Pearson's correlation coefficient filters out key driving variables affecting runoff. Next, the original data are decomposed by variational modal decomposition (VMD) to intrinsic modal function (IMF); Then, each IMF is decomposed by complementary ensemble empirical modal decomposition (CEEMD) to capture more data details. Following, the runoff point prediction portion is realized by the attention mechanism fusion gated recurrent unit (AM-GRU). In this study, data from Dunhuang and Panjiazhuang stations, located in the upper and lower reaches of the Shule River in China, were used to validate and analyze the VMD-CEEMD-ISSA-AM-GRU (VCIAG) model. The results show that (1) the VCIAG model has the best fitting effect which the NSE values of Dunhuang and Panjiazhuang stations are 0.97 and 0.96, respectively. (2) In the multi-period prediction in advance, the highest prediction accuracy is achieved when the prediction period is 1 day and the accuracy of the prediction decreases gradually as the prediction period becomes longer. (3) In flood early warning, the VCIAG performs well at both stations, which suggests that the proposed model can take precautionary measures in advance before the floods come. (4) In terms of interval prediction, the VCIAG model has the narrowest prediction interval width and the highest prediction accuracy, which enhances the application value of the model.</p>\",\"PeriodicalId\":21987,\"journal\":{\"name\":\"Stochastic Environmental Research and Risk Assessment\",\"volume\":\"38 1\",\"pages\":\"\"},\"PeriodicalIF\":3.9000,\"publicationDate\":\"2024-08-29\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Stochastic Environmental Research and Risk Assessment\",\"FirstCategoryId\":\"93\",\"ListUrlMain\":\"https://doi.org/10.1007/s00477-024-02780-6\",\"RegionNum\":3,\"RegionCategory\":\"环境科学与生态学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, CIVIL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Stochastic Environmental Research and Risk Assessment","FirstCategoryId":"93","ListUrlMain":"https://doi.org/10.1007/s00477-024-02780-6","RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
Research on runoff interval prediction method based on deep learning ensemble modeling with hydrological factors
Precise prediction of runoff is not only conducive to the prevention of floods and droughts but also to the rational use of water resources. Due to the frequency of weather extremes and the complexity of runoff variability, achieving accurate runoff predictions is challenging. This research develops a deep-learning ensemble model for interval prediction based on meteorological and hydrological factors. The model can be divided into four stages: feature extraction, decomposition, point prediction, and interval prediction. First, Pearson's correlation coefficient filters out key driving variables affecting runoff. Next, the original data are decomposed by variational modal decomposition (VMD) to intrinsic modal function (IMF); Then, each IMF is decomposed by complementary ensemble empirical modal decomposition (CEEMD) to capture more data details. Following, the runoff point prediction portion is realized by the attention mechanism fusion gated recurrent unit (AM-GRU). In this study, data from Dunhuang and Panjiazhuang stations, located in the upper and lower reaches of the Shule River in China, were used to validate and analyze the VMD-CEEMD-ISSA-AM-GRU (VCIAG) model. The results show that (1) the VCIAG model has the best fitting effect which the NSE values of Dunhuang and Panjiazhuang stations are 0.97 and 0.96, respectively. (2) In the multi-period prediction in advance, the highest prediction accuracy is achieved when the prediction period is 1 day and the accuracy of the prediction decreases gradually as the prediction period becomes longer. (3) In flood early warning, the VCIAG performs well at both stations, which suggests that the proposed model can take precautionary measures in advance before the floods come. (4) In terms of interval prediction, the VCIAG model has the narrowest prediction interval width and the highest prediction accuracy, which enhances the application value of the model.
期刊介绍:
Stochastic Environmental Research and Risk Assessment (SERRA) will publish research papers, reviews and technical notes on stochastic and probabilistic approaches to environmental sciences and engineering, including interactions of earth and atmospheric environments with people and ecosystems. The basic idea is to bring together research papers on stochastic modelling in various fields of environmental sciences and to provide an interdisciplinary forum for the exchange of ideas, for communicating on issues that cut across disciplinary barriers, and for the dissemination of stochastic techniques used in different fields to the community of interested researchers. Original contributions will be considered dealing with modelling (theoretical and computational), measurements and instrumentation in one or more of the following topical areas:
- Spatiotemporal analysis and mapping of natural processes.
- Enviroinformatics.
- Environmental risk assessment, reliability analysis and decision making.
- Surface and subsurface hydrology and hydraulics.
- Multiphase porous media domains and contaminant transport modelling.
- Hazardous waste site characterization.
- Stochastic turbulence and random hydrodynamic fields.
- Chaotic and fractal systems.
- Random waves and seafloor morphology.
- Stochastic atmospheric and climate processes.
- Air pollution and quality assessment research.
- Modern geostatistics.
- Mechanisms of pollutant formation, emission, exposure and absorption.
- Physical, chemical and biological analysis of human exposure from single and multiple media and routes; control and protection.
- Bioinformatics.
- Probabilistic methods in ecology and population biology.
- Epidemiological investigations.
- Models using stochastic differential equations stochastic or partial differential equations.
- Hazardous waste site characterization.