Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting
Abstract Daily inflow forecasting is of vital importance in reservoir economic operation. In the context of hydrometeorological forecasting, the effectiveness of the data‐driven models has been demonstrated as bias correctors for physically‐based models or direct forecasting models. However, existin...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2023-03-01
|
| Series: | Journal of Flood Risk Management |
| Subjects: | |
| Online Access: | https://doi.org/10.1111/jfr3.12854 |
| _version_ | 1811164515734126592 |
|---|---|
| author | Xinran Luo Pan Liu Qianjin Dong Yanjun Zhang Kang Xie Dongyang Han |
| author_facet | Xinran Luo Pan Liu Qianjin Dong Yanjun Zhang Kang Xie Dongyang Han |
| author_sort | Xinran Luo |
| collection | DOAJ |
| description | Abstract Daily inflow forecasting is of vital importance in reservoir economic operation. In the context of hydrometeorological forecasting, the effectiveness of the data‐driven models has been demonstrated as bias correctors for physically‐based models or direct forecasting models. However, existing studies only highlight the performance improvements provided by the data‐driven model, lacking a comprehensive investigation on whether the data‐driven model should be used as bias correctors or direct forecasting models. This study constructs long short‐term memory (LSTM)‐based preprocessing and postprocessing techniques for a hydrological model, which are tested by linear scaling preprocessing and autoregressive (AR) postprocessing models. The integrated model is compared with the LSTM‐only model. The Shuibuya and Zuojiang reservoirs in China are selected as case studies. Results indicate that: (1) LSTM‐based bias correctors are effective in both preprocessing and postprocessing and (2) the integrated model is comparable to the LSTM‐only model when trained with four or more years of data, while it is better than the LSTM‐only model when trained with less data. These findings demonstrate that data‐driven methods can effectively correct the bias in physically‐based model output, and integrating the physical and data‐driven models is useful in improving multi‐step ahead reservoir inflow forecasting if limited data can be obtained. |
| first_indexed | 2024-04-10T15:22:33Z |
| format | Article |
| id | doaj.art-581af5cbb7364964877fdb37d77e3062 |
| institution | Directory Open Access Journal |
| issn | 1753-318X |
| language | English |
| last_indexed | 2024-04-10T15:22:33Z |
| publishDate | 2023-03-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of Flood Risk Management |
| spelling | doaj.art-581af5cbb7364964877fdb37d77e30622023-02-14T12:22:52ZengWileyJournal of Flood Risk Management1753-318X2023-03-01161n/an/a10.1111/jfr3.12854Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecastingXinran Luo0Pan Liu1Qianjin Dong2Yanjun Zhang3Kang Xie4Dongyang Han5State Key Laboratory of Water Resources and Hydropower Engineering Science Wuhan University Wuhan ChinaState Key Laboratory of Water Resources and Hydropower Engineering Science Wuhan University Wuhan ChinaState Key Laboratory of Water Resources and Hydropower Engineering Science Wuhan University Wuhan ChinaState Key Laboratory of Water Resources and Hydropower Engineering Science Wuhan University Wuhan ChinaState Key Laboratory of Water Resources and Hydropower Engineering Science Wuhan University Wuhan ChinaState Key Laboratory of Water Resources and Hydropower Engineering Science Wuhan University Wuhan ChinaAbstract Daily inflow forecasting is of vital importance in reservoir economic operation. In the context of hydrometeorological forecasting, the effectiveness of the data‐driven models has been demonstrated as bias correctors for physically‐based models or direct forecasting models. However, existing studies only highlight the performance improvements provided by the data‐driven model, lacking a comprehensive investigation on whether the data‐driven model should be used as bias correctors or direct forecasting models. This study constructs long short‐term memory (LSTM)‐based preprocessing and postprocessing techniques for a hydrological model, which are tested by linear scaling preprocessing and autoregressive (AR) postprocessing models. The integrated model is compared with the LSTM‐only model. The Shuibuya and Zuojiang reservoirs in China are selected as case studies. Results indicate that: (1) LSTM‐based bias correctors are effective in both preprocessing and postprocessing and (2) the integrated model is comparable to the LSTM‐only model when trained with four or more years of data, while it is better than the LSTM‐only model when trained with less data. These findings demonstrate that data‐driven methods can effectively correct the bias in physically‐based model output, and integrating the physical and data‐driven models is useful in improving multi‐step ahead reservoir inflow forecasting if limited data can be obtained.https://doi.org/10.1111/jfr3.12854hydrological modelinglong short‐term memorypostprocessingpreprocessingreservoir inflow forecasting |
| spellingShingle | Xinran Luo Pan Liu Qianjin Dong Yanjun Zhang Kang Xie Dongyang Han Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting Journal of Flood Risk Management hydrological modeling long short‐term memory postprocessing preprocessing reservoir inflow forecasting |
| title | Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting |
| title_full | Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting |
| title_fullStr | Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting |
| title_full_unstemmed | Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting |
| title_short | Exploring the role of the long short‐term memory model in improving multi‐step ahead reservoir inflow forecasting |
| title_sort | exploring the role of the long short term memory model in improving multi step ahead reservoir inflow forecasting |
| topic | hydrological modeling long short‐term memory postprocessing preprocessing reservoir inflow forecasting |
| url | https://doi.org/10.1111/jfr3.12854 |
| work_keys_str_mv | AT xinranluo exploringtheroleofthelongshorttermmemorymodelinimprovingmultistepaheadreservoirinflowforecasting AT panliu exploringtheroleofthelongshorttermmemorymodelinimprovingmultistepaheadreservoirinflowforecasting AT qianjindong exploringtheroleofthelongshorttermmemorymodelinimprovingmultistepaheadreservoirinflowforecasting AT yanjunzhang exploringtheroleofthelongshorttermmemorymodelinimprovingmultistepaheadreservoirinflowforecasting AT kangxie exploringtheroleofthelongshorttermmemorymodelinimprovingmultistepaheadreservoirinflowforecasting AT dongyanghan exploringtheroleofthelongshorttermmemorymodelinimprovingmultistepaheadreservoirinflowforecasting |