A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Bill Morris
Date: 2017 Jul 3, 15:24 -0700
Noell and David,
The median is preferred over the mean because it is a more robust estimate of central tendency in that it is less sensitive to skewed results and outliers. It also requires no calculation.
Early dedicated aeronautical sextants required the observer to decide when an observation was "good" and to take a reading or press a button. However, a constant acceleration of the aircraft could give a stationary bubble even though vertical was not being sensed by it, so automatic devices that read at regular intervals were devised, that averaged or gave a median, as in the Mk IX A and A10-A , or integrated as in the German SOLD. All that was then required of the observer was to try to keep the bubble centred on the body and to keep the time. The latter could be done with a split seconds stop watch, starting it off the navigational watch, stopping the first hand at the beginning of the observation and stopping the other hand at the end, or with the help of an assistant.
Random errors decrease in proportion to the square root of the number of observations and by about 60 observations, diminishing returns become very prominent. Additionally, the period of slow oscillation of many aircraft, called "phugoid oscillations" by Frederick Lanchester, is about 2 minutes, so 60 observations in 2 minutes tended to become the norm. No doubt this felt a very long time to be flying straight and level over hostile territory and some instruments like the MK IX BM and SOLD allowed for shorter periods.