Journal of the Korean Society of Marine Environment & Safety (해양환경안전학회지)

The Korean Society of Marine Environment and safety (해양환경안전학회)
권호 표지
DOI QR코드

DOI QR Code

Preliminary Study on the Reproduction of Dissolved Oxygen Concentration in Jinhae Bay Based on Deep Learning Model

딥러닝 모형 기반 진해만 용존산소농도 재현을 위한 기초연구

  • Park, Seongsik (Department of Ocean Engineering Pukyong National University) ;
  • Kim, Kyunghoi (Department of Ocean Engineering Pukyong National University)
  • Received : 2022.02.15
  • Accepted : 2022.04.27
  • Published : 2022.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

We conducted a case study to determine the optimal model parameters and predictors of Long Short-Term Memory (LSTM) for the reproduction of dissolved oxygen (DO) concentration in Jinhae Bay. The model parameter case study indicated the lowest accuracy when the Hidden node=10, Epoch=100. This was caused by underfitting of machine learning. The accuracy increased as the Hidden node and Epoch increased. The accuracy was the highest when the Hidden node=80 and Epoch=100 with R2=0.99. In the bottom DO reproduction of Step 1 of the predictors case study, accuracy was highest when the water temperature was used as a predictor with R2=0.81. In Step 2, The R2 value increased up to 0.92 when the water temperature and SiO2 were used as a predictor. This was caused by a high correlation between the bottom DO and SiO2 concentrations. Consequently, we determined the optimal model parameters and predictors of LSTM for the reproduction of DO concentration in Jinhae Bay.

본 연구에서는 진해만의 DO 농도 재현을 목표로 LSTM 모형의 최적 매개변수 조건과 예측변수를 선별하기 위한 Case study를 진행하였다. 모형 매개변수 Case study 결과, 가장 적은 Hidden node와 Epoch인 Hidden node=10, Epoch=100에서 가장 낮은 정확도를 보였다. 이는 모형이 과소적합(Underfitting) 상태인 것으로 판단된다. Hidden node=80, Epoch=1200에서 R2 값은 0.99로 가장 높은 정확도를 보였다. 예측변수 Case study 결과, 1개의 환경변수만을 예측변수로 사용한 Step 1에서 수온을 예측변수로 했을 때 저층 DO 농도 재현의 R2 값은 0.81로 가장 높은 정확도를 보였다. 이후 2개의 환경변수를 사용한 Step 2에서는 수온과 SiO2를 예측변수로 했을 때 R2 값은 0.92로 수온만 사용했을 때보다 정확도가 급격히 증가하였다. 이는 저층 DO 농도와 SiO2 농도간의 높은 상관성$({\mid}R{\mid}=0.70)$에 기인한 것으로 판단된다. 상기 결과로부터 진해만의 DO 농도 재현에 적합한 LSTM 모형의 매개변수와 예측변수를 찾을 수 있었다.

Keywords

Acknowledgement

이 논문은 2021년도 정부(교육부)의 재원으로 한국연구재단의 지원을 받아 수행된 기초연구사업임(Grant 2021R1I1A306037411).

References

  1. Ahn, S., I. Yeon, Y. Han, and J. Lee(2001), Water quality forecasting at Gongju station in Geum River using neural network model, Journal of Korea Water Resources Association, Vol. 34, pp. 701-711 (in Korean).
  2. Breitburg, D., L. A. Levin, A. Oschlies, M. Gregoire, F. P. Chavez, D. J. Conley, V. Garcon, D. Gilbert, D. Gutierrez, K. Isensee, G. S. Jacinto, K. E. Limburg, I. Montes, S. W. A. Naqvi, G. C. Pitcher, N. N. Rabalais, M. R. Roman, K. A. Rose, B. A. Seibel, M. Telszewski, M. Yasuhara, and J. Zhang(2018), Declining oxygen in the global ocean and coastal waters, Science, Vol. 359, No. 6371.
  3. Diaz, R. J. and R. Rosenberg(2008), Spreading dead zones and consequences for marine ecosystems, Science, Vol. 321, pp. 926-929. https://doi.org/10.1126/science.1156401
  4. Eze, E. and T. Ajmal(2020), Dissolved Oxygen Forecasting in Aquaculture: A Hybrid Model Approach, Appl. Sci., Vol. 10, No. 20, 7079. https://doi.org/10.3390/app10207079
  5. Huck, P. M. and G. J. Farquhar(1974), Water quality models using the Box-Jenkins method. Journal of the Environmental Engineering, Vol. 100, pp. 733-751.
  6. Keeling, R. F., A. Kortzinger, and A. Gruber(2010), Ocean deoxygenation in a warming world, Marine Science, Vol. 2, No. 199, pp. 463-93.
  7. Lee, I. C., H. H. Kong, and S. J. Yoon(2008), Numerical Prediction for Reduction of Oxygen Deficient Water Mass by Ecological Model in Jinhae Bay, Journal of Ocean Engineering and Technology, Vol. 22, No. 5, pp. 75-82 (in Korean).
  8. Levin, L. and D. Breitburg(2015), Linking coasts and seas to address ocean deoxygenation, Nature Clim Change, Vol. 5, pp. 401-403. https://doi.org/10.1038/nclimate2595
  9. Li, W., H. Wu, N. Zhu, Y. Jiang, J. Tan, and Y. Guo(2021), Prediction of dissolved oxygen in a fishery pond based on gated recurrent unit (GRU), Information Processing in Agriculture, Vol. 8, No. 1, pp. 185-193. https://doi.org/10.1016/j.inpa.2020.02.002
  10. Matear, R. J. and A. C. Hirst(2003), Long-term changes in dissolved oxygen concentrations in the ocean caused by protracted global warming, Global Biogeochem. Cycles, Vol. 17, No. 4, 1125.
  11. NIFS(2021), Monitoring the Fishing Environment, http://www.nifs.go.kr/femo/.
  12. Park, J. and B. Ra(2014), Positive Research About Water Aeration Improvement to Break Thermal Stratification of Dam, Int. J. of Fluid Mach. Syst., Vol. 17, No. 5, pp. 37-42.
  13. Park, S. and K. Kim(2021), Prediction of DO Concentration in Nakdong River Estuary through Case Study Based on Long Short Term Memory Model, Journal of Korean Society of Coastal and Ocean Engineers, Vol. 33, No. 6, pp. 1-8 (in Korean). https://doi.org/10.9765/KSCOE.2021.33.1.1
  14. Rumelhart, D., G. Hinton, and R. Williams(1986), Learning representations by back-propagating errors, Nature, Vol. 323, pp. 533-536. https://doi.org/10.1038/323533a0
  15. Tealab, A.(2018), Time series forecasting using artificial neural networks methodologies: A systematic review, Future Computing and Informatics Journal, Vol. 3, No. 2, pp. 334-340. https://doi.org/10.1016/j.fcij.2018.10.003
  16. Xue, P., C. Chen, and R. C. Beardsley(2012), Observing system simulation experiments of dissolved oxygen monitoring in Massachusetts Bay, J. Geophys. Res., Vol. 117.
  17. Zhang, J., D. Gilbert, A. J. Gooday, L. Levin, S. W. A. Naqvi, J. J. Middelburg, M. Scranton, W. Ekau, A. Pena, B. Dewitte, T. Oguz, P. M. S. Monteiro, E. Urban, N. N. Rabalais, V. Ittekkot, W. M. Kemp, O. Ulloa, R. Elmgren, E. Escobar-Briones, and A. K. Van der Plas(2010), Natural and human-induced hypoxia and consequences for coastal areas: synthesis and future development, Biogeosciences Vol. 7, pp. 1443-1467. https://doi.org/10.5194/bg-7-1443-2010