gps-wizard
Altitude - WIP
Overview
Big spikes in even Doppler-derived speeds can occur and these may not picked up by either HDOP or satellite filters.
Whilst an issue is usually evident in SDOP / SDOS / sAcc, speed accuracy estimates are not recorded by all GPS devices, including the COROS watches.
Previous experience of Doppler spikes from the GT-11 and GT-31 have indicated that elevation / altitude is a useful diagnostic.
Looking at changes in elevation may be particularly when the acceleration is still within the accepted range.
Example
A GT-31 track from 9 Oct 2012 contains an 81.72 knot Doppler spike which coincided with a notable change in altitude.
HDOP was fine at 0.8 - 1.0 and satellites were decent at 7 - 9. Only the acceleration of 12.5 m/s2 would have resulted in the speed being filtered out.
However, this example still illustrates how altitude can be a useful diagnostic of data reliability and maybe help to identify subtle spikes.
