-------------------------------
Here in Denmark (and Greenland) we allways measure the declination of the target marks through the windows. Else we often would have air turbulence between hot and cold air.
But we also calibrate the target mark (do the sun observation) with closed window to the target mark. (but sun shots through open window to avoid the error)
In this way we avoid the error (or have the same error when we measure the declination of the target mark as when we calibrate the target mark)
Lars W. Pedersen
DTU Space
-------------------------------
Here in Canada I have found the calibration to be stable through winter with +/- 20C outside temperatures. What is the angle of incidence on the glass? Both our target and absolute pier are normal to the window.
David Calp
Physical Scientist, Geomagnetism
Hazards, Adaptation, and Operations Branch
Natural Resources Canada / Government of Canada
-----------------------------------
At Niemegk we made a comparison between observations with glass and
without glass for two azimuth marks. The glass has good optical
characteristics and is clearer than the normal glass windows in Niemegk.
One azimuth mark has an angle of about 65° degrees between the
line-of-sight and the glass. There was a difference (bias) of 0.0007 gon
(corresponding to 0.25 nT) between observations with glass and without
glass.
Another azimuth mark (our main azimuth mark) has a line of sight almost
perpendicular to the glass. Here, we didn't observe a difference (bias)
in the observations with and without glass. We also investigated if
there is a difference in the scatter (noise) of the measurements and
found no difference between observations with and without glass.
These results are based on more than 500 measurements over more than 1
year to investigate potential seasonal changes and no seasonal
dependence was observed.
Dr. Jurgen Matzka
Section 2.3, Geomagnetism
Tel.: +49 (0)33843/624-18
------------------------------------
Here in the US, we use double and triple pane windows at all of our high latitude observatories and use single pane windows at our mid and lower latitude observatories. We try to find a glass company that is capable of producing optical quality glass. We then test the individual panes for refraction and then assemble the window and retest for refraction. Ideally we want no refraction, but will settle for a maximum of +/- 0.1 minutes of refraction. Usually most of our triple pane windows have close to 0.0 minutes of refraction. It is a tedious process but worth it. We also measure through the center of the glass/window and make sure the center is close to the line of sight between the instrument and target. In addition, we also make sure that the window is orthogonal to the line of sight.
Usually our windows are 12 x 18 inches or 30 x 46 cm. We don't want them too big in the colder climates, to reduce heat loss.
Bill Worthington
----------------------------
I think that, perhaps, better results will be obtained with the window opened, just in case that the glass itself causes some kind of distortion. After a few minutes, you can close the window and carry on the magnetic part of measurements, although there is no problem if you leave the window open all the time. So, to avoid any possible error introduced by window glass, keep it open. Good luck, sincerely yours
Dragan Popeskov
----------------------------
Yes, we do the same even in Antarctica.
Regards, Santi
------------------------------------
Hello ,
I think what matters in this optical refraction situation is:
1.- Thickness of glass pane,
2.- Distance from the theo to glass pane and
3.- Angle of incidence on glass pane.
thicker glass gives more error
Shorter distance gives more error
The more departure from 90° incidence the more error.
Have fun!
J.
Jean RASSON, Chef de Travaux Principal Honoraire, Institut Royal Metorologique,
Rue du Centre de Physique, 1,
5670 Viroinval (Dourbes)
---------------------------------------
To illustrate this, I have attached a simple calculator that I have used in the past. The calculation assumes that the front and the back of the glass are parallel (i.e. float glass). If this is not the case then the error angle can be much larger. It also assumes that the refractive index of air is the same inside and out (=1), which is a reasonable approximation for typical air temperatures.
Chris Turbitt
Observatories Manager, Geomagnetism Team
British Geological Survey, The Lyell Centre, Riccarton, Edinburgh EH14 4AP, UK
------------------------------------
You have got lot of comments. Shortly my notes:
We did renovation of the absolute house in 2009. Same time we changed glasses to improve insulation, but the target observation glass was not changed. However, we added extra two glasses, which can be moved away during reading target. Our target is directly trough glasses, angle near zero in horizontal and vertical(!) direction. The installed two glasses packet causes 7 arc sec shift. During the coldest conditions we are not able to measure without these glasses, but at least in 1 position in the end of the observation (glass got frost from moisture rapidly and it takes time to melt) and we can correct it from observations.
Tero Raita
Geofyysikko
Sodankylan geofysiikan observatorio
--------------------------------
For the AutoDIF installation at Canberra observatory we installed a small single-pane window (300 x 300 mm) and a new azimuth pillar. In this situation it was possible to choose the location of the azimuth pillar for perpendicular line-of-sight through the glass.
The aluminium window frame was designed to include adjustment screws so that, during installation, the angle of the glass was fine-tuned for perpendicular angle (to front side of window glass) using the reflected beam of a small laser mounted on a theodolite.
Andrew Lewis
Geomagnetism Lead Scientist | Observe Opps & Data
GEOSCIENCE AUSTRALIA
------------------------------------
Maybe one of solutions:
Completely remove the window, just leave a small hole with cower. Cca
10x10 cm (or even less) is more than enough for targeting the azimuth
mark. When you take absolute observations, during the azimuth mark
readings just remove the cower, take the readings, put the cower back
and close the hole (it lasts 1 minute or less). Continue with D and I
readings. This way you avoid possible errors due to glass refraction and
environmental changes are reduced to minimum.
Igor Mandic, Ph.D.
Department of Geophysics
Faculty of Science
University of Zagreb
---------------------------------------------------
For new installations, we use the "hole" method as Igor named it. Here, a simple PVC tube with a locking cap easy to remove from the inside.
We also have a small aluminium cap (8cm diameter but can be smaller I guess). It is easy to open/close. Mosquitoes have little time to realise it is open during azimuth sighting.
If you have huge temperature difference between inside and outside it might not be best but you have no glass error to calculate and never glass cleaning to make.
Benoit Heumez
Institut de physique du globe de Paris
Observatoire Magnetique National
Carrefour des 8 routes, 45340 Chambon la foret - France
--------------------------------------
