What qualifies as a spike?

I am looking at todays data based on trackpoints and I have a 10 sec speed that breaks 30 knots. Gps display read Max 25.3. On doppler in realspeed it disappears. looking at the data it seems to progress in a feasible pattern rather than an instant peak 24.47,26.07,27.23,28.11,28.8,29.25,29.59,30.05,30.24,29.43.
How does this occur?

Thanks Decrepit
using V1.62 (B1115)
Yesterday getting a two second result that shows 13.634 seconds and distance of 203m. Also top 10 sec shows same data.
Another days 2 sec data shows 7.728 seconds and 120m and 10 sec 10.728 seconds.
Looking forward to using the doppler in future to simplify things.
Altitude shows -77 to +201m I didnt think the swell at Dutchies was that big!

This is the advantage of using the Doppler data. If there is a large discrepancy between the Doppler and the Trackpoint speeds, it is almost certain that there is a problem with the trackpoint speeds. I have seen this happen a few times and when you know what to look for it is quite easy to spot.

This is a very good reason to use the GT-11 and not the Garmin Foretrex/Geko. With the Garmin's, when this happens, you have no way of double checking. I have had tracks sent to me in the past from Garmin Foretrex that were suspect because the speed on one run seemed too good to be true and significantly faster than the max speed readout on the screen. But when you look at the trackpoints they seem to have a normal progression. On other words, they don't look like a 'spike'. This situation leaves you with a dilemma as you have no other proof that it is an error.

If someone went out tomorrow and did a 50 knot run with a Garmin Foretrex, how would we know it was not one of these errors?

At least with the data from a GT-11 we have lots of other ways to validate the speeds.

Of course, one way to validate GPS data is to simply wear two or more GPS. Then if one disagrees significantly from the other you know there is a problem, but with Garmins you still may not know the cause or which one is correct.

I ALWAYS carry two (or more) GPS. There is no way I am willing to risk losing a good session through the failure of a GPS or an error in a GPS so I always have at least one backup and a comparison for validation.

Yesterday I ran a 5hx i747 logger for comparison with a Foretrex and GT-11. When we got home after a good session I found the log in the 5hz logger was blank. I still don't know why but at least I had the Garmin and GT-11 logs so there was no loss apart from the opportunity to compare.

I have had bizzare readings on my Foretrex 101, on both occassions after big stacks, up to 20 track points were recorded which looked potentially legit, save for the fact that I knew I did'nt sail in that spot, and had strange properties like the path was dead straight, or all the speeds were identical to 2 decimal places.

Yes the garmins do that haggar. There's onboard software that hates to admit to being lost. I had Lao's problem once, legitimate looking results that I knew I couldn't have done. To try and figure out what happened I played tricks on my Geko by popping it in a steel can while hiking around an oval.

I found that when it loses satellites it lays out track points at the last speed for about 30 seconds. When it regains reception it will try and blend in the results so you don't notice. If you reverse direction just after popping it in a steel can, it will play catch up when released form the can and thus smoothly record higher speeds than you actually did. So look out for crashes or spikes in the minute or so before any unusually high results. If you're using a garmin and blow a gybe give yourself a minute or two of lazy sailing before you go for it.

I've since found out that when glider pilots first started using gps one smart pilot figured this out and in the competions around landmarks he would wrap the gps in alfoil and execute the turn a hundred or metres or so short of the mark. His downloaded tracks showed him neatly rounding the mark. But he was found out.