Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

Comparison of physics-based and data-driven models for streamflow simulation of the Mekong river

메콩강 유출모의를 위한 물리적 및 데이터 기반 모형의 비교·분석

  • Lee, Giha (Department of Disaster Prevention and Environmental Engineering, Kyungpook National University) ;
  • Jung, Sungho (Department of Disaster Prevention and Environmental Engineering, Kyungpook National University) ;
  • Lee, Daeeop (Department of Disaster Prevention and Environmental Engineering, Kyungpook National University)
  • 이기하 (경북대학교 과학기술대학 건설방재공학과) ;
  • 정성호 (경북대학교 과학기술대학 건설방재공학과) ;
  • 이대업 (경북대학교 과학기술대학 건설방재공학과)
  • Received : 2018.02.12
  • Accepted : 2018.03.15
  • Published : 2018.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

In recent, the hydrological regime of the Mekong river is changing drastically due to climate change and haphazard watershed development including dam construction. Information of hydrologic feature like streamflow of the Mekong river are required for water disaster prevention and sustainable water resources development in the river sharing countries. In this study, runoff simulations at the Kratie station of the lower Mekong river are performed using SWAT (Soil and Water Assessment Tool), a physics-based hydrologic model, and LSTM (Long Short-Term Memory), a data-driven deep learning algorithm. The SWAT model was set up based on globally-available database (topography: HydroSHED, landuse: GLCF-MODIS, soil: FAO-Soil map, rainfall: APHRODITE, etc) and then simulated daily discharge from 2003 to 2007. The LSTM was built using deep learning open-source library TensorFlow and the deep-layer neural networks of the LSTM were trained based merely on daily water level data of 10 upper stations of the Kratie during two periods: 2000~2002 and 2008~2014. Then, LSTM simulated daily discharge for 2003~2007 as in SWAT model. The simulation results show that Nash-Sutcliffe Efficiency (NSE) of each model were calculated at 0.9(SWAT) and 0.99(LSTM), respectively. In order to simply simulate hydrological time series of ungauged large watersheds, data-driven model like the LSTM method is more applicable than the physics-based hydrological model having complexity due to various database pressure because it is able to memorize the preceding time series sequences and reflect them to prediction.

최근 기후변화 및 유역개발로 인하여 메콩강 유역의 수문환경이 급격히 변화하고 있으며, 메콩강을 공유하는 국가의 수재해 예방 및 지속가능한 수자원개발을 위해서는 메콩강 주요지점에서의 유량 정보의 분석 및 예측이 요구된다. 본 연구에서는 물리적 기반의 수문모형인 SWAT과 데이터기반 딥러닝 알고리즘인 LSTM을 이용하여 메콩강 하류 Kratie 지점의 유출모의를 수행하고, 유출모의 정확도 및 두 가지 방법론의 장 단점을 비교 분석한다. SWAT 모형의 구축을 위해 범용 입력자료(지형: HydroSHED, 토지이용: GLCF-MODIS, 토양: FAO-Soil map, 강우: APHRODITE 등)을 이용하였으며 warming-up 및 매개변수 보정 후 2003~2007년 일유량 모의를 수행하였다. LSTM을 이용한 유출모의의 경우, 딥러닝 오픈소스 라이브러리인 TensorFlow를 활용하여 Kratie 지점기준 메콩강 상류 10개 수위관측소의 두 기간(2000~2002, 2008~2014) 일수위 정보만을 이용하여 심층신경망을 학습하고, SWAT 모형과 마찬가지로 2003~2007년을 대상으로 Kratie 지점에 대한 일수위 모의 후 수위-유량관계곡선식을 이용하여 유출량으로 환산하였다. 두 모형의 모의성능 비교 검토를 위하여 모의기간에 대해 NSE (Nash-Sutcliffe Efficiency)을 산정한 결과, SWAT은 0.9, LSTM은 보다 높은 0.99의 정확도를 나타내는 것으로 분석되었다. 메콩강과 같은 대유역의 특정 지점에 대한 수문시계열 자료의 모의를 위해서는 다양한 입력자료를 요구하는 물리적 수문모형 대신 선행 시계열자료의 변동성을 기억 학습하여 이를 예측에 반영하는 LSTM 기법 등 데이터기반의 심층신경망 모형의 적용이 가능할 것으로 판단된다.

Keywords

References

  1. Baran, E., and Myschowoda, C. (2009). "Dams and fisheries in the Mekong basin." Aquatic Ecosystem Health & Management, Vol. 12, No. 3, pp. 227-234. https://doi.org/10.1080/14634980903149902
  2. Environmental Protection Agency (2017). An overview of rainfall- runoff model types. EPA Report, EPA/600/R-14/152, p. 30.
  3. Hitogoto, M., Sakuraba, M., and Sei, Y. (2016) "Development of the real-time river stage prediction method using deep learning." Annual Journal of Hydraulic Engineering (B1), JSCE, Vol. 72, No. 4, pp. 187-192.
  4. Hochreiter, S., and Schmidhuber, J. (1997). "Long short-term memory." Neural Computation, Vol. 9, No. 8, pp. 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  5. Johnston, R. M., and Kummu, M. (2012). "Water resource models in the Mekong basin: a review." Water Resources Management, Vol. 26, pp. 429-455. https://doi.org/10.1007/s11269-011-9925-8
  6. Jung, S. H., Lee, D. E., and Lee, K. S. (2018), "Prediction of river water level using deep-learning open library." Journal of the Korean Society of Hazard Mitigation, Vol. 18, No. 1, pp. 1-11.
  7. Kim, S., and Tachikawa, Y. (2018), "Real-time river-stage prediction with artificial neural network based on only upstream observation data." Annual Journal of Hydraulic Engineering, JSCE, Vol. 62, pp. 1375-1380.
  8. Kokkonen, T. S., and Jakeman, A. J. (2001), "A comparison of metric and conceptual approaches in rainfall-runoff modeling and its implications." Water Resources Research, Vol. 37, No. 9, pp. 2345-2352. https://doi.org/10.1029/2001WR000299
  9. Kummu, M., Tes, S., Yin, S., Adamson, P., Jozsa, J., Koponen, J., Richey, J., and Sarkkula, J. (2014), "Water balance analysis for the Tonle Sap lake-floodplain system." Hydrological Processes, Vol. 28, pp. 1722-1733. https://doi.org/10.1002/hyp.9718
  10. Lauri, H., Rasanen, T. A., and Kummu, M. (2014). "Using reanalysis and remotely sensed temperature and precipitation data for hydrological modeling in monsoon climate: Mekong river case study." Journal of Hydrometeorology, Vol. 15, No. 4, pp. 1532-1545. https://doi.org/10.1175/JHM-D-13-084.1
  11. Lee, D. E., Yu, W. S., and Lee G. H. (2018), "Large scale rainfall-runoff analysis using SWAT model: Case study: Mekong river basin", Journal of the Korean Society of Agricultural Engineers, Vol. 60, No. 1, pp. 47-57. https://doi.org/10.5389/KSAE.2018.60.1.047
  12. Lee, G. H. (2008). Assessment of prediction uncertainty due to various sources involved in rainfall-runoff modeling. Doctoral Thesis, Kyoto University, Japan.
  13. Minns, A. W., and Hall, M. J. (1996), "Artificial neural networks as rainfall-runoff models." Hydrological Sciences Journal, Vol. 41, No. 3, pp. 399-417. https://doi.org/10.1080/02626669609491511
  14. MRC (2009), The flow of the Mekong. MRC management information booklet series, No. 2, p. 16.
  15. Olah, C. (2015). "Understanding lstm networks." GITHUB blog, http://colah.github.io/posts/2015-08-Understanding-LSTMs/.
  16. Raghavan, S. V., Vu, M. T., and Liong, S. Y. (2012). "Assessment of future stream flow over the Sesan catchment of the lower Mekong basin in Vietnam." Hydrological Processes, Vol. 26, No. 24, pp. 3661-3668. https://doi.org/10.1002/hyp.8452
  17. Reungsang, P., Kanwar, R. S., and Srisuk, K. (2010). "Application of SWAT model in simulating stream flow for the Chi river Subbasin II in Northeast Thailand." Trends Research in Science and Technology, Vol. 2, No. 1, pp. 23-28.
  18. Shreenivas, L., and Shrikant, C. (2010) "Comparison of data driven modelling techniques for river flow forecasting." Hydrological Sciences Journal, Vol. 55, No. 7, pp. 1163-1174. https://doi.org/10.1080/02626667.2010.512867
  19. Shrestha, B., Babel, M. S., Maskey, S., Griensven, A. V., Uhlenbrook, S., Green, A., and Akkharath, I. (2013). "Impact of climate change on sediment yield in the Mekong river basin: a case study of the Nam Ou basin, Lao PDR." Hydrology and Earth System Sciences, Vol. 17, No. 1, pp. 1-20. https://doi.org/10.5194/hess-17-1-2013
  20. Sok, K., and Oeurng, C. (2016). "Application of HEC-HMS model to assess streamflow and water resources availability in Stung Sangker Catchment of Mekong' Tonle Sap lake basin in Cambodia." Preprints, 2016, 28 December.
  21. Sun, W. C., Ishidaira, H., and Bastola, S. (2010). "Towards improving river discharge estimation in ungauged basins: calibration of rainfall-runoff models based on satellite observations of river flow width at basin outlet." Hydrology and Earth System Sciences, Vol. 14, No. 10, pp. 2011-2022. https://doi.org/10.5194/hess-14-2011-2010
  22. Tran, Q. K., and Song, S. K. (2017). "Water level forecasting based on deep learning: a use case of rinity river-Texas-The United States." Journal of KIISE, Vol. 44, No. 6, pp. 607-612. https://doi.org/10.5626/JOK.2017.44.6.607
  23. Vilaysane, B., Takara, K., Luo, P., Akkharath, I., and Duan, W. (2015). "Hydrological stream flow modelling for calibration and uncertainty analysis using SWAT model in the Xedone river basin, Lao PDR." Procedia Environmental Sciences, Vol. 28, pp. 380-390. https://doi.org/10.1016/j.proenv.2015.07.047
  24. Vu, M. T., Raghavan, S. V., and Liong, S. Y. (2012). "SWAT use of gridded observations for simulating runoff - A Vietnam river basin study." Hydrology and Earth System Sciences, Vol. 16, No. 8, pp. 2801-2811. https://doi.org/10.5194/hess-16-2801-2012
  25. Wang, W., Lu, H., Yang, D., Sothea, K., Jiao, Y., Gao, B., Peng, X., and Pang, Z. (2016). "Modelling hydrologic processes in the Mekong river basin using a distributed model driven by satellite precipitation and rain gauge observations." PloS one, Vol. 11, No. 3, e0152229. https://doi.org/10.1371/journal.pone.0152229