There is also the statement about independent verification of the magnetic variation data used by Williams. Can anyone point me to the source material that was used to verify this data? Would this have came from the U.S. Hydrographic Office or the U.K. Hydrographic Dept that was producing the Air Almanac?
You spend a lot of time worrying about variation, did you ever wonder how the variation was determined in the first place so that it could be displayed on charts? Actually it is pretty simple, you compute what the true azimuth of the sun is and then measure, with an accurate compass, the magnetic azimuth to the sun and the difference is the variation at that point. Of course, in order to be able to compute the true azimuth of the sun you have to know where you are and in order to get a valid compass reading you must not be near any large metal objects since they distort the earth's magnetic field and give erroneous magnetic compass readings. Then after you get your readings you just connect the points of equal variation and draw in the isogonic lines. To be away from large metal objects you must be standing on dirt, not on the deck of a steel ship. Now you might have noticed that there are not very many spots with dry dirt in the pacific so it is not possible to get many of the variation readings so it is unlikely that you get many with exactly the same variation. So, when you draw in the isogonic lines you end up using a lot if imagination because variation does not necessarily change smoothly between the data points.
It is obviously important on ships to have accurate compasses so normal practice is to compare the ship's compass with the computed azimuth of the sun every morning at sunrise when the sun is low on the horizon making it easier to acccurate measure its azimuth. We can see that they did this on the Itasca. For example, looking at the deck log for July 1, 1937 we see "Gyro compass error 0° by amplitude of sun." (See attached.) Amplitude is a special case of azimuth which is the azimuth at sunrise which can be computed from simpler tables. You find the same entries for the other days in the deck log. Although not in the log they also check the ship's master magnetic compass at the same time.
Here
is a link to a description of variation (declination to surveyors.) Remember that they did not have computer models of variation in 1937. Here is the warning that comes with the models.
"Magnetic reference field models give results that are typically accurate to
about 30 minutes of arc, but the difference between the model value and the true value of the magnetic field at a given location is dependent on a number of factors:
* The accuracy of the model will worsen at locations close to the magnetic poles.
* As the time from the epoch of the model increases, uncertainties in the estimate of secular variation will result in an increasingly large difference.
* Magnetic minerals in local geological formations cause magnetic anomalies that can sometimes be very large. These cannot be reproduced by reference field models.
* Large magnetic storms can temporarily cause large changes in the magnetic field, especially at high latitudes."
But my point is that you put way to much effort into your computations as they relate to variation, which can not be perfectly known at sea.
gl