Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting
This study applied a GR4J model in the Xiangjiang and Qujiang River basins for rainfall-runoff simulation. Four recurrent neural networks (RNNs)—the Elman recurrent neural network (ERNN), echo state network (ESN), nonlinear autoregressive exogenous inputs neural network (NARX), and long sh...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2018-11-01
|
| Series: | Water |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4441/10/11/1655 |
| _version_ | 1811310676192264192 |
|---|---|
| author | Ye Tian Yue-Ping Xu Zongliang Yang Guoqing Wang Qian Zhu |
| author_facet | Ye Tian Yue-Ping Xu Zongliang Yang Guoqing Wang Qian Zhu |
| author_sort | Ye Tian |
| collection | DOAJ |
| description | This study applied a GR4J model in the Xiangjiang and Qujiang River basins for rainfall-runoff simulation. Four recurrent neural networks (RNNs)—the Elman recurrent neural network (ERNN), echo state network (ESN), nonlinear autoregressive exogenous inputs neural network (NARX), and long short-term memory (LSTM) network—were applied in predicting discharges. The performances of models were compared and assessed, and the best two RNNs were selected and integrated with the lumped hydrological model GR4J to forecast the discharges; meanwhile, uncertainties of the simulated discharges were estimated. The generalized likelihood uncertainty estimation method was applied to quantify the uncertainties. The results show that the LSTM and NARX better captured the time-series dynamics than the other RNNs. The hybrid models improved the prediction of high, median, and low flows, particularly in reducing the bias of underestimation of high flows in the Xiangjiang River basin. The hybrid models reduced the uncertainty intervals by more than 50% for median and low flows, and increased the cover ratios for observations. The integration of a hydrological model with a recurrent neural network considering long-term dependencies is recommended in discharge forecasting. |
| first_indexed | 2024-04-13T10:03:07Z |
| format | Article |
| id | doaj.art-71490d7f46ad45a1bb356b486aba97ea |
| institution | Directory Open Access Journal |
| issn | 2073-4441 |
| language | English |
| last_indexed | 2024-04-13T10:03:07Z |
| publishDate | 2018-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Water |
| spelling | doaj.art-71490d7f46ad45a1bb356b486aba97ea2022-12-22T02:51:11ZengMDPI AGWater2073-44412018-11-011011165510.3390/w10111655w10111655Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow ForecastingYe Tian0Yue-Ping Xu1Zongliang Yang2Guoqing Wang3Qian Zhu4School of Hydrology and Water Resources, Nanjing University of Information Science & Technology, Nanjing 210044, ChinaDepartment of Civil Engineering, Institute of Hydrology and Water Resources, Zhejiang University, Hangzhou 310058, ChinaDepartment of Geological Sciences, University of Texas at Austin, Austin, TX 78712, USAState Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Nanjing Hydraulic Research Institute, Nanjing 210029, ChinaSchool of Civil Engineering, Southeast University, Nanjing 211189, ChinaThis study applied a GR4J model in the Xiangjiang and Qujiang River basins for rainfall-runoff simulation. Four recurrent neural networks (RNNs)—the Elman recurrent neural network (ERNN), echo state network (ESN), nonlinear autoregressive exogenous inputs neural network (NARX), and long short-term memory (LSTM) network—were applied in predicting discharges. The performances of models were compared and assessed, and the best two RNNs were selected and integrated with the lumped hydrological model GR4J to forecast the discharges; meanwhile, uncertainties of the simulated discharges were estimated. The generalized likelihood uncertainty estimation method was applied to quantify the uncertainties. The results show that the LSTM and NARX better captured the time-series dynamics than the other RNNs. The hybrid models improved the prediction of high, median, and low flows, particularly in reducing the bias of underestimation of high flows in the Xiangjiang River basin. The hybrid models reduced the uncertainty intervals by more than 50% for median and low flows, and increased the cover ratios for observations. The integration of a hydrological model with a recurrent neural network considering long-term dependencies is recommended in discharge forecasting.https://www.mdpi.com/2073-4441/10/11/1655recurrent neural networkslong-short term memoryhydrological modelinguncertaintystream flow forecasting |
| spellingShingle | Ye Tian Yue-Ping Xu Zongliang Yang Guoqing Wang Qian Zhu Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting Water recurrent neural networks long-short term memory hydrological modeling uncertainty stream flow forecasting |
| title | Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting |
| title_full | Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting |
| title_fullStr | Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting |
| title_full_unstemmed | Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting |
| title_short | Integration of a Parsimonious Hydrological Model with Recurrent Neural Networks for Improved Streamflow Forecasting |
| title_sort | integration of a parsimonious hydrological model with recurrent neural networks for improved streamflow forecasting |
| topic | recurrent neural networks long-short term memory hydrological modeling uncertainty stream flow forecasting |
| url | https://www.mdpi.com/2073-4441/10/11/1655 |
| work_keys_str_mv | AT yetian integrationofaparsimonioushydrologicalmodelwithrecurrentneuralnetworksforimprovedstreamflowforecasting AT yuepingxu integrationofaparsimonioushydrologicalmodelwithrecurrentneuralnetworksforimprovedstreamflowforecasting AT zongliangyang integrationofaparsimonioushydrologicalmodelwithrecurrentneuralnetworksforimprovedstreamflowforecasting AT guoqingwang integrationofaparsimonioushydrologicalmodelwithrecurrentneuralnetworksforimprovedstreamflowforecasting AT qianzhu integrationofaparsimonioushydrologicalmodelwithrecurrentneuralnetworksforimprovedstreamflowforecasting |