Logo


Description of Preliminary MAG-L Data Level 2.3



The preliminary MAG-L data are the first, official Ørsted magnetic vector data released. The data are preliminary since they are not yet decimated to the planned 1 Hz rate, and the data may still contain some errors. Also, not all known disturbances have been corrected for. The vast majority of the data have an accuracy of 7 nT or better, see section on accuracy below. The data are distributed in what is expected to be the format of the final MAG-L data, but note that the preliminary data are not exactly as described in the MAG-L Format Description. See section below for details of the differences. The last section of this document contains a list of known errors and deficiencies of the preliminary MAG-L data.

What's new with regard to Level 2.2

Absolute Accuracy of the Ørsted Vector Data

It is possible to fit a degree/order 13 field model to the vector data with an rms misfit of about 5 nT, which indicates that the data are within specifications. However, the 5 nT rms data noise (corresponding to 9 nT rss) is not equally distributed over the three components. Almost all the noise (8 nT rms) is in the component perpendicular to both the bore-sight of the star imager and the direction of the main field. (For nominal satellite attitude the component of maximum noise is the radial direction at low latitudes and to North-South-direction at high latitudes).

The source of this noise in the vector data is probably the intrinsic noise (of about 30 arcsecs rms) in one of the attitude angles measured by the star imager. We are currently working on methods of data pre-processing (filtering/smoothing of the attitude data) to reduce the effect.

The user of the Oersted vector data should be aware of this noise which might result in relatively large (10 nT amplitude or more) point-to-point variations in one of the vector components.

Differences Between Preliminary and Final MAG-L Data

Errors and Deficiencies of the Preliminary MAG-L Data

Timing

The timestamps of the SIM data have been intensively processed in order to obtain the correct timing. The SIM timing accuracy should now be better than 0.5 ms, with occasional errors of 1 ms. Finally, a static timeshift have been applied to the SIM data based on theoretical and practical studies.

Temperature Correction of Vector Magnetometer

Due to the regular lack of house-keeping data, before July 14, 1999, only 60% of the vector magnetometer (CSC) data before this date are properly temperature corrected.This error is assumed to contribute less than a few nT.

Other Attitude Effects

The noise in the SIM data is an-isotropic causing a 4 to 7-times larger error in the rotation around the boresight angle compared to the rect-ascension and declination angles of the boresight axis itself (as is also described above). This contributes to up to 5 nT rms noise in the vector components.

Boom vibrations are causing aliasing effects in and possibly slight degradation of the attitude. This is probably negligible at night-side, but typically a few effects (10 nT peak-to-peak) per orbit are seen at the day-side.

The Euler angles of the SIM-CSC transformation are expected to be known with an accuracy of 20" or better, which corresponds to a vector uncertainty of less than 5 nT.

Charge Particle counts

CPD counts have been calibrated with the following factors to get UNITS of counts/s cm2 ster.
Pe,med and Pe,high 0.0553 cm2 ster
Pi,med and Pi,high 0.273 cm2 ster

Back The Ørsted Science Data Centre Homepage

Back The Ørsted Homepage


E-Mail address: Osdc@dmi.dk

Last modified 010924 by Lars Tøffner-Clausen.