K index scales

[halojatapäivää]

Hello everyone. I'm working on a project that involves converting magnetometer data into local K indices. I'm however really confused about the official K index limit scales because they don't seem to reproduce consistent probability distributions. 

I downloaded a bunch of archived magnetometer data from SuperMAG. I then processed all the data from Hartland which is one of the magnetometers that is used to calculate the official Kp index. By comparing the local K indices from 1983 to 2019 with the Kp indices from the same time period, the probability distributions ended up looking drastically different. I also ran some data from the Kiruna magnetometer from a slightly different time frame. The distribution was again completely different. I added two bar charts down below to visualize this. The y-axis represents occurrences (%) and x-axis represents indices.

If the reason behind the differences between Kiruna and Hartland is that there’s more geomagnetic activity in Kiruna, then why is K0 also more common there? Or is the K index limit scale simply just flawed (by definition)? Or would the distributions eventually smoothen out? I find it pretty difficult to believe that the latter would be the case with this large of an inconsistency.

I just recently found out about the new H30 index which is like the K index but with higher cadence (30 minutes vs 180 minutes). In this paper (p. 11) the H30 scale is defined in a way that the H30 probability distribution ends up looking exactly as the magnetometer’s K index distribution. What’s noteworthy is that all the individual limit values are set individually. As a result, the different H30 scales end up not being proportional to each other unlike K scales. So should the modern K scales be like this too? Not proportional to each other in order to produce similar probability distributions.

Edited April 21, 2022 by halojatapäivää
Made text more clear

[Vancanneyt Sander]

there are always differences between magnetometer stations and each station has its magnetometer data to local K values.
Did you use officiel Kp potsdam? also don't forget the complete scale 0, 0+, 1-, 1, 1+...

[halojatapäivää]

13 minutes ago, Vancanneyt Sander said:

there are always differences between magnetometer stations and each station has its magnetometer data to local K values.
Did you use officiel Kp potsdam? also don't forget the complete scale 0, 0+, 1-, 1, 1+...

Thanks for the reply.

Yes, I used the official Kp data from Gfz Potsdam. This doesn't however remove the fact that the local K index distribution graphs look completely different especially when compared between magnetic latitudes. I analyzed some more data and I noticed that stations with low limit values (Hartland, Wingst, Fredericksburg, K9=500nt) have a narrow distribution and magnetometers with high limits have a much more spread out distribution (Meanook, Kiruna, Sodankylä). The Kp distribution is in between of these.

It just doesn't seem right to use the same quasi-logarithmic scale at both high and low latitudes.