MK.tempAggr: MK test and Sen slope calculator
In mannkendall/R: MannKendall trends for geophysical time series

Description Usage Arguments Details Value Author(s) References

View source: R/MK_tempAggr.R

'MK_tempAggr' performs the MK test and computes the Sen slope on the given time granularity.

MK.tempAggr(
  data,
  PW.method = "3PW",
  resolution,
  alpha.mk = 95,
  alpha.cl = 90,
  alpha.xhomo = 90,
  alpha.ak = 95,
  seasonal = FALSE,
  seasons = NULL
)

`data`	a data.frame with the first column being a POSIXct object with tz = "UTC" and the second column the variable to be analysed
`PW.method`	the PW method used: e.g. one among: 3PW, PW, TFPW.Y, TFPW.WS and VCTFPW. The default is 3PW.
`resolution`	It is taken into account to determine the number of ties; a good guess it is the resolution of the instrument or a little bit higher. This parameters can be determinant for the results but not very sensitive
`alpha.mk`	confidence limit for Mk test in percentage. Default value is 95
`alpha.cl`	confidence limit for the confidence limits of the Sen's slope in percentage. Default value is 90
`alpha.xhomo`	confidence limit for the homogeneity between seasons in percentage. Default value is 90
`alpha.ak`	confidence limit for the first lag autocorrelation in percentage. Default value is 95
`seasonal`	set to TRUE if the analysis needs to be performed over user-defined seasons (default is FALSE)
`seasons`	a vector of the same lenght of the number of records in data.tempAgg used to split the data into seasons. It is used only if seasonal = TRUE.

The function implements three prewhitening (PW) methods: PW (Yue et al., 2002) and TFPW.Y (trend free PW, Wang and Swail, 2001) to compute the statistical significance and VCTFPW (Wang, W., Chen, Y., Becker, S., \& Liu, B. (2015). Variance Correction Prewhitening Method for Trend Detection in Autocorrelated Data. J. Hydrol. Eng., 20(12), 4015033-1-04015033-10. https://doi.org/10.1061/(ASCE)HE.1943-5584.0001234.) to compute the Sen\'s slope. Only the statistically significant (ss) autocorrelation are taken into account for the prewhitening. The ss of the trends is taken at 95% confidence limit. The upper and lower confidence limits are given by the 90% of the all intervals differences distribution. The significance level is given by the MK test and has therefore no direct relation to the confidences limits. If seasonal Mann-Kendall is applied, the yearly trend is assigned only if the results of the seasonal test are homogeneous

P probability for the statistical significance. If 3PW is applied, P = max(P.PW, P.TFPW.Y)

ss statistical significance: alpha % if the test is ss at the alpha confidence level. Default = 95%. 0 if the test is not ss at the alpha confidence level; -1 if the test is a TFPW.Y false positive at alpha confidence level; -2 if the test is a PW false positive at alpha confidence level

slope Sen's slope in units/y

UCL upper confidence level in units/y

LCL lower confidence level in units/

Martine Collaud Coen (martine.collaud@meteoswiss.ch), MeteoSwiss (CH) and Alessandro Bigi (abigi@unimore.it), University of Modena and Reggio Emilia (IT)

WMO-GAW publication N. 133, annex E, p. 26, MULTMK/PARTMK by C. Libiseller

Gilbert, R. O.: Statistical Methods for Environmental Pollution Monitoring, Van Nostrand Reinhold Company, New York, 1987

Collaud Coen, M., Andrews, E., Bigi, A., Romanens, G., Martucci, G., and Vuilleumier, L.: Effects of the prewhitening method, the time granularity and the time segmentation on the Mann-Kendall trend detection and the associated Sen's slope, Atmos. Meas. Tech., https://doi.org/10.5194/amt-2020-178, 2020.

mannkendall/R documentation built on Feb. 2, 2022, 7:31 p.m.