Notes on the Bz offsets in MFE data

Peter Chi (pchi@igpp.ucla.edu) 10/12/2005

In the first two years (1996-1998) of the Polar mission, the MFE measurements of the magnetic field in the Z component (along the spin axis of the satellite) were calibrated by conducting sensor flips. Starting from May 1998, the flipper did not function properly, and the small deviation (or "offset") in the Z component cannot be estimated except during the occasions when the mirror mode waves were observed or when the satellite was in the solar wind. The offsets in the X and Y components (both in the spin plane) can still be accurately calculated, and the uncalibrated offsets in the Z component can be up to a few nT, which is usually negligible for high-magnetic-field regions but may be of importance for weak-field cases.

In order to estimate the onboard offsets in the Z-component starting from May 1998, we carefully estimated the offsets in Z for several mirror-mode events and solar wind intervals observed by Polar. The estimated offsets exhibit a linear trend, and the best linear fit is:

OBz [nT] = (Year - 1998.297) * 1.237 (between 1600 UT April 19, 1998 and 0200 UT June 11, 1999)

OBz [nT] = (Year - 1998.297) * 1.237 - 1.328 (between 0200 UT June 11, 1999 and 0110 March 14, 2005)

At this moment, the Bz data provided at the UCLA web server as well as at CDAWeb have not been corrected using the estimated offset (except for the intervals starting from December 2004). The correction in these databases will be performed soon after we complete the processing of recent data. In the meantime, users of the MFE data for the intervals between April 20, 1998 and November 30, 2004 are advised that the true Bz in the spacecraft coordinates can be calculated in the following way:

Bz (true) = Bz (data) + OBz

At 0110 UT of March 14, 2005, a command was sent to the Polar MFE instrument for correcting the onboard offsets including that for Bz. Similar commands will be sent for maintaining negligible onboard offsets whenever necessary.