Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

Possibility analysisof future droughts using long short term memory and standardized groundwater level index

LSTM과 SGI를 이용한 미래 가뭄 발생 가능성 분석

  • Lim, Jae Deok (School of Civil and Environmental Engineering, Kookmin University, College of Creative Engineering) ;
  • Yang, Jeong-Seok (School of Civil and Environmental Engineering, Kookmin University, College of Creative Engineering)
  • 임재덕 (국민대학교 창의공과대학 건설시스템공학부) ;
  • 양정석 (국민대학교 창의공과대학 건설시스템공학부)
  • Received : 2020.01.13
  • Accepted : 2020.02.01
  • Published : 2020.02.29
Download PDF
⟨ Previous Next ⟩

Abstract

The purpose of this study is to analyze the possibility of future droughts by calculating the Standardized Groundwater level Index(SGI) after predicting groundwater level using Long Short Term Memory (LSTM) model. The groundwater level of the Kumho River basin was predicted for the next three years by using the LSTM model, and it was validated through RMSE after learning with observation data except the last three years. The temporal SGI was calculated by using the prediction data and the observation data. The calculated SGI was interpolated within the study area, and the spatial SGI was calculated as the average value for each catchment using the interpolated SGI. The possibility of spatio-temporal drought was analyzed using calculated spatio-temporal SGI. It is confirmed that there is a spatio-temporal difference in the possibility of drought. Through the improvement of deep learning model and diversification of validation method, it is expected to obtain more reliable prediction results and the expansion of study area can be used to respond to drought nationwide, and furthermore it can provide important information for future water resource management.

본 연구는 심층학습 기법인 Long Short Term Memory (LSTM)를 이용하여 지하수위를 예측 후 표준지하수위지수(Standardized Groundwater level Index, SGI)를 산정함으로써 미래 가뭄 발생 가능성의 분석을 목적으로 하고 있다. LSTM 모형을 이용하여 금호강 유역의 지하수위를 미래 3년에 대해 예측을 하였으며, 예측시 최근 3년을 제외한 관측 자료로 학습 후 RMSE를 통해 검증하였다. 예측 자료와 관측 자료를 이용하여 시간적 SGI를 산정하였다. 산정된 SGI는 연구 지역 내 보간을 하였고, 보간된 SGI는 소유역별 평균값으로 공간적 SGI를 산정하였다. 산정된 시공간적 SGI를 이용하여 시공간적 가뭄 발생 가능성에 대해 분석하였다. 시공간별로 가뭄 발생 가능성에서 차이가 발생하는 것을 확인하였다. 향후 심층학습 모형의 개선 및 검증 방법의 다양화를 통해 신뢰성이 더욱 높은 예측 결과를 도출할 수 있고, 연구 적용 지역의 확대를 통해 전국적인 가뭄 대응 정책에 활용이 될 수 있으며, 더 나아가 미래 수자원 관리 차원에서 중요한 정보를 제공할 수 있을 것이다.

Keywords

References

  1. Bengio, Y., Simard, P., and Frasconi, P. (1994). "Learning longterm dependencies with gradient descent is difficult." IEEE Transactions on Neural Networks, IEEE Neural Networks Council, Vol. 5, No. 2, pp. 157-166. https://doi.org/10.1109/72.279181
  2. Bloomfield, J.P., and Marchant, B.P. (2013). "Analysis of groundwater drought building on the standardised precipitation index approach." Hydrology and Earth System Sciences, European Geophysical Society, Vol. 17, No. 12, pp. 4769-4788. https://doi.org/10.5194/hess-17-4769-2013
  3. Chang, F.J., Chen, P.A., Lu, Y.R., Huang, E., and Chang, K.Y. (2014). "Real-time multi-step-ahead water level forecasting by recurrent neural networks for urban flood control." Journal of Hydrology, Crossref, Vol. 517, pp. 836-846. https://doi.org/10.1016/j.jhydrol.2014.06.013
  4. Ha, J.H., Lee, Y.H., and Kim, Y.H. (2016). "Forecasting the precipitation of the next day using deep learning." Journal of Korean Institute of Intelligent Systems, Korean Institute of Intelligent Systems, Vol. 26, No. 2, pp. 93-98. https://doi.org/10.5391/JKIIS.2016.26.2.093
  5. Hochreiter, S. (1991). Untersuchungen zu dynamischen neuronalen Netzen. Master's thesis, Technical University of Munich, Germany, pp. 1-65.
  6. Hochreiter, S., and Schmidhuber, J. (1997). "Long short-term memory." Neural Computation, Massachusetts Institute of Technology Press, M.A., U.S., Vol. 9, No. 8, pp. 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  7. Intergovernmental Panel on Climate Change (IPCC) (2015). Climate change 2014 synthesis report. IPCC, Switzerland.
  8. Jang, S.K., Lee, J.K., Kwon, H.H., Oh, J.H., and Jo, J.W. (2017). "Development and application of meteorological and hydrological drought indices for the drought monitoring and forecasting." Journal of Korean Society of Hazard Mitigation, Korean Society of Hazard Mitigation, Vol. 17, No. 6, pp. 37-51. https://doi.org/10.9798/KOSHAM.2017.17.6.37
  9. Jung, S.H., Lee, D.E., and Lee, K.S. (2018). "Prediction of river water level using deep-learning open library." Journal of Korean Society of Hazard Mitigation, Korean Society of Hazard Mitigation, Vol. 18, No. 1, pp. 1-11. https://doi.org/10.9798/KOSHAM.2018.18.1.1
  10. Kim, B.S., Kwon, H.H., and Kim, H.S. (2011). "Impact assessment of climate change on drought risk." Journal of Wetlands Research, Korean Wetlands Society, Vol. 13, No. 1, pp. 1-11.
  11. Kim, J.H., Lee, S.H., and Kim, B.S. (2018). "An assessment of past and future droughts in North Korea using standardized precipitation evapotranspiration index." Crisisonomy, Crisis and Emergency Management : Theory and Praxis, Vol. 14, No. 2, pp. 139-151.
  12. Kumar, R., Musuuza, J.L., Van Loon, A.F., Teuling, A.J., Barthel, R., Broek, J.T., Mai, J., Samaniego, L., and Attinger, S. (2016). "Multiscale evaluation of the standardized precipitation index as a groundwater drought indicator." Hydrology and Earth System Sciences, European Geophysical Society, Vol. 20, No. 3, pp. 1117-1132. https://doi.org/10.5194/hess-20-1117-2016
  13. Lee, J.J., Kang, S.U., Jeong, J.H., and Chun, G.I. (2018). "Development of groundwater level monitoring and forecasting technique for drought analysis (I) - groundwater drought monitoring using standardized groundwater level index (SGI)." Journal of Korea Water Resources Association, Korea Water Resources Association, Vol. 51, No. 11, pp. 1011-1020. https://doi.org/10.3741/JKWRA.2018.51.11.1011
  14. Lee, J.J., Kang, S.U., Kim, T.H., and Chun, G.I. (2018). "Development of groundwater level monitoring and forecasting technique for drought analysis (II) - groundwater drought forecasting using SPI, SGI and ANN." Journal of Korea Water Resources Association, Korea Water Resources Association, Vol. 51, No. 11, pp. 1021-1029. https://doi.org/10.3741/JKWRA.2018.51.11.1021
  15. National Drought Information-Analysis Center (NDIAC) (2019). Republic of Korea, accessed 24 December 2019, .
  16. Song, S.H. (2018). "Assessment of drought effects on groundwater system in rural area using standardized groundwater level index (SGI)." Journal of Soil and Groundwater Environment, Korean Society of Soil and Groundwater Environment, Vol. 23, No. 3, pp. 1-9. https://doi.org/10.7857/JSGE.2018.23.3.001
  17. Tran, Q.K., and Song, S.K. (2016). "Water level forecasting based on deep learning: a use-case of trinity River-Texas-The United States." Journal of Korean Institute of Information Scientists and Engineers, Korean Institute of Information Scientists and Engineers, Vol. 44, No. 6, pp. 607-612.
  18. Water Environment Information System (WEIS) (2019). Republic of Korea, accessed 24 December 2019, .
  19. Yeh, H.F., Chang, C.F., Lee, J.W., and Lee, C.H. (2016). "Using standardized groundwater index and standardized precipitation index to assess drought characteristics of the Kaoping River Basin, Taiwan." Water Resources, Springer, Vol. 46, No. 5, pp. 670-678.
  20. Zhang, D., Lindholm, G., and Ratnaweera, H. (2018). "Use long short-term memory to enhance internet of things for combined sewer overflow monitoring." Journal of Hydrology, Elsevier, Vol. 556, pp. 409-418. https://doi.org/10.1016/j.jhydrol.2017.11.018