Help understanding why magnetometer readings are not impacted by starting date.

[MitchCraun]

Hello,

I'm playing with the magnetometer model and have found that the magnetometer readings do not seem to be impacted by the starting date I select. I would assume that the magnetic field readings would be determined by the vehicle's ECEF position. I am specifying a starting position in ECI so altering the starting date should impact the ECI2ECEF transformation and thus my initial ECEF position. However, no mater what starting date I select the magnetometer readings are always the same, almost as if they are determined by the vehicles ECI position not ECEF position. I am attaching my BSK python file. Can someone help me understand what is going on, am I doing something wrong?

-Mitch

MAGNITOMETER_BSK.py

[rmz-oz]

In your script the line that links the field model to the Earth state is commented out, so magModule.planetPosInMsg is never connected. When that message isn't linked, the magnetic field base class keeps a zeroed planet state with J20002Pfix set to identity (magneticFieldBase.cpp#L50-L52) and only reads the real one if the message is linked (L188-L192). The position it uses for the field is then just your ECI position, which is why the start date makes no difference. Your own r_ecef changes because you compute it from the SPICE message yourself, but the dipole model never sees that rotation.

Subscribe it to the SPICE Earth message you already have:

magModule.planetPosInMsg.subscribeTo(gravFactory.spiceObject.planetStateOutMsgs[earthIdx])

I ran your file with Basilisk 2.12.0 and printed the first truth reading for two start times 6 hours apart:

not linked  2019 MAY 01 00:00  B = [-2898.8, -4409.8, 25665.5] nT
not linked  2019 MAY 01 06:00  B = [-2898.8, -4409.8, 25665.5] nT
linked      2019 MAY 01 00:00  B = [ 7646.5,  2560.8, 25672.8] nT
linked      2019 MAY 01 06:00  B = [ 4992.1, -3882.6, 25670.2] nT

With the link the date only enters through Earth's rotation, since the centered dipole uses fixed g10/g11/h11 with no secular variation. If you also want the field itself to change with the date, magneticFieldWMM has an epochInMsg you can feed from the same epoch.

[MitchCraun]

I implemented your fix and now things are working correctly. Thanks for taking the time to explain this to me.