KSCE Journal of Civil and Environmental Engineering Research (대한토목학회논문집)

Korean Society of Civil Engeneers (대한토목학회)
권호 표지
DOI QR코드

DOI QR Code

Comparing Monthly Precipitation Predictions Using Time Series Analysis with Deep Learning Models

시계열 분석 및 딥러닝 모형을 활용한 월 강수량 예측 비교

  • Chung, Yeon-Ji (Kyonggi University) ;
  • Kim, Min-Ki (Kyonggi University) ;
  • Um, Myoung-Jin (Kyonggi University)
  • 정연지 (경기대학교 토목공학과 ) ;
  • 김민기 (경기대학교 토목공학과) ;
  • 엄명진 (경기대학교 사회에너지시스템공학과)
  • Received : 2024.02.08
  • Accepted : 2024.03.14
  • Published : 2024.08.01
Download PDF
⟨ Previous Next ⟩

Abstract

This study sought to improve the accuracy of precipitation prediction by utilizing monthly precipitation data for each region over the past 30 years. Using statistical models (ARIMA, SARIMA) and deep learning models (LSTM, GBM), we learned monthly precipitation data from 1983 to 2012 in Gangneung, Gwangju, Daegu, Daejeon, Busan, Seoul, Jeju, and Chuncheon. Based on this, monthly precipitation was predicted for 10 years from 2013 to 2022. As a result of the prediction, most models accurately predicted the precipitation trend, but showed a tendency to underpredict the actual precipitation. To solve these problems, appropriate models were selected for each region and season. The LSTM model showed suitable results in Gangneung, Gwangju, Daegu, Daejeon, Busan, Seoul, Jeju, and Chuncheon. When comparing forecasting power by season, the SARIMA model showed particularly suitable forecasting performance in winter in Gangneung, Gwangju, Daegu, Daejeon, Seoul, and Chuncheon. Additionally, the LSTM model showed higher performance than other models in the summer when precipitation is concentrated. In conclusion, closely analyzing regional and seasonal precipitation patterns and selecting the optimal prediction model based on this plays a critical role in increasing the accuracy of precipitation prediction.

본 연구는 지역별로 과거 30년간의 월 강수량 데이터를 활용하여 강수량 예측의 정확성을 높이고자 하였다. 통계적 모형(ARIMA, SARIMA)과 딥러닝 모형(LSTM, GBM)을 사용하여 강릉, 광주, 대구, 대전, 부산, 서울, 제주 그리고 춘천 지역에서 1983년부터 2012년까지의 월 강수량 데이터를 학습하고 이를 바탕으로 2013년부터 2022년까지 10년간 월 강수량을 예측하였다. 예측 결과, 대부분 모형에서 강수량의 추세는 정확하게 예측했으나, 실제 강수량보다 과소 예측하는 경향을 보였다. 이러한 문제점을 해결하고자 지역별, 계절별 적합한 모델을 선정하였다. 강릉, 광주, 대구, 대전, 부산, 서울, 제주 그리고 춘천에는 LSTM 모형이 적합한 결과를 보였다. 계절별로 나누어 예측력을 비교하면, SARIMA 모형은 강릉, 광주, 대구, 대전, 서울 그리고 춘천 지역에서 겨울철에 특히 적합한 예측 성능을 나타냈다. 또한, LSTM 모형은 강수가 집중된 여름철에 다른 모형에 비해 높은 성능을 보였다. 결론적으로, 지역별 및 계절별 강수 패턴을 면밀하게 분석하고 이에 기반한 적합한 예측 모형을 선택하는 것은 강수량 예측의 정확도를 높이는 데 결정적인 역할을 한다.

Keywords

Acknowledgement

This work was supported by Kyonggi University's Graduate Research Assistantship 2024 and this work was also supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. NRF-2022R1A2C2004034).

References

  1. YoosefDoost, A., Sadeghian, M. S., NodeFarahani, M. and Rasekhi, A. (2017). "Comparison between performance of statistical and low cost ARIMA model with GFDL, CM2. 1 and CGM 3 atmosphere-ocean general circulation models in assessment of the effects of climate change on temperature and precipitation in Taleghan Basin." American Journal of Water Resources, Vol. 5, No. 4, pp. 92-99, https://doi.org/10.12691/ajwr-5-4-1.
  2. Choi, I. H., Cha, K. H., Kim, K. M. and Ahn, J.-H. (2023). "Development of forecasting model for machine learning- based landfill leachate generation using linear interpolation." Journal of Korean Society of Environmental Engineers, Vol. 45, No. 1, pp. 11-20, https://doi.org/10.4491/KSEE.2023.45.1.11.
  3. Shenify, M., Danesh, A. S., Gocic, M., Taher, R. S., Abdul Wahab, A. W., Gani, A., Shamshirband, S. and Petkovic, D. (2016). "Precipitation estimation using support vector machine with discrete wavelet transform." Water Resources Management, Vol. 30, pp. 641-652, https://doi.org/10.1007/s11269-015-1182-9.
  4. Yang, C., Zhai, J. and Tao, G. (2020). "Deep learning for price movement prediction using convolutional neural network and long short-term memory." Mathematical Problems in Engineering, Vol. 2020, 2746845, https://doi.org/10.1155/2020/2746845.
  5. Mekanik, F., Imteaz, M. A., Gato-Trinidad, S. and Elmahdi, A. (2013). "Multiple regression and Artificial Neural Network for long-term rainfall forecasting using large scale climate modes." Journal of Hydrology, Vol. 503, pp. 11-21, https://doi.org/10.1016/j.jhydrol.2013.08.035.
  6. Kuligowski, R. J. and Barros, A. P. (1998). "Localized precipitation forecasts from a numerical weather prediction model using artificial neural networks." Weather and Forecasting, Vol. 13, No. 4, pp. 1194-1204, https://doi.org/10.1175/1520-0434(1998)013<1194:LPFFAN>2.0.CO;2.
  7. Kashid, S. S. and Maity, R. (2012). "Prediction of monthly rainfall on homogeneous monsoon regions of India based on large scale circulation patterns using Genetic Programming." Journal of Hydrology, Vol. 454-455, pp. 26-41, https://doi.org/10.1016/j.jhydrol.2012.05.033.
  8. Mirza, M. Q., Warrick, R. A., Ericksen, N. J. and Kenny, G. J. (1998). "Trends and persistence in precipitation in the Ganges, Brahmaputra and Meghna river basins." Hydrological Sciences Journal, Vol. 43, No. 6, pp. 845-858, https://doi.org/10.1080/02626669809492182.
  9. Moustris, K. P., Larissi, I. K., Nastos, P. T. and Paliatsos, A. G. (2011). "Precipitation forecast using artificial neural networks in specific regions of Greece." Water Resources Management, Vol. 25, pp. 1979-1993, https://doi.org/10.1007/s11269-011-9790-5.
  10. Jung, S., Cho, H., Kim, J. and Lee, G. (2018). "Prediction of water level in a tidal river using a deep-learning based LSTM model." Journal of Korea Water Resources Association, Vol. 51, No. 12, pp. 1207-1216, https://doi.org/10.3741/JKWRA.2018.51.12.1207.
  11. Abbot, J. and Marohasy, J. (2014). "Input selection and optimisation for monthly rainfall forecasting in Queensland, Australia, using artificial neural networks." Atmospheric Research, Vol. 138, pp. 166-178, https://doi.org/10.1016/j.atmosres.2013.11.002.
  12. Box, G. E. and Pierce, D. A. (1970). "Distribution of residual autocorrelations in autoregressive-integrated moving average time series models." Journal of the American Statistical Association, Vol. 65, No. 332, pp. 1509-1526, https://doi.org/10.1080/01621459.1970.10481180.
  13. Kim, B. S., Kim, B. K., Kyung, M. S. and Kim, H. S. (2008). "Impact assessment of climate change on extreme rainfall and IDF analysis." Journal of Korea Water Resources Association, Vol. 41, No. 4, pp. 379-394, https://doi.org/10.3741/JKWRA.2008.41.4.379.
  14. Boorman, D. B. and Sefton, C. E. M. (1997). "Recognising the uncertainty in the quantification of the effects of climate change on hydrological response." Climatic Change, Vol. 35, No. 4, pp. 415-434, https://doi.org/10.1023/A:1005372407881.
  15. Venkata Ramana, R., Krishna, B., Kumar, S. R. and Pandey, N. G. (2013). "Monthly rainfall prediction using wavelet neural network analysis." Water Resources Management, Vol. 27, pp. 3697-3711, https://doi.org/10.1007/s11269-013-0374-4.
  16. Park, M., Park, M., Kim, S. and Joo, J. (2013). "Extreme storm estimation by climate change using precipitable water." Journal of the Korean Society of Hazard Mitigation, Vol. 13, No. 1, pp. 121-127, https://doi.org/10.9798/KOSHAM.2013.13.1.121.
  17. Luk, K. C., Ball, J. E. and Sharma, A. (2001). "An application of artificial neural networks for rainfall forecasting." Mathematical and Computer Modelling, Vol. 33, No. 6-7, pp. 683-693, https://doi.org/10.1016/S0895-7177(00)00272-7.
  18. Prudhomme, C., Jakob, D. and Svensson, C. (2003). "Uncertainty and climate change impact on the flood regime of small UK catchments." Journal of Hydrology, Vol. 277, No. 1-2, pp. 1-23, https://doi.org/10.1016/S0022-1694(03)00065-9.
  19. Dimri, T., Ahmad, S. and Sharif, M. (2020). "Time series analysis of climate variables using seasonal ARIMA approach." Journal of Earth System Science, Vol. 129, pp. 1-16, https://doi.org/10.1007/s12040-020-01408-x.
  20. Aliyu, A. S., Auwal, A. M. and Adenomon, M. O. (2021). "Application of Sarima Models in Modelling and Forecasting Monthly Rainfall in Nigeria." Asian Journal of Probability and Statistics, Vol. 13, No. 3, pp. 30-43.
  21. Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q. and Liu, T. Y. (2017). "Lightgbm: A highly efficient gradient boosting decision tree." Advances in Neural Information Processing Systems, 30.
  22. Gellens, D. and Roulin, E. (1998). "Streamflow response of Belgian catchments to IPCC climate change scenarios." Journal of Hydrology, Vol. 210, No. 1-4, pp. 242-258, https://doi.org/10.1016/S0022-1694(98)00192-9.
  23. Zhang, Y., Zhu, C. and Wang, Q. (2020). "LightGBM-based model for metro passenger volume forecasting." IET Intelligent Transport Systems, Vol. 14, No. 13, pp. 1815-1823, https://doi.org/10.1049/iet-its.2020.0396.
  24. Saelthun, N. R. (1998). "Climate Change Impacts on Runoff and Hydropower in the Nordic Countries: Final Report from the Project "Climate Change and Energy Production"." Nordic Council of Ministers.
  25. Dabral, P. P. and Murry, M. Z. (2017). "Modelling and forecasting of rainfall time series using SARIMA." Environmental Processes, Vol. 4, No. 2, pp. 399-419, https://doi.org/10.1007/s40710-017-0226-y.
  26. Bari, S. H., Rahman, M. T., Hussain, M. M. and Ray, S. (2015). "Forecasting monthly precipitation in Sylhet city using ARIMA model." Civil and Environmental Research, Vol. 7, No. 1, pp. 69-77.
  27. Panagoulia, D. and Dimou, G. (1997). "Sensitivity of flood events to global climate change." Journal of Hydrology, Vol. 191, No. 1-4, pp. 208-222, https://doi.org/10.1016/S0022-1694(96)03056-9.
  28. Hyndman, R. J. and Khandakar, Y. (2008). "Automatic time series forecasting: the forecast package for R." Journal of Statistical Software, Vol. 27, pp. 1-22, https://doi.org/10.18637/jss.v027.i03.
  29. Kite, G. W. (1993). "Application of a land class hydrological model to climatic change." Water Resources Research, Vol. 29, No. 7, pp. 2377-2384, https://doi.org/10.1029/93WR00582.
  30. Lukic, B. (2020). Applied Machine Learning Methods with Long-Short Term Memory Based Recurrent Neural Networks for Multivariate Temperature Prediction.
  31. Naz, S. (2015). "Forecasting daily maximum temperature of Umea." Department of Mathematics and Mathematical Statistics, Umea University Master Thesis.
  32. Meehl, G. A. and Tebaldi, C. (2004). "More intense, more frequent, and longer lasting heat waves in the 21st century." Science, Vol. 305, No. 5686, pp. 994-997.
  33. Balibey, M. and Turkyilmaz, S. (2015). "A time series approach for precipitation in Turkey." Gazi University Journal of Science, Vol. 28, No. 4, pp. 549-559.
  34. Box, G. E., Jenkins, G. M., Reinsel, G. C. and Ljung, G. M. (2015). Time series analysis: forecasting and control, John Wiley & Sons.
  35. Cortes, C. and Vapnik, V. (1995). "Support-vector networks." Machine Learning, Vol. 20, pp. 273-297, https://doi.org/10.1007/BF00994018.
  36. Olah, C. (2015). Understanding LSTM networks. https://research.google/pubs/understanding-lstm-networks/
  37. IPCC. (2021). Climate Change 2021: Global Warming of 1.5. IPCC.