Forecasting the Sea Surface Temperature in the Tropical Pacific by Neural Network Model

신경망 모델을 이용한 적도 태평양 표층 수온 예측

  • Chang You-Soon (Marine Meteorology & Earthquake Research Laboratory, Meteorological Research Institute) ;
  • Lee Da-Un (Marine Meteorology & Earthquake Research Laboratory, Meteorological Research Institute) ;
  • Seo Jang-Won (Marine Meteorology & Earthquake Research Laboratory, Meteorological Research Institute) ;
  • Youn Yong-Hoon (Marine Meteorology & Earthquake Research Laboratory, Meteorological Research Institute)
  • 장유순 (기상청 기상연구소 해양기상지진연구실) ;
  • 이다운 (기상청 기상연구소 해양기상지진연구실) ;
  • 서장원 (기상청 기상연구소 해양기상지진연구실) ;
  • 윤용훈 (기상청 기상연구소 해양기상지진연구실)
  • Published : 2005.04.01

Abstract

One of the nonlinear statistical modelling, neural network method was applied to predict the Sea Surface Temperature Anomalies (SSTA) in the Nino regions, which represent El Nino indices. The data used as inputs in the training step of neural network model were the first seven empirical orthogonal functions in the tropical Pacific $(120^{\circ}\;E,\;20^{\circ}\;S-20^{\circ}\;N)$ obtained from the NCEP/NCAR reanalysis data. The period of 1951 to 1993 was adopted for the training of neural network model, and the period 1994 to 2003 for the forecasting validation. Forecasting results suggested that neural network models were resonable for SSTA forecasting until 9-month lead time. They also predicted greatly the development and decay of strong E1 Nino occurred in 1997-1998 years. Especially, Nino3 region appeared to be the best forecast region, while the forecast skills rapidly decreased since 9-month lead time. However, in the Nino1+2 region where they are relatively low by the influence of local effects, they did not decrease even after 9-month lead time.

대표적인 엘니뇨 지수인 태평양 Nino 해역의 표층 수온을 예측하기 위해 비선형 통계모델 중의 하나인 신경망 기법을 적용하였다. 신경망 모델 학습 과정의 입력 자료로 1951년부터 1993년까지의 태평양 해역$(120^{\circ}\;E,\;20^{\circ}\;S-20^{\circ}\;N)$ NCEP/NCAR의 재분석 표층 수온 편차의 경험적 직교함수 7개 주모드를 사용하였고, 그 중 1994년부터 2003년까지의 10년 결과를 분석하였다. 모든 해역에서의 9개월까지의 신경망 모델의 예측력은 비교적 우수하였으며, 특히 1997년과 1998년의 강한 엘니뇨의 발달 및 소멸도 잘 예측함을 확인할 수 있었다. 해역별로는 Nino3 지역의 예측성능이 가장 높았으며, 9개월 이후부터는 그 예측력이 급격히 감소하였다. 한편 지역적인 영향이 커 예측력이 낮은 동태평양 연안의 Nino1+2 지역은 9개월 이후에도 예측력의 감소가 관찰되지 않았다.

Keywords

References

  1. 강용균, 2004, 연안 실시간 정보를 이용한 태풍해일 예측. 2004년도 춘계해양학회 요약집, 7-17
  2. 김호준, 백희정, 권원태, 최병철, 2001, 구간 연산 신경망을 이용한 강수량 장기예측 기법. 한국기상학회지, 37 (5), 443-452
  3. 이호만, 이다운, 장유순, 서장원, 2004, 인공신경망 모델을 이용한 폭풍해일예측 연구.2004년도 추계기상학회 요약집, 348-349
  4. Barnett, T.P., Preisendorfer, R., 1987, Origins and levels of monthly and seasonal forecast skill fur United States surface air temp eratures deterrnin ed by canonical c orrelation analysis. Monthly Weather Review, 115, 1825-1850 https://doi.org/10.1175/1520-0493(1987)115<1825:OALOMA>2.0.CO;2
  5. Bamston, A.G, and Coauthors, 1994, Long-lead seasonal forecasts-where do we stand? Bulletin of the American Meteorological Society, 75,2097-2114 https://doi.org/10.1175/1520-0477(1994)075<2097:LLSFDW>2.0.CO;2
  6. Chang, Y.-S., .An, H.S., Jeon, D., Nam, J.-C., and Seo, J.W., 2002, El Nifio signals contained in precipitation at the East China Sea. Climate Research, 23 (1), 31-38 https://doi.org/10.3354/cr023031
  7. Chang, Y-S., Jeon, D., Lee, H., .An, H.S., Seo, J.-W, and Youn, Y-H., 2004, Interannual variability and lagged correlation during strong El Nifio events in the Pacific Ocean. Climate Research, 27 (1), 51-58 https://doi.org/10.3354/cr027051
  8. Derr. V.E., and Slutz, R.J., 1994, Prediction of El Nifio events in the Pacific by means of neural networks. AI Application, 8, 51-63
  9. Elsner,J.B., and Tsonis , A.A., 1992, Nonlinear prediction, chaos, and noise. Bulletin of the American Meteorological Society, 73, 49-60 https://doi.org/10.1175/1520-0477(1992)073<0049:NPCAN>2.0.CO;2
  10. Galton, F. J, 1885, Regression towards mediocrity in hereditary stature. The Journal of the Royal Anthropological Institute, 15, 246-263
  11. Hotelling, H., 1936, Relations between two sets of variates. Biometrika, 28, 321 -366
  12. Lorenz, E.N., 1956, Empirical orthogonal functions and statistical weather predicton. Statistical Forecasting Project, Department of the Meteorology, Massachusetts Institute of Technology, Cambridge, MA, 49pp
  13. Lorenz, E.N., 1984, Irregularity: a fundamental property of the atmosphere. Tellus, 36A, 98-110 https://doi.org/10.1111/j.1600-0870.1984.tb00230.x
  14. Suga, T., Kato, A., and Hanawa, K., 2000, North Pacific Tropical Water: its climatology and temporal changes associated with the climate regime shift in the 1970s. Progress in Oceanography, 47, 223-256 https://doi.org/10.1016/S0079-6611(00)00037-9
  15. Tang. B., Flato, GM., and Holloway, G, 1994, A study of Arctic sea ice and sea-level pressure using POP and neural network methods. Atmosphere-Ocean, 32, 507-529 https://doi.org/10.1080/07055900.1994.9649510
  16. Tangang, F.T., Hsieh, W.W, and Tang, B., 1997, Forecsting the equatorial Pacific sea surface temperature by neural network model. Climate Dynamics, 13, 135-147 https://doi.org/10.1007/s003820050156
  17. Tangang, F.T., Hsieh, W.W., and Tang, B., 1998, Forecasting the regional sea surface temperature of the tropical Pacific by neural network models, with wind stress and sea level pressure as predictors. Journal of Geophysical Research, 103, 7511-7522 https://doi.org/10.1029/97JC03414
  18. Weare, B.C., Navato, A.R., and Newell, R.E., 1976, Empirical orthogonal analysis of Pacific sea surface temperature. Journal of Physical Oceanography, 6, 671-678 https://doi.org/10.1175/1520-0485(1976)006<0671:EOAOPS>2.0.CO;2