Magnetometer question

[Jesterface23]

Hello, I want to make sure I am looking at the X or H data right right way on a magnetometer. Off of the attached picture where I added some extra lines, would you go off of the black or red lines to see how far south the aurora can go, or is it a different way?

 

 

Edit: Never mind, after going though a bit of old data, I think I've figured it out.

[Vancanneyt Sander]

Hi and welcome on SWL!

you can start with reading our help article about the Kiruna magnetometer which explains the method used to know the deflection. Each magnetometer has its own unique K-indices with the corresponding deflection to know how strong the aurora is. 

[Marcel de Bont]

Black line

http://www.spaceweatherlive.com/en/help/the-kiruna-magnetometer

[Jesterface23]

Okay, Thanks.

[Waldo Hazeleger]

On 12/1/2015 at 1:52 AM, Marcel de Bont said:

Black line

http://www.spaceweatherlive.com/en/help/the-kiruna-magnetometer

Marcel, if i'm right, the K-index is based on the horizontal component (H) of the magnetic field. See this source from the gfz-potsdam. Magnetic fields are measured by XYZ components or HDF components. This picture explains the components of the magnetic field.

The following formulas exist :

H = sqrt(X^2 + Y^2)

D = arctan (Y / X)

F = sqrt(X^2 + Y^2 + Z^2) = sqrt(H^2 + Z^2)

or if you have HDZ components:

X = H * cos(D)

Y = H * sin(D)

Z = H * tan(I)

So, if you want to calculate the K-index, you take the maximum fluctuations in the H-component within a 3-hour period and convert it to K-index via a conversion table that varies per observatory.

The table for the Boulder magnetometer is:

K	nT
0	0-5
1	5-10
2	10-20
3	20-40
4	40-70
5	70-120
6	120-200
7	200-330
8	330-500
9	>500

So, in the picture in the opening post, the maximum fluctuation is around 200nT, making the K-index 6.

Was this helpfull ?

[Jesterface23]

Yep, its helpful. I'm still partly confused from on the 10th, there was a drop of 35nT in H on the Boulder magnetometer and was K6 while the Fredericksburg magnetometer had a drop of 60 nT and was K4 (K9 is >500nT for both stations). The plot here, http://www.swpc.noaa.gov/products/boulder-magnetometer, has the D component too, so I'm guessing it may have something to do with Boulder having a drop of 21deg and Fredericksburg having a drop of 10deg at the time.

[Waldo Hazeleger]

The K-index is calculated based on the maximum variation in the H-component in a 3-hour period, not only based on one drop.

 

[Jesterface23]

Here are the plots of the H component when Boulder was at K6 at 3-6z on the 10th (Sorry if I am managing to make this overly complicated lol)

[Waldo Hazeleger]

I cannot explain the K6 for boulder. If I look at your graphic and also the magnetometer data for bou and frd, the K should be both 4. The maximum H variation for Boulder is 47.69, the maximum for Frediricksburg is 68.14, making K-index = 4.
When I look at the data from cmo (College), that had a maximum H variation of 283.57 nT, making K=6 (cmo k9 value is 1000). Maybe you are mistaken with this?

[Jesterface23]

Some station K indices are in the files here, ftp://ftp.swpc.noaa.gov/pub/lists/geomag/201512AK.txt.

 

Edit: After going through some of the data again, the pattern isn't the drop in the D component, it looks like it has something to do with the rise or how high the D component angle is. Then that would go along with the variation in the H component.

Another edit: So, while searching for something else I managed to come across this, http://k9la.us/Where_Do_the_K_and_A_Indices_Come_From.PDF. I'm guessing it has changed slightly over 12 years, but that might solve a big chunk of what the D component has to do with the K index. And by using Boulders current chart on the SWPC's page, http://www.swpc.noaa.gov/products/boulder-magnetometer, and the real data, the variation in the D component needs to be multiplied by ~6, not sure about other stations.