Survex Digest, Vol 4, Issue 3

[M.J. Green]

> With daily variations in magnetic declination in the UK typically 0.2 - 
> 0.5 degrees (see http://www.geomag.bgs.ac.uk/observatories.html for 
> examples) it seems to me to be good practice to calibrate the compass 
> against a known physical baseline at the start of each day's work. Just 
> taking the published long-term rate of change of declination would be a 
> poor second-best and not something to be encouraged!
> 
> Mike
> 

Hi,

Thanks for pointing out the magnitude of the daily fluctuations. In order 
to make it worth while correcting for these daily variations, the accuracy 
to which you can calculate the daily fluctuation needs to be similar to the 
inaccuracy that you are correcting for.

Using a magnetic compass normally, I would guess that from a single reading 
0.5-1 degree errors are quite common. In order to correct for the 0.2-0.5 
degree errors (decreasing error by a half), then I would guess you would 
need about 4 independent measurements(n^-0.5) to do just as good a job as 
ignoring these deviations. With 9 measurements you would be getting to a 
position where you are probably getting an improvement (in the UK). So it 
seems to me that it would be quite reasonable to either correct for this or 
not, depending on the previous data available, the willingness of surveyors 
to do detailed calibration and the daily variation at the location that you 
are surveying.

Of course in order to determine errors in the compass, some calibration is 
needed, but this data can be collected over multiple survey trip 
calibrations. This is based on the assumption that the compass errors are 
long term rather than short term, due for example by occasionally going out 
of calibration when hit, or perhaps someone has information about that 
assumption.

Regards,
Martin