Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

Radar-based rainfall prediction using generative adversarial network

적대적 생성 신경망을 이용한 레이더 기반 초단시간 강우예측

  • Yoon, Seongsim (Department of Hydro Science and Engineering Research, Korea Institute of Civil Engineering and Building Technology) ;
  • Shin, Hongjoon (Hydro Power Department, Korea Hydro & Nuclear Power Co. Ltd.) ;
  • Heo, Jae-Yeong (Department of Civil and Environmental Engineering, Sejong University)
  • 윤성심 (한국건설기술연구원 수자원하천연구본부) ;
  • 신홍준 (한국수력원자력 수력처) ;
  • 허재영 (세종대학교 공과대학 건설환경공학과)
  • Received : 2023.04.18
  • Accepted : 2023.07.27
  • Published : 2023.08.31
Download PDF
⟨ Previous Next ⟩

Abstract

Deep learning models based on generative adversarial neural networks are specialized in generating new information based on learned information. The deep generative models (DGMR) model developed by Google DeepMind is an generative adversarial neural network model that generates predictive radar images by learning complex patterns and relationships in large-scale radar image data. In this study, the DGMR model was trained using radar rainfall observation data from the Ministry of Environment, and rainfall prediction was performed using an generative adversarial neural network for a heavy rainfall case in August 2021, and the accuracy was compared with existing prediction techniques. The DGMR generally resembled the observed rainfall in terms of rainfall distribution in the first 60 minutes, but tended to predict a continuous development of rainfall in cases where strong rainfall occurred over the entire area. Statistical evaluation also showed that the DGMR method is an effective rainfall prediction method compared to other methods, with a critical success index of 0.57 to 0.79 and a mean absolute error of 0.57 to 1.36 mm in 1 hour advance prediction. However, the lack of diversity in the generated results sometimes reduces the prediction accuracy, so it is necessary to improve the diversity and to supplement it with rainfall data predicted by a physics-based numerical forecast model to improve the accuracy of the forecast for more than 2 hours in advance.

적대적 생성 신경망 기반의 딥러닝 모델은 학습된 정보를 바탕으로 새로운 정보를 생성하는데 특화되어 있다. 구글 딥마인드에서 개발한 deep generative model of rain (DGMR) 모델은 대규모 레이더 이미지 데이터의 복잡한 패턴과 관계를 학습하여, 예측 레이더 이미지를 생성하는 적대적 생성 신경망 모델이다. 본 연구에서는 환경부 레이더 강우관측자료를 이용하여 DGMR 모델을 학습하고, 2021년 8월 호우사례를 대상으로 적대적 생성 신경망을 이용하여 강우예측을 수행하고 기존 예측기법들과 정확도를 비교하였다. DGMR은 대체적으로 선행 60분까지는 강우 분포 위치가 관측강우와 가장 유사하였으나, 전체 영역에서 강한 강우가 발생한 사례에서는 강우가 지속적으로 발달하는 것으로 예측하는 경향이 있었다. 통계적 평가에서도 DGMR 기법이 1시간 선행예측에서 임계성공지수 0.57~0.79, 평균절대오차 0.57~1.36 mm로 나타나 타 기법 대비 효과적인 강우예측 기법임을 보여주었다. 다만, 생성 결과의 다양성이 부족한 경우가 발생하여 예측 정확도를 저하하므로 다양성을 개선하기 위한 연구와 2시간 이상의 선행예측에 대한 정확도 개선을 위해 물리기반 수치예보모델 예측강우 자료를 이용한 보완이 필요할 것으로 판단되었다.

Keywords

Acknowledgement

본 논문은 한국수력원자력(주)에서 재원을 부담하여 한국건설기술연구원에서 수행한 연구결과입니다(No. 2018-기술-20).

References

  1. Ayzel, G., Scheffer, T., and Heistermann, M. (2020). "RainNet v1.0: A convolutional neural network for radar-based precipitation nowcasting." Geoscientific Model Development, Vol. 13, pp. 2631-2644. https://doi.org/10.5194/gmd-13-2631-2020
  2. Choi, S., and Kim, Y. (2022). "Rad-cGAN v1.0: Radar-based precipitation nowcasting model with conditional generative adversarial networks for multiple dam domains." Geoscientific Model Development, Vol. 15, No. 15, pp. 5967-5985. doi: 10.5194/gmd-15-5967-2022.
  3. Clark, A., Donahue, J., and Simonyan, K. (2019). "Adversarial video generation on complex datasets." arXiv:1907.06571. doi: 10.48550/arXiv.1907.06571.
  4. Espeholt, L., Agrawal, S., Sonderby, C., Kumar, M., Heek, J., Bromberg, C., Gazen, C., Carver, R., Andrychowicz, M., Hickey, J., Bell, A., and Kalchbrenner, N., (2022). "Deep learning for twelve hour precipitation forecasts." Nature Communications, Vol. 13, 5145. doi: 10.1038/s41467-022-32483-x.
  5. Karras, T., Aila, T., Laine, S., and Lehtinen, J. (2017). "Progressive growing of gans for improved quality, stability, and variation." arXiv preprint arXiv:1710.10196.
  6. Kim, Y., and Hong, S. (2022). "Very short-term rainfall prediction using ground radar observations and conditional generative adversarial networks," IEEE Transactions on Geoscience and Remote Sensing, Vol. 60, 4104308. doi: 10.1109/TGRS.2021.3108812.
  7. Lin, C., Vasic, S., Kilambi, A., Turner, B., and Zawadzki, I. (2005). "Precipitation forecast skill of numerical weather prediction mod?els and radar nowcasts." Geophysical Research Letters, Vol. 32, No. 14, L14801, doi: 10.1029/2005GL023451.
  8. Palmer, T.N., and Raisanen, J. (2002). "Quantifying the risk of extreme seasonal precipitation events in a changing climate." Nature, Vol. 415, pp. 512-514. doi: 10.1038/415512a.
  9. Ravuri, S., Lenc, K., Willson, M, Kangin, D., Lam, R., Mirowski, P., Fitzsimons, M., Athanassiadou, M., Kashem, S., Madge, S., Prudden, R., Mandhane, A., Clark, A., Brock, A., Simonyan, K., Hadsell, R., Robinson, N., Clancy, E., Arribas, A., and Mohamed, S. (2021). "Skilful precipitation nowcasting using deep generative models of radar." Nature, Vol. 597, pp. 672-677. doi: 10.1038/s41586-021-03854-z.
  10. Reichstein, M., Camps-Valls, G., Stevens, B., Jung, M., Denzler, J., Carvalhais, N., and Prabhat. (2019). "Deep learning and process understanding for data-driven Earth system science." Nature, Vol. 566, pp. 195-204. doi: 10.1038/s41586-019-0912-1.
  11. Richardson, D.S. (2000). " Skill and relative economic value of the ECMWF ensemble prediction system." Quarterly Journal of the Royal Meteorological Society, Vol. 126, pp. 649-667. https://doi.org/10.1002/qj.49712656313
  12. Seo, H.Y., Lee, G.H., Hyein, M.S., and Won, Y.S. (2020). "Current status of improvement of rainfall radar utilization." Water for future, Korea Water Resources Association, Vol. 53, p. 35-45.
  13. Shi, X., Chen, Z., Wang, H., Yeung, D., Wong, W., and Woo, W. (2015). "Convolutional LSTM Network: A machine learning approach for precipitation nowcasting." In Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 1 (NIPS'15), Montreal, Canada, MIT Press, Cambridge, MA, U.S., pp. 802-810.
  14. Shi, X., Gao, Z., Lausen, L., Wang, H., Yeung D., Wong, W., and Woo, W. (2017). "Deep learning for precipitation nowcasting: A benchmark and a new model." 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, U.S.
  15. Shin, H.J., Yoon, S.S., and Choi, J.M. (2021). "Radar rainfall prediction based on deep learning considering temporal consistency." Journal of Korea Water Resources Association, Vol. 54, No. 5, pp. 301-309. https://doi.org/10.3741/JKWRA.2021.54.5.301
  16. Sun, J., Xue, M., Wilson, J.W., Zawadzki, I., Ballard, S.P., OnvleeHooimeyer, J., Joe, P., Barker, D.M., Li, P.-W., Golding, B., Xu, M., and Pinto, J. (2014). "Use of NWP for nowcasting convective precipitation: recent progress and challenges." Bulletin of the American Meteorological Society, Vol. 95, No. 3 pp. 409-426, doi: 10.1175/BAMS-D-11-00263.1.
  17. Tran, Q.K., and Song, S.K. (2019). "Computer vision in precipitation nowcasting: applying image quality assessment metrics for training deep neural networks." Atmosphere, Vol. 10, No. 5, 244. doi: 10.3390/atmos10050244.
  18. Turner, B.J., Zawadzki, I., and Germann, U. (2004). "Predictability of precipitation from continental radar images. Part III: Operational Nowcasting Implementation (MAPLE)." Journal of Applied Meteorology, Vol. 43, No. 2, pp. 231-248. https://doi.org/10.1175/1520-0450(2004)043<0231:POPFCR>2.0.CO;2
  19. Xie, P.F., Li, X.T., Ji, X.Y., Chen, X.L., Chen, Y.Z., Liu, J., and Ye, Y.M. (2022). "An energy-based generative adversarial forecaster for radar echo map extrapolation." IEEE Geoscience and Remote Sensing Letters, Vol. 19, 3500505. doi: 10.1109/lgrs.2020.3023950.
  20. Yoon, S.S., Park, H.S., and Shin, H.J. (2020). "Very short-term rainfall prediction based on radar image learning using deep neural network." Journal of Korea Water Resources Association, Vol. 53, No. 12, pp. 1159-1172. https://doi.org/10.3741/JKWRA.2020.53.12.1159
  21. Zhang, Y., Long, M., Chen, K, Xing, L., Jin, R., Jordan, M.I., and Wang, J. (2023). "Skilful nowcasting of extreme precipitation with NowcastNet." Nature, Vol. 619, pp. 526-532. doi: 10.1038/s41586-023-06184-4
  22. Zheng, K., Liu, Y., Zhang, J.B., Luo, C., Tang, S.Y., Ruan, H.H., Tan, Q.Y., Yi, Y.L., and Ran, X.T. (2022). "GAN-argcPred Net v1.0: A generative adversarial model for radar echo extrapolation based on convolutional recurrent units." Geoscientific Model Development, Vol. 15, No. 4, pp. 1467-1475. doi: 10.5194/gmd-15-1467-2022.