4. Circular statistics

This vignette teaches you how to retrieve the mean direction of stress datasets.

library(tectonicr)
library(ggplot2) # load ggplot library

Mean direction

Directional data is \(\pi\)-periodical. Thus, for the calculation of mean, the average of 35 and 355\(^{\circ}\) should be 15 instead of 195\(^{\circ}\). tectonicr provides the circular mean (circular_mean()) and the quasi-median (circular_median()) as metrics to describe average direction:

data("san_andreas")
circular_mean(san_andreas$azi)
#> [1] 10.66086
circular_median(san_andreas$azi)
#> [1] 36

Quality weighted mean direction

Because the stress data is heteroscedastic, the data with less precise direction should have less impact on the final mean direction The weighted mean or quasi-median uses the reported measurements weighted by the inverse of the uncertainties:

circular_mean(san_andreas$azi, 1 / san_andreas$unc)
#> [1] 9.971235
circular_median(san_andreas$azi, 1 / san_andreas$unc)
#> [1] 36

The spread of directional data can be expressed by the standard deviation (for the mean) or the quasi-interquartile range (for the median):

circular_sd(san_andreas$azi, 1 / san_andreas$unc) # standard deviation
#> [1] 38.43917
circular_IQR(san_andreas$azi, 1 / san_andreas$unc) # interquartile range
#> [1] 33.10713

Statistics in the Pole of Rotation (PoR) reference frame

Because the \(\sigma_{SHmax}\) orientations are subjected to angular distortions in the geographical coordinate system, it is recommended to express statistical parameters using the transformed orientations of the PoR reference frame.

data("cpm_models")
por <- subset(cpm_models, model == "NNR-MORVEL56") |>
  equivalent_rotation("na", "pa")
san_andreas.por <- PoR_shmax(san_andreas, por, type = "right")

circular_mean(san_andreas.por$azi.PoR, 1 / san_andreas$unc)
#> [1] 137.9861
circular_sd(san_andreas.por$azi.PoR, 1 / san_andreas$unc)
#> [1] 37.30334

circular_median(san_andreas.por$azi.PoR, 1 / san_andreas$unc)
#> [1] 135.5753
circular_IQR(san_andreas.por$azi.PoR, 1 / san_andreas$unc)
#> [1] 25.82388

Rose diagram

tectonicr provides a rose diagram, i.e. histogram for angular data.

rose(san_andreas$azi, weights = 1 / san_andreas$unc, col = "grey", main = "North pole")

rose(san_andreas.por$azi, weights = 1 / san_andreas$unc, col = "grey", main = "PoR")
rose_line(135, radius = 1.1, col = "#009E73") # show the predicted direction

Statistical tests

Test for random distribution

Uniformly distributed orientation can be described by the von Mises distribution (Mardia and Jupp, 1999). If the directions are distributed randomly can be tested with the Rayleigh Test:

rayleigh_test(san_andreas.por$azi.PoR)
#> Reject Null Hypothesis
#> $statistic
#> [1] 0.7165417
#> 
#> $p.value
#> [1] 5.430805e-242
#> 
#> $p.value2
#> [1] 7.551614e-273

Here, the test rejects the Null Hypothesis (statistic > p.value). Thus the \(\sigma_{SHmax}\) directions have a preferred orientation.

Alternative statistical tests for circular uniformity are kuiper_test() and watson_test(). Read help() for more details…

Test for goodness-of-fit

Assuming a von Mises Distribution (circular normal distribution) of the orientation data, a \((1-\alpha \%)/100\) confidence interval can be calculated (Mardia and Jupp, 1999):

confidence_interval(san_andreas.por$azi.PoR, conf.level = 0.95, w = 1 / san_andreas$unc)
#> $mu
#> [1] 137.9861
#> 
#> $conf.angle
#> [1] 4.940328
#> 
#> $conf.interval
#> [1] 133.0458 142.9264

The prediction for the \(\sigma_{SHmax}\) orientation is \(135^{\circ}\). Since the prediction lies within the confidence interval, it can be concluded with 95% confidence that the orientations follow the predicted trend of \(\sigma_{SHmax}\).

The (weighted) circular dispersion of the orientation angles around the prediction can be measured by:

circular_dispersion(san_andreas.por$azi.PoR, y = 135, w = 1 / san_andreas$unc)
#> [1] 0.09769022

The value of the dispersion ranges between 0 and 2.

The standard error and the confidence interval of the calculated circular dispersion can be estimated by bootstrapping via:

circular_dispersion_boot(san_andreas.por$azi.PoR, y = 135, w = 1 / san_andreas$unc, R = 1000)
#> $MLE
#> [1] 0.1909694
#> 
#> $sde
#> [1] 0.009944572
#> 
#> $CI
#> [1] 0.1713139 0.2103890

The statistical test for the goodness-of-fit is the (weighted) Rayleigh Test:

weighted_rayleigh(san_andreas.por$azi.PoR, prd = 135, unc = san_andreas$unc)
#> Reject Null Hypothesis
#> $statistic
#> [1] 0.8046196
#> 
#> $p.value
#> [1] 2.491718e-47

If the Null Hypothesis is rejected, the alternative, that is the distribution around the prediction, cannot be excluded.

References

Mardia, K. V., and Jupp, P. E. (Eds.). (1999). “Directional Statistics” Hoboken, NJ, USA: John Wiley & Sons, Inc. doi: 10.1002/9780470316979.

Ziegler, Moritz O., and Oliver Heidbach. 2017. “Manual of the Matlab Script Stress2Grid” GFZ German Research Centre for Geosciences; World Stress Map Technical Report 17-02. doi: 10.5880/wsm.2017.002.

Tobias Stephan

2024-04-09