Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors
Study region: The Honghu Lake (HHL) and Changhu Lake (CHL) in middle China. Study focus: Large-scale and high-precision estimation of water quality parameters (WQPs) is critical in explaining the spatiotemporal dynamics and clarifying their response to short-term hydrometeorological factors. Six mac...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-02-01
|
| Series: | Journal of Hydrology: Regional Studies |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2214581824004671 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832591817334325248 |
|---|---|
| author | Wentong Hu Donghao Miao Chi Zhang Zixian He Wenquan Gu Dongguo Shao |
| author_facet | Wentong Hu Donghao Miao Chi Zhang Zixian He Wenquan Gu Dongguo Shao |
| author_sort | Wentong Hu |
| collection | DOAJ |
| description | Study region: The Honghu Lake (HHL) and Changhu Lake (CHL) in middle China. Study focus: Large-scale and high-precision estimation of water quality parameters (WQPs) is critical in explaining the spatiotemporal dynamics and clarifying their response to short-term hydrometeorological factors. Six machine learning models optimized by intelligent optimization algorithms (IOA-ML) were developed to retrieve WQPs using paired in situ measurements and near-synchronous Sentinel-2 reflectance (Rrs). Furthermore, the response of pixel-based WQPs to short-term hydrometeorological factors were explored by generalized additive model (GAM). New hydrological insights for the region: The results showed that Rrs curves were significantly correlated with WQPs concentration, which provided a solid foundation for WQPs retrieval. The best IOA-ML model for total phosphorus (TP), total nitrogen (TN), and permanganate index (CODMn) was extreme gradient boosting optimized by genetic algorithm (GA-XGB), while that for dissolved oxygen (DO) and turbidity was categorical boosting regression optimized by GA (GA-CBR). Coefficient of determination (R2) of the best retrieval models for the test sets of TP, TN, turbidity, CODMn and DO were 0.545, 0.418, 0.794, 0.798, and 0.653, respectively. The best retrieval models were applied to two big inland lakes and revealed that TP, TN, CODMn, and turbidity in HHL increased rapidly from 2016 to 2022, especially during 2021–2022. 81.4 %-91.5 % of the WQPs variations in HHL and 63.4 %-92 % in CHL can be explained by hydrometeorological factors. |
| format | Article |
| id | doaj-art-0acd7c042453441c8f8a41f23d80df33 |
| institution | Kabale University |
| issn | 2214-5818 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Journal of Hydrology: Regional Studies |
| spelling | doaj-art-0acd7c042453441c8f8a41f23d80df332025-01-22T05:42:05ZengElsevierJournal of Hydrology: Regional Studies2214-58182025-02-0157102118Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factorsWentong Hu0Donghao Miao1Chi Zhang2Zixian He3Wenquan Gu4Dongguo Shao5State Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan 430072, ChinaState Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan 430072, ChinaState Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan 430072, ChinaState Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan 430072, ChinaCorresponding authors.; State Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan 430072, ChinaCorresponding authors.; State Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan 430072, ChinaStudy region: The Honghu Lake (HHL) and Changhu Lake (CHL) in middle China. Study focus: Large-scale and high-precision estimation of water quality parameters (WQPs) is critical in explaining the spatiotemporal dynamics and clarifying their response to short-term hydrometeorological factors. Six machine learning models optimized by intelligent optimization algorithms (IOA-ML) were developed to retrieve WQPs using paired in situ measurements and near-synchronous Sentinel-2 reflectance (Rrs). Furthermore, the response of pixel-based WQPs to short-term hydrometeorological factors were explored by generalized additive model (GAM). New hydrological insights for the region: The results showed that Rrs curves were significantly correlated with WQPs concentration, which provided a solid foundation for WQPs retrieval. The best IOA-ML model for total phosphorus (TP), total nitrogen (TN), and permanganate index (CODMn) was extreme gradient boosting optimized by genetic algorithm (GA-XGB), while that for dissolved oxygen (DO) and turbidity was categorical boosting regression optimized by GA (GA-CBR). Coefficient of determination (R2) of the best retrieval models for the test sets of TP, TN, turbidity, CODMn and DO were 0.545, 0.418, 0.794, 0.798, and 0.653, respectively. The best retrieval models were applied to two big inland lakes and revealed that TP, TN, CODMn, and turbidity in HHL increased rapidly from 2016 to 2022, especially during 2021–2022. 81.4 %-91.5 % of the WQPs variations in HHL and 63.4 %-92 % in CHL can be explained by hydrometeorological factors.http://www.sciencedirect.com/science/article/pii/S2214581824004671Water quality parameters retrievalIOA-ML modelsSpatiotemporal dynamicsGeneralized additive modelHydrometeorological factors |
| spellingShingle | Wentong Hu Donghao Miao Chi Zhang Zixian He Wenquan Gu Dongguo Shao Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors Journal of Hydrology: Regional Studies Water quality parameters retrieval IOA-ML models Spatiotemporal dynamics Generalized additive model Hydrometeorological factors |
| title | Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors |
| title_full | Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors |
| title_fullStr | Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors |
| title_full_unstemmed | Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors |
| title_short | Retrieval of water quality parameters based on IOA-ML models and their response to short-term hydrometeorological factors |
| title_sort | retrieval of water quality parameters based on ioa ml models and their response to short term hydrometeorological factors |
| topic | Water quality parameters retrieval IOA-ML models Spatiotemporal dynamics Generalized additive model Hydrometeorological factors |
| url | http://www.sciencedirect.com/science/article/pii/S2214581824004671 |
| work_keys_str_mv | AT wentonghu retrievalofwaterqualityparametersbasedonioamlmodelsandtheirresponsetoshorttermhydrometeorologicalfactors AT donghaomiao retrievalofwaterqualityparametersbasedonioamlmodelsandtheirresponsetoshorttermhydrometeorologicalfactors AT chizhang retrievalofwaterqualityparametersbasedonioamlmodelsandtheirresponsetoshorttermhydrometeorologicalfactors AT zixianhe retrievalofwaterqualityparametersbasedonioamlmodelsandtheirresponsetoshorttermhydrometeorologicalfactors AT wenquangu retrievalofwaterqualityparametersbasedonioamlmodelsandtheirresponsetoshorttermhydrometeorologicalfactors AT dongguoshao retrievalofwaterqualityparametersbasedonioamlmodelsandtheirresponsetoshorttermhydrometeorologicalfactors |