Geography portal import

[dpastorgalan]

Konnichiwa,

At the import in the Geography portal there is a "default" import but it is not explained how that file works. Apparently it works in specimens, but which is the order and kind of data can we add (like age, bedding or so on...)

Also, is it going to be possible to add literature data? I would like it but skipping the resampling (which adds 0 value in that particular portal where all the rest are already averages)

Cheers!

[Jollyfant]

The "default" input for data is the same as for the statistics portal (click the example):

https://beta.paleomagnetism.org/statistics/

We should discuss whether this is good enough. How do you mean literature data, mean values + a95?

[dpastorgalan]

Ahhh, Ok! Got it. Well I think that is a quite good format. Unless we and the community adopts other way to give specimen results propagating errors to the statistic level, but for now the format is fair, and easy to understand. The only thing that should be clear is what if the coordinates are given in geographic? I guess that the CSV should say "geographic" but it is not stated in the file.

About literature data, yes I just mean adding Plat, Plong, A95, Age (or Dec, Inc, A95, Age or both?). Maybe in such case the plot should show a different symbol? It is not that I think it is fully necessary, I can still plot it in excel or in R, but there is still many people (and many decent papers) that do not publish specimen results, but I guess since now it is a different portal it is a good idea (?? maybe not)

[Jollyfant]

Maybe I will add the option to add literature data in an update. But you run in to the problem of give me: Dec, Inc, a95.. wait is that A95 or a95?!!?! 🤔

[dpastorgalan]

Well that is not that hard, but there should be some degree of flexibility and "letting be" because that is not the user data. The absolute minimum to plot in map and the graphs is 5 columns, although a NAME, and a REFERENCE (DOI?) would be cool:

Pole_long, Pole_lat, A95, Geographic_LAT, GeographicLONG, {NAME}, {DOI}_

The problem is that many old papers give only a95 for directions and do not give statistics on poles at all or they give dp (error in the colatitude, p) and dm (error in the colongitude, m). It is possible to approximate A95 using a95. dp and dm can be obtained easily:

dp = 1/2*a95(1+3cos^2 p) dm = a95 ((sin p) / (cos I)) --> where I is the mean inclination

Then we can calculate an approximation.

A95 ≈ sqrt (dp * dm)

This is not the best procedure, but it is better than assuming a95 = A95, and allows us to use literature data to compare.

We can add this calculation somewhere (in palpole? in the geographic portal?). People can calculate already Plong and Plat in palpole, we could add the A95 estimation. I think that is useful to use data from literature in the same software and with the possibility of saving it with the file (in which the file should say Literature pole, DOI:10.1139/e71-047 -This is a real 71 paper on paleomagnetism, basically anything has a DOI right now, only very obscure stuff do not have it, in which people can just write: DOI: OBSCURE RUSSIAN POLE blah blah)

--