Variational Bayesian Neural Network for Ensemble Flood Forecasting
Disastrous floods are destructive and likely to cause widespread economic losses. An understanding of flood forecasting and its potential forecast uncertainty is essential for water resource managers. Reliable forecasting may provide future streamflow information to assist in an assessment of the be...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-09-01
|
| Series: | Water |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4441/12/10/2740 |
| _version_ | 1797552160049725440 |
|---|---|
| author | Xiaoyan Zhan Hui Qin Yongqi Liu Liqiang Yao Wei Xie Guanjun Liu Jianzhong Zhou |
| author_facet | Xiaoyan Zhan Hui Qin Yongqi Liu Liqiang Yao Wei Xie Guanjun Liu Jianzhong Zhou |
| author_sort | Xiaoyan Zhan |
| collection | DOAJ |
| description | Disastrous floods are destructive and likely to cause widespread economic losses. An understanding of flood forecasting and its potential forecast uncertainty is essential for water resource managers. Reliable forecasting may provide future streamflow information to assist in an assessment of the benefits of reservoirs and the risk of flood disasters. However, deterministic forecasting models are not able to provide forecast uncertainty information. To quantify the forecast uncertainty, a variational Bayesian neural network (VBNN) model for ensemble flood forecasting is proposed in this study. In VBNN, the posterior distribution is approximated by the variational distribution, which can avoid the heavy computational costs in the traditional Bayesian neural network. To transform the model parameters’ uncertainty into the model output uncertainty, a Monte Carlo sample is applied to give ensemble forecast results. The proposed method is verified by a flood forecasting case study on the upper Yangtze River. A point forecasting model neural network and two probabilistic forecasting models, including hidden Markov Model and Gaussian process regression, are also applied to compare with the proposed model. The experimental results show that the VBNN performs better than other comparable models in terms of both accuracy and reliability. Finally, the result of uncertainty estimation shows that the VBNN can effectively handle heteroscedastic flood streamflow data. |
| first_indexed | 2024-03-10T15:56:02Z |
| format | Article |
| id | doaj.art-8e4793cfa9d6486a9942f19af2c7ed02 |
| institution | Directory Open Access Journal |
| issn | 2073-4441 |
| language | English |
| last_indexed | 2024-03-10T15:56:02Z |
| publishDate | 2020-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Water |
| spelling | doaj.art-8e4793cfa9d6486a9942f19af2c7ed022023-11-20T15:41:09ZengMDPI AGWater2073-44412020-09-011210274010.3390/w12102740Variational Bayesian Neural Network for Ensemble Flood ForecastingXiaoyan Zhan0Hui Qin1Yongqi Liu2Liqiang Yao3Wei Xie4Guanjun Liu5Jianzhong Zhou6China Southern Power Grid Power Generation Company, Guangzhou 510663, ChinaSchool of Hydropower and Information Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Hydropower and Information Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaChangjiang River Scientific Research Institute of Changjiang Water Resources Commission, Wuhan 430074, ChinaSchool of Hydropower and Information Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Hydropower and Information Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Hydropower and Information Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaDisastrous floods are destructive and likely to cause widespread economic losses. An understanding of flood forecasting and its potential forecast uncertainty is essential for water resource managers. Reliable forecasting may provide future streamflow information to assist in an assessment of the benefits of reservoirs and the risk of flood disasters. However, deterministic forecasting models are not able to provide forecast uncertainty information. To quantify the forecast uncertainty, a variational Bayesian neural network (VBNN) model for ensemble flood forecasting is proposed in this study. In VBNN, the posterior distribution is approximated by the variational distribution, which can avoid the heavy computational costs in the traditional Bayesian neural network. To transform the model parameters’ uncertainty into the model output uncertainty, a Monte Carlo sample is applied to give ensemble forecast results. The proposed method is verified by a flood forecasting case study on the upper Yangtze River. A point forecasting model neural network and two probabilistic forecasting models, including hidden Markov Model and Gaussian process regression, are also applied to compare with the proposed model. The experimental results show that the VBNN performs better than other comparable models in terms of both accuracy and reliability. Finally, the result of uncertainty estimation shows that the VBNN can effectively handle heteroscedastic flood streamflow data.https://www.mdpi.com/2073-4441/12/10/2740Bayesian neural networksflood forecastvariational inferenceforecast uncertainty |
| spellingShingle | Xiaoyan Zhan Hui Qin Yongqi Liu Liqiang Yao Wei Xie Guanjun Liu Jianzhong Zhou Variational Bayesian Neural Network for Ensemble Flood Forecasting Water Bayesian neural networks flood forecast variational inference forecast uncertainty |
| title | Variational Bayesian Neural Network for Ensemble Flood Forecasting |
| title_full | Variational Bayesian Neural Network for Ensemble Flood Forecasting |
| title_fullStr | Variational Bayesian Neural Network for Ensemble Flood Forecasting |
| title_full_unstemmed | Variational Bayesian Neural Network for Ensemble Flood Forecasting |
| title_short | Variational Bayesian Neural Network for Ensemble Flood Forecasting |
| title_sort | variational bayesian neural network for ensemble flood forecasting |
| topic | Bayesian neural networks flood forecast variational inference forecast uncertainty |
| url | https://www.mdpi.com/2073-4441/12/10/2740 |
| work_keys_str_mv | AT xiaoyanzhan variationalbayesianneuralnetworkforensemblefloodforecasting AT huiqin variationalbayesianneuralnetworkforensemblefloodforecasting AT yongqiliu variationalbayesianneuralnetworkforensemblefloodforecasting AT liqiangyao variationalbayesianneuralnetworkforensemblefloodforecasting AT weixie variationalbayesianneuralnetworkforensemblefloodforecasting AT guanjunliu variationalbayesianneuralnetworkforensemblefloodforecasting AT jianzhongzhou variationalbayesianneuralnetworkforensemblefloodforecasting |