Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

Streamflow Generation by Boostrap Method and Skewness

Bootstrap 방법에 의한 하천유출량 모의와 왜곡도

  • Published : 2002.06.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this study, a method of random resampling of residuals from stochastic models such as the Monte-Carlo model, the lag-one autoregressive model(AR(1)) and the periodic lag-one autoregressive model(PAR(1)), has been adopted to generate a large number of long traces of annual and monthly steamflows. Main advantage of this resampling scheme called the Bootstrap method is that it does not rely on the assumption of population distribution. The Bootstrap is a method for estimating the statistical distribution by resampling the data. When the data are a random sample from a distribution, the Bootstrap method can be implemented (among other ways) by sampling the data randomly with replacement. This procedure has been applied to the Yongdam site to check the performance of Bootstrap method for the streamflow generation. and then the statistics between the historical and generated streamflows have been computed and compared. It has been shown that both the conventional and Bootstrap methods for the generation reproduce fairly well the mean, standard deviation, and serial correlation, but the Bootstrap technique reproduces the skewness better than the conventional ones. Thus, it has been noted that the Bootstrap method might be more appropriate for the preservation of skewness.

본 연구에서는 Monte-Carlo 모형, AR(1)모형, PAR(1) 모형과 같은 추계학적 모형의 잔차값을 무작위적 복원추출하여 연 및 월 하천 유출량자료를 모의발생하였다. Bootstrap이라고 불리우는 이 복원추출방법은 자료의 모집단의 가정이 필요없다는 장점이 있으며 자료로부터 직접 통계적 분포형을 추정하는 방법으로써 자료의 순위변동법을 이용한다. 본 연구에서는 이 방법을 용담지점에 적용하였으며 Bootstrap 방법으로 모의발생된 하천 유출량자료의 거동을 검토하기 하기 위해 관측 유출량과 모의 발생된 유출량의 통계치를 산정하여 비교하였다. 그 결과 기존의 방범과 Bootstrap 방법 모두 평균, 표준편차, 자기상관성은 잘 재현하였으나 왜곡도 계수의 경우 Bootstrap 방법이 더 뛰어남을 확인할 수 있었다.

Keywords

References

  1. 김형수, 서병하, 정상만, 문영일, 강경석 (1999). '지속 기간별 갈수유량의 최소, 최대치가 확률 갈수량에 미치는 영향.' 99년 한국수자원학회 학술발표회 논문집, 한국수자원학회, pp. 87-92
  2. 윤용남, 이동률 (1987). '河川流量의 模擬?生을 위한 長技 및 短期 推計?的 模型의 比較硏究.' 한국수자원학회논문집, 한국수자원학회, 제20권, 제4호, pp. 257-266
  3. Box, G. E. P., and Cox, D. R. (1964). 'An analysis of Transformation', J.R. Stat. Soc., Vol. 26, pp. 211-243
  4. Cover, K. A., and Unny, T. E. (1986). 'Application of computer intensive statistics to parameter uncertainty in streamflow synthesis.' Water Resour. Bull., Vol 22(3), pp. 495-507 https://doi.org/10.1111/j.1752-1688.1986.tb01905.x
  5. Dahmen, E. R., and Hall, M. J. (1990). Screening of Hydrological Data. ; Tests for Stationarity and Relative Consistence, International Institute of Land and Reclamation
  6. Ezio Todini (1980). 'The presevation of Skewness in linear Disaggregation Schemes.' Journal of Hydrology, Vol. 47, pp. 199-214 https://doi.org/10.1016/0022-1694(80)90093-1
  7. Gary, D. T., and Paul Dunne (1997). 'Bootstrap Position Analysis for Forecasting Low Flow Frequency.' Journal of Water Resources Planning and Management, Nov-Dev pp. 355-367 https://doi.org/10.1061/(ASCE)0733-9496(1997)123:6(359)
  8. Lall, U., and Sharma, A. (1996). 'A nearest neighbor bootstrap for resampling hydrologic time series.' Water Resour. Res., Vol. 32(3), pp. 679-693 https://doi.org/10.1029/95WR02966
  9. Matalas, N. (1967). 'Mathematical assessment of synthetic hydrology.' Water Resour. Res., Vol. 3(4), pp. 937-945 https://doi.org/10.1029/WR003i004p00937
  10. Moss, M. E., and Tasker, G. D. (1991). 'An intercomparision of hydrological network-design technologies.' Hydrological Sci. J., Vol. 36(3), pp. 200-221 https://doi.org/10.1080/02626669109492504
  11. Mthsoft (1999). S-Plus 2000 : Modem Statistics and Advanced Graphics
  12. Pereira, M. V. F., oliveria, G. C. Costa, C. G., and Kelman, J. (1984). 'Stochastic streamflow models for hydroelectric system' Water Resour. Res., Vol. 20(3), pp. 379-390 https://doi.org/10.1029/WR020i003p00379
  13. Sharma, A., Tarboton, D. G., and Lall, U. (1997). 'Streamflow simulation: a non-parametric approach.' Water Resour. Res., Vol. 33(2), pp. 291-308 https://doi.org/10.1029/96WR02839
  14. Srikanthan, R. (1978). 'Sequential Generation of Monthly Streamflows.' Journal of Hydrology, Vol. 38, pp. 71-80 https://doi.org/10.1016/0022-1694(78)90133-6
  15. Srinivasan, K., and M. C. Philipose (1993). 'Stochastic Modeling for Drought Analysis.' ASCE symposium on Engineering, pp. 323-328
  16. Tasker, G. D. (1987). 'Comparison of methods for estimating low flow characteristics of streams.' Water Resour. Bull., Vol. 23(6), 1077-1083 https://doi.org/10.1111/j.1752-1688.1987.tb00858.x
  17. Thomas, H. A., and Burden, R. P. (1976). 'Statistical analysis of reseruoir yieid relation, Chapter 1: Operation sresearch on uxuer quality rnmagemeni,' Final Rep, Cambrige, Mass., pp. 1-21
  18. Vogel, R. M, and Shallcross, A L. (1996). 'The moving blocks bootstrap versus parametric time series models.' Water Resour. Res., Vol. 32(6), pp. 1875-1882 https://doi.org/10.1029/96WR00928
  19. Woo, M. K. (1989). 'Confidence intervals of optimum risk-based hydraulic design parameters.' Water Resour. J., Vol. 14(2), pp. 10-16 https://doi.org/10.4296/cwrj1402010
  20. Zucchini, W., and Adamson, P. T. (1989). 'Bootstrap confidence intervals for des-ign storms from exceedence series.' Hydrological Sci. J., Vol. 34(1), pp. 41-48 https://doi.org/10.1080/02626668909491307