The material that follows is very tentative and further data is being collected to confirm the validity.

The plot below shows the longitude (East-West) error from smooth the average of 20 days data. Note the measured longitude appears to move east during the day and west during the night. As the error in the plot is the measured value minus the true value, this would correspond with the measured longitude being slightly biased away from the sun.

The latitude plot is quite different. Intuitively, if the sun has an effect, that effect should be roughly the same (whatever it is) near sunrise and sunset. That appears to be roughly true. At about those times, the location appears to be farther south (as the errors in the plot is the measured latitude minus true latitude).

The plot below is the un-smoothed 20 day average of the altitude errors. This plot is shown to compare with the next one to demonstrate how much smoothing improves our ability to perceive the general trends.

The plot below is the smoothed average, as was shown for latitude and longitude. It is a fairly good sine wave. As this was a Garmin, 12XL, the known vertical positive offset is seen. Added to that is a gradual increase in the measured height during the day and a gradual decrease in the measured height at night.

The apparent effects above are tentative as more experimental data needs to be examined.

( Return to http://www.erols.com/dlwilson/gps.htm )