This data is from the event August 16, 2001. In this event we want to determine the direction of wave front arrivals.
To make the data we culled the time and data columns from the raw data, and put them in the files 3E.12.evt, etc. This meant combining the x, y and z components into a single file, using the script cullData.
When done, we have a series of evt files in the 12 and 34 directories. These files have the time signature, x component, y component, and z component in the four columns.
There are thus 12 such files, for the E, EDot, and/or B fields of each of the four satellites for the 12 and 34 event.
Since the timing from each satellite is different, we want the files so that the times almost match. This means we use the nearest neighbor algorithm to find the sample whose time code is closest to the given time code in each file. Our range of time codes is the max of the min times codes to the min of the max time codes across the four satellites for event 12 and event 34.
To do this, we used the perl script interFiles.pl.
It reads in 4 satellites' files, each with the fields time index, x, y, z columns, and finds the min and max time of the files, and interpolates all data of those files into the directories 12 and 34, sub-directoriees Edot, E, and B, each with files time, 1x, 1y, 1z, 2x, 2y, 2z, 3x, 3y, 3z, 4x, 4y, 4z.
We put zeroes in for missing data.
It is the files of each leaf directory that we cross filter and
Here is a series of plots showing the results.
The results are summarized below.
Click on an image to get the full sized version.
|Name||Data (Time 34208 to 39585||NoMean||Butterworth||AutoCor||Events (Time 34500 to 38000 for 1,2, and 4 and 36500 to 40000 for 3)||XCor Evt 4 : Sat I||Semblance Evt 4 : Sat I|
|EDot 12 X|
|EDot 12 Y|
|EDot 12 Z|
|B 34 X|
|B 34 Y|
|B 34 Z|
|EDot 34 X|
|EDot 34 Y|
|EDot 34 Z|