Advertisement

Deutsche Hydrografische Zeitschrift

, Volume 51, Supplement 10, pp 181–185 | Cite as

Statistische Trendanalysen bei marinen Monitoringdaten—Aktuelle Diskussionen

  • Steffen UhligEmail author
Monitoring und Qualitätssicherung
  • 55 Downloads

Zusammenfassung

Ausgangspunkt der vorliegenden Arbeit sind Untersuchungen der ICES-Arbeitsgruppe WGSAEM zur Vereinheitlichung der Trendanalyse von Monitoringzeitreihen. Dabei stehen die folgenden Elemente einer Trendanalyse von Jahreswerten im Vordergrund: eine grafische Darstellung der Zeitreihe, ein formaler Trend-Test, die Bestimmung des linearen Trendanteils sowie die Überprüfung der Differenz zwischen aktuellem Niveau und einem gegebenen Referenzniveau. Dabei ist hervorzuheben, dass nicht alle Elementen grundsätzlich mit demselben statistischen Instrumentarium bearbeitet werden müssen. Verschiedene statistische Methoden zur Realisierung dieser Elemente werden vorgestellt und ihre Stärken und Schwächen diskutiert.

Summary

Based on ICES-WGSAEM investigations on trend assessment, in this paper desirable components of a package of trend assessment tools for the analysis of a series of annual indices are discussed: a graphical presentation of the time series, a formal test of trend, a measure of the tendency to increase/decrease, and a comparison of the current level against some reference level or a level in a previous year. Separation of these components gives greater flexibility in choosing a statistical method for each component which is most appropriate for the purpose of a specific monitoring programme. A review of possible methods is given, guidelines to when they are appropriate, and a summary of their strengths and weaknesses.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Literatur

  1. Cleveland, W.S., 1979: Robust locally-weighted regression and smoothing scatterplots.Journal of the American Statistical Association,74, 829–836.CrossRefGoogle Scholar
  2. Fryer R.J. and Nicholson M.D., 1999: Using smoothers for comprehensive assessments of contaminant time series in marine biota. Submitted to ICESJ. Mar. Sci. Google Scholar
  3. Fryer R.J. and Nicholson M.D. 1999: Assessing Power of Trend Tests for Monotonic Trends. In: ICES-WGSAEM Report 1999, Annex 4.Google Scholar
  4. Hastie T.J. andTibshirani R.J., 1990: Generalized Additive Models. London: Chapman and Hall.Google Scholar
  5. ICES, 1999: ICES-WGSAEM Report 1999.Google Scholar
  6. Kendall, M.G., 1975. Rank Correlation Methods, 4th ed. London: Charles Griffin.Google Scholar
  7. Maxwell, D., 1999: The Compound Mann-Kendall Test. In ICES-WGSAEM Report 1999, Annex 6.Google Scholar
  8. Nicholson, M.D., Fryer R.J. and Maxwell D.L., 1997: Report of the ICES/OSPAR Workshop on the identification of statistical methods for temporal trends. Annex 9. A study of the power of various methods for detecting trends. ICESCM 1997/Env: 11.Google Scholar
  9. Nicholson, M.D., Fryer, R.J. and Maxwell, D., 1997: A study of the power of various methods for detecting trends. Report of the ICES/OSPAR, Workshop on the Identification of Statistical Methods for Trend Detection. ICES CM 1997/Env: 11.Google Scholar
  10. Nicholson, M.D., Fryer, R.J. and Maxwell, D., 1998: The influence of individual outlying observations on four methods of detecting trends. Report of the ICES Working Group on Statistical Aspects of Environmental Monitoring Report of the ICES/OSPAR. ICES CM 1998/E:8.Google Scholar
  11. Uhlig, S., 1999: Assessing Accuracy of Estimators for the Percentage of Reduction of the Contaminant Level. In ICES-WGSAEM Report 1999, Annex 5.Google Scholar

Copyright information

© Bundesamt für Seeschiffahrt und Hydrographie 1999

Authors and Affiliations

  1. 1.Institut für Statistik und ÖkonometrieFreie Universität Berlin FB WirtschaftswissenschaftenBerlin

Personalised recommendations