A Study on Outlier Adjustment for Multibeam Echosounder Data

다중빔 음향측심기 자료의 이상치 보정에 관한 연구

  • 이정숙 (이화여자대학교 통계학과) ;
  • 김수영 (이화여자대학교 통계학과) ;
  • 이용국 (한국해양연구소 환경기후연구본부) ;
  • 신동완 (이화여자대학교 통계학과) ;
  • 주형태 (한국해양연구소 환경기후연구본부) ;
  • 김한준 (한국해양연구소 환경기후연구본부)
  • Published : 2001.02.28


Multibeam echosounder data, collected to investigate seabed features and topography, are usually subject to outliers resulting from the ship's irregular movements and insufficient correction for pressure calibration to the positions of beams. We introduce a statistical method which adjusts the outliers using the ARMA (Autoregressive Moving Average) technique. Our method was applied to a set of real data acquired in the East Sea. In our approach, autocorrelation of the data is modeled by an AR (1) model. If an observation is substantially different from that obtained from the estimated AR (1) model, it is declared as an outlier and adjusted using the estimated AR (1) model. This procedure is repeated until no outlier is found. The result of processing shows that outliers that are far greater than signals in amplitude were successfully removed.



  1. Biometrika v.66 Bayesian analysis of some outlier problems in time series Abragam, B.;Box, G.E.P.
  2. Technometrics v.30 Estimation of time series parameters in the presence of outliers Chang, I.;C.G. Tiao;C. Chen
  3. J. Am. Stat. Assoc. v.70 Intervention analysis with applications to economic and environmental problems Box, G.E.P.;G.C. Tiao
  4. J. Roy. Stat. Soc. Series B v.34 Outliers in time series Fox, A.J.
  5. Geophysical Signal Analysis Roginson, E.A.
  6. J. Am. Stat. Assoc. v.81 Time series model specification in the presence of outliers Tsay, R.S.
  7. Can. J. Stat. v.6 Effect of correlation on the estimation of a mean in the presence of spurious observations Guttman, I.;G.C. Tiao
  8. Introduction to Statistical Time Series. (2nd ed.) Fuller, W.A.