Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

A comparative study of conceptual model and machine learning model for rainfall-runoff simulation

강우-유출 모의를 위한 개념적 모형과 기계학습 모형의 성능 비교

  • Lee, Seung Cheol (Department of Civil Engineering, Jeonbuk National University) ;
  • Kim, Daeha (Department of Civil Engineering, Jeonbuk National University)
  • 이승철 (전북대학교 토목환경자원에너지공학부) ;
  • 김대하 (전북대학교 토목환경자원에너지공학부)
  • Received : 2023.07.21
  • Accepted : 2023.09.05
  • Published : 2023.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

Recently, climate change has affected functional responses of river basins to meteorological variables, emphasizing the importance of rainfall-runoff simulation research. Simultaneously, the growing interest in machine learning has led to its increased application in hydrological studies. However, it is not yet clear whether machine learning models are more advantageous than the conventional conceptual models. In this study, we compared the performance of the conventional GR6J model with the machine learning-based Random Forest model across 38 basins in Korea using both gauged and ungauged basin prediction methods. For gauged basin predictions, each model was calibrated or trained using observed daily runoff data, and their performance was evaluted over a separate validation period. Subsequently, ungauged basin simulations were evaluated using proximity-based parameter regionalization with Leave-One-Out Cross-Validation (LOOCV). In gauged basins, the Random Forest consistently outperformed the GR6J, exhibiting superiority across basins regardless of whether they had strong or weak rainfall-runoff correlations. This suggest that the inherent data-driven training structures of machine learning models, in contrast to the conceptual models, offer distinct advantages in data-rich scenarios. However, the advantages of the machine-learning algorithm were not replicated in ungauged basin predictions, resulting in a lower performance than that of the GR6J. In conclusion, this study suggests that while the Random Forest model showed enhanced performance in trained locations, the existing GR6J model may be a better choice for prediction in ungagued basins.

최근 기후변화로 인해 유역의 기상자료에 대한 반응이 달라지고 있어 강우-유출 모의에 대한 연구는 중요해지고 있다. 아울러 최근 기계학습 기법에 대한 높은 관심으로 이를 통한 강우-유출 모의 역시 활발하게 증가하고 있으나 기계학습 모형이 전통적으로 사용되어온 개념적 모형에 비해 활용성이 높은지는 아직 확실치 않다. 본 연구에서는 개념적 모형인 GR6J와 기계학습 모형인 Random Forest 성능을 한국 전역의 38개 계측 유역에 대해 계측 유역 예측기법과 미계측 유역 예측기법을 이용해 평가하였다. 먼저 계측 유역 적용기법 평가를 위해 각 모형을 관측 일 유량자료에 학습시키고 분리된 평가기간에 대한 모의성능을 비교하였다. 이후 미계측 유역 모의성능 평가를 위해 인접성 기반 지역화 방법을 Leave-One-Out Cross-Validation (LOOCV)을 이용해 평가하였다. 그 결과 계측 유역 평가에서는 Random Forest 기법이 GR6J 모형보다 일관되게 높은 성능을 보였다. 학습된 데이터를 출력 값으로 재생산하도록 구조화되어 있는 기계학습 기법이 개념적 이론을 통한 모형보다 높은 재현성을 갖기 때문으로 판단된다. 하지만 Random Forest 모형의 성능은 미계측 유역의 예측기법으로는 재현되지 않았고 GR6J 모형보다 성능이 더 낮은 것이 확인되었다. 본 연구는 기계학습 모형은 계측 유역의 유출예측에는 적용성이 높을 수 있으나 미계측 유역에 대한 적용가능성은 전통적인 개념적 모형보다 낮을 수 있음을 제시한다.

Keywords

Acknowledgement

본 연구는 한국수자원공사(K-water)의 개방형 혁신 R&D(OTSK_2022_021) 사업의 일환으로 수행되었습니다.

References

  1. Alpert, P., Ben-Gai, T., Bahard, A., Benjamini, Y., Yekutieli, D., Colacino, M., Diodato, L., Ramis, C., Homar, V., Romero, R., Michaelides, S., and Manes, A. (2002). "The paradoxical increase of Mediterranean extreme daily rainfall in spite of decrease in total values." Geophysical Research Letters, Vol. 29, No. 11, pp. 31-1-31-4. doi: 10.1029/2001gl013554.
  2. Beven, K. (1997). "TOPMODEL: A critique." Hydrological Processes, Vol. 11, No. 9, pp. 1069-1085. doi: 10.1002/(SICI)1099-1085(199707)11:9<1069::AID-HYP545>3.0.CO;2-O.
  3. Breiman, L. (2001). "Random forest." Machine Learning, Vol. 45, No. 1, pp. 5-32. doi: 10.1023/a:1010933404324.
  4. Caloiero, T., Veltri, S., Caloiero, P., and Frustaci, F. (2018). "Drought analysis in Europe and in the Mediterranean Basin Using the standardized precipitation index." Water, Vol. 10, No. 8, 1043. doi: 10.3390/w10081043.
  5. Coron, L., Thirel, G., Delaigue, O., Perrin, C., and Andreassian, V. (2017). "The suite of lumped GR hydrological models in an R package." Environmental Modelling & Software, Vol. 94, pp. 166-171. doi: 10.1016/j.envsoft.2017.05.002.
  6. Daly, C., Halbleib, M., Smith, J.I., Gibson, W.P., Doggett, M.K., Taylor, G.H., Curtis, J., and Pasteris, P.P. (2008). "Physiographically sensitive mapping of climatological temperature and precipitation across the conterminous United States." International Journal of Climatology, Vol. 28, No. 15, pp. 2031-2064. doi: 10.1002/joc.1688.
  7. Dawson, C.W., Abrahart, R.J., Shamseldin, A.Y., and Wilby, R.L. (2006). "Flood estimation at ungauged sites using artificial neural networks." Journal of Hydrology, Vol. 319, No. 1, pp. 391-409. doi: 10.1016/j.jhydrol.2005.07.032.
  8. El-Mottaleb, S.A.A., Metwalli, A., Chehri, A., Ahmed, H.Y., Zeghid, M., and Khan, A.N. (2022). "A QoS classifier based on machine learning for next-generation optical communication." Electronics, Vol. 11, No. 16, 2619. doi: 10.3390/electronics11162619.
  9. Founda, D., Giannakopoulos, C., Pierros, F., Kalimeris, A., and Petrakis, M. (2013). "Observed and projected precipitation variability in Athens over a 2.5 century period." Atmospheric Science Letters, Vol. 14, No. 2, pp. 72-78. doi: 10.1002/asl2.419.
  10. Georgakakos, A.P., Yao, H., Kistenmacher, M., Georgakakos, K.P., Graham, N.E., Cheng, F.Y., Spencer, C., and Shamir, E. (2012). "Value of adaptive water resources management in Northern California under climatic variability and change: Reservoir management." Journal of Hydrology, Vol. 412-413, pp. 34-46. doi: 10.1016/j.jhydrol.2011.04.038.
  11. Ghiggi, G., Humphrey, V., Seneviratne, S.I., and Gudmundsson, L. (2019). "GRUN: An observation-based global gridded runoff dataset from 1902 to 2014." Earth System Science Data, Vol. 11, No. 4, pp. 1655-1674. doi: 10.5194/essd-11-1655-2019.
  12. Ghiggi, G., Humphrey, V., Seneviratne, S.I., and Gudmundsson, L. (2021). "GRUN ENSEMBLE: A multi forcing observation based global runoff reanalysis." Water Resources Research, Vol. 57, No. 5, e2020WR028787. doi: 10.1029/2020wr028787.
  13. Grisel, O., Mueller. A., Lars, Gramfort, A., Louppe, G., Thomas, J.F., Prettenhofer, P., Blondel, M., Niculae, V., Nothman, J., Joly, A., Lemaitre, G., Vanderplas, J., Esteve, L., du Boisberranger, J., Kumar, M., Qin, H., Hug, N., Varoquaux, N., Layton, R., Metzen, J.H., Jalali, A., Rajagopalan, V.R., Schonberger, J., Yurchak, R., Jerphanion, J., la Tour, T.D., Li, W., Marmo, C., and Woolam, C. (2023). scikit-learn/scikit-learn: Scikit-learn 1.3.0, accessed 18 August 2023, <https://doi.org/10.5281/zenodo.8098905>.
  14. Guner Bacanli, U. (2017). "Trend analysis of precipitation and drought in the Aegean region, Turkey." Meteorological Applications, Vol. 24, No. 2, pp. 239-249. doi: 10.1002/met.1622.
  15. Guo, Y., Zhang, Y., Zhang, L., and Wang, Z. (2021). "Regionalization of hydrological modeling for predicting streamflow in ungauged catchments: A comprehensive review." WIREs Water, Vol. 8, No. 1, e1487. doi: 10.1002/wat2.1487.
  16. Gupta, H.V., Kling, H., Yilmaz, K.K., and Martinez, G.F. (2009). "Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling." Journal of Hydrology, Vol. 377, No. 1, pp. 80-91. doi: 10.1016/j.jhydrol.2009.08.003.
  17. Head, T., Kumar, M., Nahrstaedt, H., Louppe, G., and Shcherbatyi, I. (2021). scikit-optimize/scikit-optimize, accessed 18 August 2023, <https://doi.org/10.5281/zenodo.5565057>.
  18. Hrachowitz, M., and Clark, M.P. (2017). "HESS Opinions: The complementary merits of competing modelling philosophies in hydrology." Hydrology and Earth System Sciences, Vol. 21, No. 8, pp. 3953-3973. doi: 10.5194/hess-21-3953-2017.
  19. Hrachowitz, M., Savenije, H.H.G., Bloschl, G., McDonnell, J.J., Sivapalan, M., Pomeroy, J.W., Arheimer, B., Blume, T., Clark, M.P., Ehret, U., Fenicia, F., Freer, J.E., Gelfan, A., Gupta, H.V., Hughes, D.A., Hut, R.W., Montanari, A., Pande, S., Tetzlaff, D., Troch, P.A., Uhlenbrook, S., Wagener T., Winsemius, H.C., Woods, R.A., Zegh E., and Cudennec, C. (2013). "A decade of Predictions in Ungauged Basins (PUB) - a review." Hydrological Sciences Journal, Vol. 58, No. 6, pp. 1198-1255. doi: 10.1080/02626667.2013.803183.
  20. Intergovernmental Panel on Climate Change (IPCC) (2023). Climate change 2022: Impacts, adaptation and vulnerability. Cambridge University Press, Cambridge, UK and New York, NY, U.S., pp. 1-3056, doi: 10.1017/9781009325844.
  21. Jeong, Y.M., and Eum, H.I. (2015). "Application of a statistical interpolation method to correct extreme values in high-resolution gridded climate variables." Journal of Climate Change Research, Vol. 6, No. 4, pp. 331-344. doi: 10.15531/ksccr.2015.6.4.331
  22. Kim, D., Chun, J.A., and Choi, S.J. (2019). "Incorporating the logistic regression into a decision-centric assessment of climate change impacts on a complex river system." Hydrology and Earth System Sciences, Vol. 23, No. 2, pp. 1145-1162. doi: 10.5194/hess-23-1145-2019.
  23. Kim, D., Jung, I.W., and Chun, J.A. (2017). "A comparative assessment of rainfall-runoff modelling against regional flow duration curves for ungauged catchments." Hydrology and Earth System Sciences, Vol. 21, No. 11, pp. 5647-5661. doi: 10.5194/hess-21-5647-2017.
  24. Kim, D., Kim, E., Lee, S.C., Kim, E., and Shin, J. (2022). "A decisioncentric impact assessment of operational performance of the Yongdam Dam, South Korea." Journal of Korea Water Resources, Vol. 55, No. 3, pp. 205-215. doi: 10.3741/jkwra.2022.55.3.205.
  25. Kim, U., and Kaluarachchi, J.J. (2008). "Application of parameter estimation and regionalization methodologies to ungauged basins of the Upper Blue Nile River Basin, Ethiopia." Journal of Hydrology, Vol. 362, No. 1, pp. 39-56. doi: 10.1016/j.jhydrol.2008.08.016
  26. Korea Meteorological Administration (KMA) (2020). Climatological normals of Korea (1991-2020), accessed 1 May 2023, <https://data.kma.go.kr/resources/normals/pdf_data/korea_pdf_0310.pdf>.
  27. Krause, P., Boyle, D.P., and Base, F. (2005). "Comparison of different efficiency criteria for hydrological model assessment." Advances in Geosciences, Vol. 5, pp. 89-97. doi: 10.5194/adgeo-5-89-2005.
  28. Li, H., and Zhang, Y. (2017). "Regionalising rainfall-runoff modelling for predicting daily runoff: Comparing gridded spatial proximity and gridded integrated similarity approaches against their lumped counterparts." Journal of Hydrology, Vol. 550, pp. 279-293. doi: 10.1016/j.jhydrol.2017.05.015.
  29. Liaw, A., and Wiener, M. (2002). "Classification and regression by randomForest." R News, Vol. 2, No. 3, pp. 18-22.
  30. Ma, N., Szilagyi, J., and Zhang, Y. (2021). "Calibration-free complementary relationship estimates terrestrial evapotranspiration globally." Water Resources Research, Vol. 57, No. 9, e2021 WR029691. doi: 10.1029/2021wr029691
  31. McIntyre, N., Lee, H., Wheater, H., Young, A., and Wagener, T. (2005). "Ensemble predictions of runoff in ungauged catchments." Water Resources Research, Vol. 41, No. 12, W12434. doi: 10.1029/2005wr004289.
  32. McMillan, H., Krueger, T., and Freer, J. (2012). "Benchmarking observational uncertainties for hydrology: Rainfall, river discharge and water quality." Hydrological Processes, Vol. 26, No. 26, pp. 4078-4111. doi: 10.1002/hyp.9384.
  33. Mlynski, D., Petroselli, A., and Walega, A. (2018). "Flood frequency analysis by an event-based rainfall-runoff model in selected catchments of Southern Poland." Soil & Water, Vol. 13, No. 3, pp. 170-176. doi: 10.17221/153/2017.
  34. Mohammadi, B. (2021). "A review on the applications of machine learning for runoff modeling." Sustainable Water Resources Management, Vol. 7, No. 6, 98. doi: 10.1007/s40899-021-00584-y
  35. Moussa, R., and Chahinian, N. (2009). "Comparison of different multi-objective calibration criteria using a conceptual rainfall-runoff model of flood events." Hydrology and Earth System Sciences, Vol. 13, No. 4, pp. 519-535. doi: 10.5194/hess-13-519-2009.
  36. Nastos, P.T., Politi, N., and Kapsomenakis, J. (2013). "Spatial and temporal variability of the Aridity Index in Greece." Atmospheric Research, Vol. 119, pp. 140-152. doi: 10.1016/j.atmosres.2011.06.017.
  37. Nearing, G.S., Kratzert, F., Sampson, A.K., Pelissier, C.S., Klotz, D., Frame, J.M., Prieto, C., and Gupta, H.V. (2021). "What role does hydrological science play in the age of machine learning?." Water Resources Research, Vol. 57, No. 3, e2020WR 028091. doi: 10.1029/2020wr028091.
  38. Nielsen, S.A., and Hansen, E. (1973). "Numerical simulation of the rainfall-runoff process on a daily basis." Hydrology Research, Vol. 4, No. 3, pp. 171-190. doi: 10.2166/nh.1973.0013.
  39. Oudin, L., Andreassian, V., Perrin, C., Michel, C., and Le Moine, N. (2008). "Spatial proximity, physical similarity, regression and ungaged catchments: A comparison of regionalization approaches based on 913 French catchments." Water Resources Research, Vol. 44, No. 3, W03413. doi: 10.1029/2007wr006240.
  40. Parajka, J., Viglione, A., Rogger, M., Salinas, J.L., Sivapalan, M., and Bloschl, G. (2013). "Comparative assessment of predictions in ungauged basins-Part 1: Runoff-hydrograph studies." Hydrology and Earth System Sciences, Vol. 17, No. 5, pp. 1783-1795. doi: 10.5194/hess-17-1783-2013.
  41. Perrin, C., Michel, C., and Andreassian, V. (2003). "Improvement of a parsimonious model for streamflow simulation." Journal of Hydrology, Vol. 279, No. 1, pp. 275-289. doi: 10.1016/S0022-1694(03)00225-7.
  42. Priestley, C.H.B., and Taylor, R.J. (1972). "On the assessment of surface heat flux and evaporation using large-scale parameters." Monthly Weather Review, Vol. 100, No. 2, pp. 81-92. doi: 10.1175/1520-0493(1972)100<0081:OTAOSH>2.3.CO;2.
  43. Pushpalatha, R., Perrin, C., Le Moine, N., Mathevet, T., and Andreassian, V. (2011). "A downward structural sensitivity analysis of hydrological models to improve low-flow simulation." Journal of Hydrology, Vol. 411, No. 1, pp. 66-76. doi: 10.1016/j.jhydrol.2011.09.034.
  44. Raje, D., and Mujumdar, P.P. (2011). "A comparison of three methods for downscaling daily precipitation in the Punjab region." Hydrological Processes, Vol. 25, No. 23, pp. 3575-3589. doi: 10.1002/hyp.8083.
  45. Rhee, J., and Cho, J. (2016). "Future changes in drought characteristics: Regional analysis for South Korea under CMIP5 projections." Journal of Hydrometeorology, Vol. 17, No. 1, pp. 437-451. doi: 10.1175/jhm-d-15-0027.1.
  46. Smakhtin, V., and Toulouse, M. (1998). "Relationships between low-flow characteristics of South African streams." WATER SA-PRETORIA- 24, Vol. 24, No. 2, pp. 107-112.
  47. Spinoni, J., Barbosa, P., De Jager, A., McCormick, N., Naumann, G., Vogt, J.V., Magni, D., Masante, D., and Mazzeschi, M. (2019). "A new global database of meteorological drought events from 1951 to 2016." Journal of Hydrology: Regional Studies, Vol. 22, 100593. doi: 10.1016/j.ejrh.2019.100593.
  48. Tyralis, H., Papacharalampous, G., and Langousis, A. (2019). "A brief review of random forests for water scientists and practitioners and their recent history in water resources." Water, Vol. 11, No. 5, 910. doi: 10.3390/w11050910.
  49. United Nations Environment Programme (UNEP) (1997). World atlas of desertification: Second edition. Knowledge Repository - UNEP, Nairobi, Kenya, p. 16.
  50. Westerberg, I.K., Gong, L., Beven, K.J., Seibert, J., Semedo, A., Xu, C.Y., and Halldin, S. (2014). "Regional water balance modelling using flow-duration curves with observational uncertainties." Hydrology and Earth System Sciences, Vol. 18, No. 8, pp. 2993-3013. doi: 10.5194/hess-18-2993-2014.
  51. Westerberg, I.K., Guerrero, J.L., Younger, P.M., Beven, K.J., Seibert, J., Halldin, S., Freer, J.E., and Xu, C.Y. (2011). "Calibration of hydrological models using flow-duration curves." Hydrology and Earth System Sciences, Vol. 15, No. 7, pp. 2205-2227. doi: 10.5194/hess-15-2205-2011.
  52. Zittis, G., Almazroui, M., Alpert, P., Ciais, P., Cramer, W., Dahdal, Y., Fnais, M., Francis, D., Hadjinicolaou, P., Howari, F., Jrrar, A., Kaskaoutis, D.G., Kulmala, M., Lazoglou, G., Mihalopoulos, N., Lin, X., Rudich, Y., Sciare, J., Stenchikov, G., Xoplaki, E., and Lelieveld, J. (2022). "Climate change and weather extremes in the eastern mediterranean and middle east." Reviews of Geophysics, Vol. 60, No. 3, e2021RG000762. doi: 10.1029/2021rg000762