Paleomagnetists frequently discarded site mean_directions whose its alpha_95 greater than e.g 20 degrees in order to get a mean_locality direction; is it possible to do it without removing the directions ? For example, by introducing weigth to each component (e.g. proportional to 1/ (alpha_95). This is usually done in ordinary statistics.
there are no strict rules in this case...
so - you can... ))
should just describe your actions and procedures...
imho...
The method used in palaeomagnetism was devised by Fisher (1953) (FISHER, R. 1953. Dispersion on a sphere. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 217, 295-305) In general after bedding-tilt correction, the selected directions group in two antipodal normal and reversed clusters on stereographic projections. Fisher statistics (Fisher, 1953) are used to calculate the mean normal and reversed ChRM directions and their α95 (95% confidence angle). The outliers and transitional directions positioned more than 45° from the normal and reversed means are then rejected. Should you use a different method to this you would need to be able to justify why, as this method is generally recognised and accepted by the palaeomag community. Another method by Vandamme (1994) (VANDAMME, D. 1994. A new method to determine paleosecular variation. Physics of the Earth and Planetary Interiors, 85, 131-142.) could also be used to statistically reduce the error. keeping outliers in will often make a palaeomagnetic correlation more difficult to justify. Hope that helps!
Dear MS Blayney, you are absolutely right. But I still do not understand, why paleomagnetists discard the site means with large A95s when they are calculating the locality mean although they are sure that there was nothing wrong during the field and the lab work.
Paleomagnetist do not discard based on A95. Never. Because A95 is n dependent. Paleomagnetist tend to discard based on the dispersion parameter k (or K if it is on poles). This is not n dependent if the number of samples is 7 or more. very low k (less than 7) means that the obtained directions have a strong component of randomness. If you collect many many samples the A95 will be relatively small, but k will be still very low. Paleomagnetists may also discard sites with very high k (>1000) since it will represent a spot reading of the field which is not suitable for directional analysis.
I just wanted to say there is now a new website that will calculate site means for you. It is very easy to use and provides plots as well as all of the statistics mentioned above. For more info go to http://www.paleomagnetism.org.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India