Journal of Korean Society on Water Environment (한국물환경학회지)

Korean Society on Water Environment (한국물환경학회)
권호 표지
DOI QR코드

DOI QR Code

Short-Term Water Quality Prediction of the Paldang Reservoir Using Recurrent Neural Network Models

순환신경망 모델을 활용한 팔당호의 단기 수질 예측

  • Jiwoo Han (Han-River Environment Research Center, National Institute of Environment Research) ;
  • Yong-Chul Cho (Han-River Environment Research Center, National Institute of Environment Research) ;
  • Soyoung Lee (Han-River Environment Research Center, National Institute of Environment Research) ;
  • Sanghun Kim (Han-River Environment Research Center, National Institute of Environment Research) ;
  • Taegu Kang (Han-River Environment Research Center, National Institute of Environment Research)
  • 한지우 (국립환경과학원 한강물환경연구소) ;
  • 조용철 (국립환경과학원 한강물환경연구소) ;
  • 이소영 (국립환경과학원 한강물환경연구소) ;
  • 김상훈 (국립환경과학원 한강물환경연구소) ;
  • 강태구 (국립환경과학원 한강물환경연구소)
  • Received : 2022.12.20
  • Accepted : 2023.01.19
  • Published : 2023.01.30
Download PDF
⟨ Previous Next ⟩

Abstract

Climate change causes fluctuations in water quality in the aquatic environment, which can cause changes in water circulation patterns and severe adverse effects on aquatic ecosystems in the future. Therefore, research is needed to predict and respond to water quality changes caused by climate change in advance. In this study, we tried to predict the dissolved oxygen (DO), chlorophyll-a, and turbidity of the Paldang reservoir for about two weeks using long short-term memory (LSTM) and gated recurrent units (GRU), which are deep learning algorithms based on recurrent neural networks. The model was built based on real-time water quality data and meteorological data. The observation period was set from July to September in the summer of 2021 (Period 1) and from March to May in the spring of 2022 (Period 2). We tried to select an algorithm with optimal predictive power for each water quality parameter. In addition, to improve the predictive power of the model, an important variable extraction technique using random forest was used to select only the important variables as input variables. In both Periods 1 and 2, the predictive power after extracting important variables was further improved. Except for DO in Period 2, GRU was selected as the best model in all water quality parameters. This methodology can be useful for preventive water quality management by identifying the variability of water quality in advance and predicting water quality in a short period.

Keywords

Acknowledgement

본 논문은 환경부의 재원으로 국립환경과학원의 지원을 받아 수행하였습니다(NIER-2022-01-01-042).

References

  1. Ahn, C. Y., Oh, H. M., and Park, Y. S. (2011). Evaluation of environmental factors on cyanobacterial bloom in eutrophic reservoir using artificial neural networks, Journal of Phycology, 47(3), 495-504. https://doi.org/10.1111/j.1529-8817.2011.00990.x
  2. Ahu, R., Yang, L., Liu, T., Wen, X., Zhang, L., and Chang, Y. (2019). Hydrological responses to the future climate change in a data scarce region, Northwest China: Application of machine learning models, Water, 11(8), 1588.
  3. Bhandari, N. S. and Nayal, K. (2008). Correlation study on physico-chemical parameters and quality assessment of Kosi river water, Uttarakhand, E-Journal of Chemistry, 5(2), 342-346. https://doi.org/10.1155/2008/140986
  4. Brieman, L. (2001). Random forests, Machine Learning, 45, 5-32. https://doi.org/10.1023/A:1010933404324
  5. Cho, K., Van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., and Bengio, Y. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation, arXiv preprint arXiv, 1406.1078.
  6. Choi, H., Suh, S. I., Kim, S. H., Han, E. J., and Ki, S. J. (2021). Assessing the performance of deep learning algorithms for short-term surface water quality prediction, Sustainability, 13(19), 10690.
  7. Garcia-Nieto, P. J., Garcia-Gonzalo, E., Alonso Fernandez, J. R., and Diaz Muniz, C. (2020). A new predictive model for evaluating Chlorophyll-a concentration in Tanes reservoir by using a gaussian process regression, Water Resources Management, 34, 4921-4941. https://doi.org/10.1007/s11269-020-02699-x
  8. Herzsprung, P., Wentzky, V., Kamjunke, N., Tumpling, W., Wilske, C., Friese, K., Boehrer, B., Reemtsma, T., Rinke, K. J., and Lechtenfeld, O. (2020). Improved understanding of dissolved organic matter processing in freshwater using complementary experimental and machine learning approaches, Environment Science & Technology, 54(21), 13556-13565. https://doi.org/10.1021/acs.est.0c02383
  9. Hochreiter, S. and Schmidhuber, J. (1997). Long short-term memory, Neural Computation, 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  10. Hu, Z., Zhang, Y., Zhao, Y., Xie, M., Zhong, J., Tu, Z., and Liu, J. (2019). A water quality prediction method based on the deep LSTM network considering correlation in smart mariculture, Sensors, 19(6), 1420.
  11. Jo, B. G., Jung, W. S., Lee, J. M., and Kim, Y. D. (2022). Analysis of water quality impact of Hapcheon dam reservoir according to changes in watershed runoff using ANN, Journal of Wetlands Research, 24(1), 25-37. [Korea Literature] https://doi.org/10.17663/JWR.2022.24.1.25
  12. Kim, M., Lee, J., Sung, K., Lim, C., Hwang, W., and Hyun, S. (2022). Potential impacts of climate change on water temperature of the streams in Han-river basin, Journal of Korean Society on Water Environment, 38(1), 19-30. [Korean Literature] https://doi.org/10.15681/KSWE.2022.38.1.19
  13. Kong, D. (2019). Evaluating effect of density flow from upstream on vertical distribution of water quality at the Paldang reservoir, Journal of Korea Society on Water Environment, 35(6), 557-566. [Korean Literature]
  14. Kuo, J. T., Hsieh, M. H., Lung, W. S., and She, N. (2007). Using artificial neural network for reservoir eutrophication prediction, Ecological Modelling, 200(1-2), 171-177. https://doi.org/10.1016/j.ecolmodel.2006.06.018
  15. Kwak, J. (2021). A study on the 3-month prior prediction of Chl-a concentration in the Daechong lake using hydrometeorological forecasting data, Journal of Wetlands Research, 23(2), 144-153. [Korean Literature]
  16. Lee, J. H., Lee, J. Y., Lee, M. H., Lee, M. Y., Kim, Y. W., Hyung, J. S., Kim, K. B., Cha, Y. K., and Koo, J. Y. (2022). Development of a short-term water quality prediction model for urban rivers using real-time water quality data, Water Supply, 22(4), 4082-4097. https://doi.org/10.2166/ws.2022.038
  17. Li, L., Jamieson, K., DeSalvo, G., Rostamizadeh, A., and Talwalkar, A. (2018). Hyperband: A novel bandit-based approach to hyperparameter optimization, Journal of Machine Learning Research, 18(1), 6765-6816.
  18. Li, M., Xie, G. Q., Dai, C. R., Yu, L. X., Li, F. R., and Yang, S. P. (2009). A study of the relationship between the water body chlorophyll-a and water quality factors of the offcoast of Dianchi lake, Yunnan Geographic Environment Research, 21(2), 102-106.
  19. Liu, P., Wang, J., Sangaiah, A., Xie, Y., and Yin, X. (2019). Analysis and prediction of water quality using LSTM deep neural networks in IoT environment, Sustainability, 11(7), 2058.
  20. Lopez-Archilla, A., Moreira, D., Lopez-Garcia, P., and Guerrero, C. (2004). Phytoplankton diversity and cyanobacterial dominance in a hypereutrophic shallow lake with biologically produced alkaline pH, Extremophiles, 8(2), 109-115. https://doi.org/10.1007/s00792-003-0369-9
  21. Lu, H. and Ma, X. (2020). Hybrid decision tree-based machine learning models for short-term water quality prediction, Chemosphere, 249, 126169.
  22. Luo, W., Zhu, S., Wu, S., and Dai, J. (2019). Comparing artificial intelligence techniques for chlorophyll-a prediction in US lakes, Environmental Science and Pollution Research, 26(29), 30524-30532. https://doi.org/10.1007/s11356-019-06360-y
  23. Min, Y. H., Hyun, D. Y., Eum, C. H., Chung, N., Kang, S. W., and Lee, S. (2011), A study on relationship of concentration of phosphorus, turbidity and pH with temperature in water and soil, Analytical Science and Technology, 24(5), 378-386. https://doi.org/10.5806/AST.2011.24.5.378
  24. Moheimani, N. R. and Borowitzka, M. A. (2006). The long-term culture of the coccolithophore Pleurochrysis carterae (Haptophyta) in outdoor raceway ponds, Journal of Applied Phycology, 18(6), 703-712. https://doi.org/10.1007/s10811-006-9075-1
  25. National Institute of Environmental Research (NIER). (2017). A comprehensive study on water quality control of Paldang watershed(III): a assessment system to diagnose and evaluate water pollutant, National Institute of Environmental Research, NIER-RP2017-223. Incheon, Korea, 1-50. [Korea Literature]
  26. National Institute of Environmental Research (NIER). (2021). Prediction on water quality variations in Paldang Reservoir by climate change(I): Application of Data-based model through high frequency monitoring, National Institute of Environmental Research, NIER-RP2021-138. Incheon, Korea, 1-50. [Korea Literature]
  27. Park, J. (2021). The effect of input variables clustering on the characteristics of ensemble machine learning model for water quality prediction, Journal of Korean Society on Water Environment, 37(5), 335-343. [Korea Literature] https://doi.org/10.15681/KSWE.2021.37.5.335
  28. Park, J. and Lee, H. (2020). Prediction of high turbidity in rivers using LSTM algorithm, Journal of Korean Society of Water and Wastewater, 34(1), 35-43. [Korean Literature] https://doi.org/10.11001/jksww.2020.34.1.035
  29. Park, S. J. and Lee, D. K. (2020). Prediction of coastal flooding risk under climate change impacts in South Korea using machine learning algorithms, Environmental Research Letters, 15(9), 094052.
  30. Park, Y., Cho, K. H., Park, J., Cha, S. M., and Kim, J. H. (2015). Development of early-warning protocol for predicting chlorophyll-a concentration using machine learning models in freshwater and estuarine reservoirs, Science of the Total Environment, 502, 31-41. https://doi.org/10.1016/j.scitotenv.2014.09.005
  31. Rixen, T., Baum, A., Sepryani, H., Pohlmann, T., Jose, C., and Samiaji, J. (2010). Dissolved oxygen and its response to eutrophication in a tropical black water river, Journal of Environmental Management, 91(8), 1730-1737. https://doi.org/10.1016/j.jenvman.2010.03.009
  32. Ruan, X. H., Shi, X. D., Zhao, Z. H., Ni, L. X., Wu, Y., and Jiao, T. (2008). Correlation between chlorophyll-a concentration and environmental factors in shallow lakes in plain river network areas of Suzhou, Journal of Lake Sciences, 20(5), 556-562. https://doi.org/10.18307/2008.0502
  33. Saraceno, J. F., Pellerin, B. A., Downing, B. D., Boss, E., Bachand, P. A., and Bergamaschi, B. A, (2009). High-frequency in situ optical measurements during a storm event: assessing relationships between dissolved organic matter, sediment concentrations, and hydrologic processes, Journal of Geophysical research, 114(G4).
  34. Seo, H. J., Kang, I. S., Son, K. R., Eun, Y., Jeong, W. S., and Kim, S. J. (2019). Evaluation of water quality characteristics using multivariate statistical analysis in the fourth reservoir, Journal of Environmental Analysis, Health and Toxicology, 22(3), 117-125. https://doi.org/10.36278/jeaht.22.3.117
  35. Shim, S. H., Kim, Y. H., Lee, H. W., Kim, M., and Choi, J. H. (2022). Comparison of chlorophyll-a prediction and analysis of influential factors in Yeongsan river using machine learning and deep learning, Journal of Korean Society on Water Environment, 38(6), 292-305. [Korean Literature] https://doi.org/10.15681/KSWE.2022.38.6.292
  36. Singh, K. P., Basant, N., and Gupta, S. (2011). Support vector machines in water quality management, Analytica Chimica Acta, 703(2), 152-162. https://doi.org/10.1016/j.aca.2011.07.027
  37. Wang, X., Zhang, H., Bertone, E., Stewart, R., and Hughes, S. (2021). Coupled data-driven and process-based model for fluorescent dissolved organic matter prediction in a shallow subtropical reservoir, Environmental Modelling & Software, 141, 105053.
  38. Woo , J. W ., K im, Y. J., a nd Y o on, J . S. ( 2022). Predictio n of salinity of Nakdong river estuary using deep learning algorithm (LSTM) for time series analysis, Journal of Korean Society of Coastal and Ocean Engineers, 34(4), 128-134. [Korea Literature] https://doi.org/10.9765/KSCOE.2022.34.4.128
  39. Wu, J. and Wang, Z. (2022). A hybrid model for water quality prediction based on an artificial neural network, wavelet transform, and long short-term memory, Water, 14(4), 610.
  40. Xu, J., Xu, Z., Kuang, J., Lin, C., Xiao, L., Huang, X., and Zhang, Y. (2021). An alternative to laboratory testing: Random Forest-based water quality prediction framework for inland and nearshore water bodies, Water, 13(22), 3262.
  41. Yi, H. S., Kim, D. S., Hwang, M. H., and An, K. G. (2016). Assessment of runoff and water temperature variations under RCP climate change scenario in Yongdam dam watershed, South Korea, Journal of Korean Society on Water Environment, 32(2), 173-182. [Korean Literature] https://doi.org/10.15681/KSWE.2016.32.2.173
  42. Yu, S., Lee, E., Park, M., Kim, K., Im, J., Ryu, I., Choi, H., Byeon, M., and Noh, H. (2018). Changes in the water environment based on the statistical data in the lake Paldang, Journal of Korean Society on Water Environment, 34(6), 688-702. [Korean Literature] https://doi.org/10.15681/KSWE.2018.34.6.688
  43. Zang, C., Huang, S., Wu, M., Du, S., Scholz, M., Gao, F., Lin, C., Guo, Y., and Dong, Y. (2010). Comparison of relationships between pH, dissolved oxygen and chlorophyll-a for aquaculture and non-aquaculture waters, Water, Air, & Soil Pollution, 219(1), 157-174. https://doi.org/10.1007/s11270-010-0695-3
  44. Zhang, J. Y., Huang, J., Yan, F., and Zhang, Z. Q. (2009). Preliminary study on characters of dissolved oxygen and the relationship with pH in Meiliang lake, Journal of Fudan University, 48(5), 623-627.
  45. Zhu, S. and Heddam, S. (2020). Prediction of dissolved oxygen in urban rivers at the Three Gorges reservoir, China: Extreme learning machines (ELM) versus artificial neural network (ANN), Water Quality Research Journal, 55(1), 106-118. https://doi.org/10.2166/wqrj.2019.053