Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

Improvement of precipitation forecasting skill of ECMWF data using multi-layer perceptron technique

다층퍼셉트론 기법을 이용한 ECMWF 예측자료의 강수예측 정확도 향상

  • Received : 2019.03.11
  • Accepted : 2019.06.20
  • Published : 2019.07.31
Download PDF
⟨ Previous Next ⟩

Abstract

Subseasonal-to-Seasonal (S2S) prediction information which have 2 weeks to 2 months lead time are expected to be used through many parts of industry fields, but utilizability is not reached to expectation because of lower predictability than weather forecast and mid- /long-term forecast. In this study, we used multi-layer perceptron (MLP) which is one of machine learning technique that was built for regression training in order to improve predictability of S2S precipitation data at South Korea through post-processing. Hindcast information of ECMWF was used for MLP training and the original data were compared with trained outputs based on dichotomous forecast technique. As a result, Bias score, accuracy, and Critical Success Index (CSI) of trained output were improved on average by 59.7%, 124.3% and 88.5%, respectively. Probability of detection (POD) score was decreased on average by 9.5% and the reason was analyzed that ECMWF's model excessively predicted precipitation days. In this study, we confirmed that predictability of ECMWF's S2S information can be improved by post-processing using MLP even the predictability of original data was low. The results of this study can be used to increase the capability of S2S information in water resource and agricultural fields.

2주에서 2개월까지 선행기간을 가지는 계절내-계절(Subseasonal-to-Seasonal, S2S) 예측결과는 산업전반에 걸쳐 다양한 분야에 활용이 가능할 것으로 기대되고 있으나, 일기예보나 중장기 예보대비 낮은 예측성으로 인하여 현재까지 활용성이 매우 낮은 실정이다. 본 연구에서는 기계학습 기법중 비선형회귀 분야에서 좋은 결과를 보여주는 다층퍼셉트론 기법을 이용하여 S2S 예측자료의 후처리를 통한 국내 영역에서의 강수예측성 향상에 관한 연구를 수행하였다. 후처리 모형의 학습을 위한 입력자료로는 ECMWF의 S2S 과거예측(Hindcast) 정보를 이용하였으며 양분예보기법에 기반하여 학습된 다층퍼셉트론 모델을 이용한 후처리 결과와의 비교 분석이 수행되었다. 비교분석 결과 편차도(Bias score)는 평균 59.7% 감소하였고, 정확도(Accuracy)는 124.3% 증가하였으며, 임계성공지수(Critical Success Index)는 88.5% 향상된 것으로 분석되었다. 탐지확률(Probability of detection)의 경우 원자료 대비 평균 9.5% 감소하였으나 이는 ECMWF의 예측모델이 강수의 발생일을 과도하게 예측하였기 때문인 것으로 분석되었다. 본 연구 수행 결과 비록 ECMWF의 S2S 예측자료의 예측성이 낮더라도 후처리를 통해 예측성을 향상 시킬 수 있음을 확인하였으며, 본 연구 결과는 향후 수자원과 농업 분야에서 S2S 자료의 활용성을 높이는데 도움이 될 수 있을 것으로 판단된다.

Keywords

SJOHCI_2019_v52n7_475_f0001.png 이미지

Fig. 1. Single-layer perceptron and multi-layer perceptron (Lee, 2018)

SJOHCI_2019_v52n7_475_f0002.png 이미지

Fig. 2. Grid point over South Korea (1: 127.5°E, 36°N; 2: 129°E, 36°N; 3: 127.5°E, 37.5°N; 4: 129°E, 37.5°N)

SJOHCI_2019_v52n7_475_f0003.png 이미지

Fig. 3. Comparison of forecasting skills

Table 1. Information summary of ECMWF S2S data

SJOHCI_2019_v52n7_475_t0001.png 이미지

Table 2. Contingency table

SJOHCI_2019_v52n7_475_t0002.png 이미지

Table 3. Summary of results

SJOHCI_2019_v52n7_475_t0003.png 이미지

References

  1. Brunet, G., Shapiro, M., Hoskins, B., Moncrieff, M., Dole, R., Kiladis, G. N., Kirtman, B., Lorenc, A., Mills, B., Morss, R., Polavarapu, S., Rogers, D., Schaake, J., and Shukla, J. (2010). "Collaboration of the weather and climate communities to advance subseasonal-to-seasonal prediction." Bulletin of the American Meteorological Society, Vol. 91, No. 10, pp. 1397-1406. https://doi.org/10.1175/2010BAMS3013.1
  2. Cha, Y. M., and Ahn, J. B. (2005). "Evaluation of artificial neural network correction skill on dynamically downscaled summer rainfall over South Korea." Journal of the Korean Meteorological Society, Vol. 41, No. 6, pp. 1125-1135.
  3. de Andrade, F. M., Coelho, C. A., and Cavalcanti, I. F. (2019). "Global precipitation hindcast quality assessment of the subseasonal to seasonal (S2S) prediction project models." Climate Dynamics, Vol. 52, No. 9-10, pp. 5451-5475. https://doi.org/10.1007/s00382-018-4457-z
  4. ECMWF, S. P. (2014). In IFS documentation CY40R1 Part IV: Physical Processes. ECMWF: Reading, UK, pp. 111-113.
  5. 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, Vol. 26, No. 2, pp. 93-98. https://doi.org/10.5391/JKIIS.2016.26.2.093
  6. Hall, T., Brooks, H .E., and Doswell III, C. A. (1999). "Precipitation forecasting using a neural network." Weather and forecasting, Vol. 14, No. 3, pp. 338-345. https://doi.org/10.1175/1520-0434(1999)014<0338:PFUANN>2.0.CO;2
  7. Kang, B. S., and Lee, B. K. (2011). "Application of artificial neural network to improve quantitative precipitation forecasts of meso-scale numerical weather prediction." Journal of Korea Water Resources Association, Vol. 44, No. 2, pp. 97-107. https://doi.org/10.3741/JKWRA.2011.44.2.097
  8. Kim, H. J., Baek, H. J., Kwon, W. T., and Choi, B. S. (2001). "Long-range forecast of precipitation using an internal arithmetic neural network." Journal of the Korean Meterological Society, Vol. 37, No. 5, pp. 443-452.
  9. Lee, S. S. (2018). Development of flood risk analysis technique using seasonal to sub-seasonal data. APEC Climate Center, Annual Report, p. 9.
  10. Nair, A., Singh, G., and Mohanty, U. C. (2018). "Prediction of monthly summer monsoon rainfall using global climate models through artificial neural network technique." Pure and Applied Geophysics, Vol. 175, No. 1, pp. 403-419. https://doi.org/10.1007/s00024-017-1652-5
  11. Robertson, A. W., Kumar, A., Pena, M., and Vitart, F. (2015). "Improving and promoting subseasonal to seasonal prediction." Bulletin of the American Meteorological Society, Vol. 96, No. 3, pp. 49-53. https://doi.org/10.1175/BAMS-D-13-00144.1
  12. Seo, J. H., Lee, Y. H., and Kim, Y. H. (2012). "Feature selection to predict very short-term heavy rainfall based on differential evolution." Journal of Korean Institute of Intelligent Systems, Vol. 22, No. 60, pp. 706-714. https://doi.org/10.5391/JKIIS.2012.22.6.706
  13. Vitart, F., and Robertson, A. W. (2018). "The sub-seasonal to seasonal prediction project (S2S) and the prediction of extreme events." npj Climate and Atmospheric Science, Vol. 1, No. 1, p. 3. https://doi.org/10.1038/s41612-018-0013-0
  14. Vitart, F., Ardilouze, C., Bonet, A., Brookshaw, A., Chen, M., Codorean, C., Deque, M., Ferranti, L., Fucile, E., Fuentes, M., and Hendon, H. (2017). "The subseasonal to seasonal (S2S) prediction project database." Bulletin of the American Meteorological Society, Vol. 98, No. 1, pp. 163-173. https://doi.org/10.1175/BAMS-D-16-0017.1
  15. Vitart, F., Robertson, A. W., and Andderson, D. L. T. (2012). "Subseasonal to seasonal prediction preject: bridging the gap between weather and climate." Bulletin of the World Meteorological Organization, Vol. 61, No. 2, p. 23.
  16. Wilks, D. S. (1995). Statistical methods in the atmospheric sciences. Academic Press, pp. 233-250.
  17. WMO Publication (2015). Seamless prediction of the earth system: from minutes to months.
  18. World Meteorological Organization (WMO) (2012). "WMO bulletin." World Meteorological Organization, Vol. 61, No. 2, p. 48.
  19. World Meteorological Organization (WMO) (2015). Switzerland, accessed 1Februray 2018, .
  20. Xingjian, S. H. I., Chen, Z., Wang, H., Yeung, D. Y., Wong, W. K., and Woo, W. C. (2015). "Convolutional LSTM network: A machine learning approach for precipitation nowcasting." In Advances in neural information processing systems, pp. 802-810.