Originally Posted by Blackroady
Surely the slower you go the slower the GPS updates ? - the change in position is slower at slower speed therefore the rate of change of position (which is basically how GPS Updates) means is does not give such an accurate reading ( %) wise .
IN reality - the change from 1-2 knots is a 100% increase in rate of chage of position. At 50 knts to 60 knts its 20% increase - when you go quicker the 'differance' becomes a lower % and therefore more accurate .
Not sure I have explainedthat too well , but the time differance between reciept of GPS signals form the satellites is greater meaning you have a bigger & therefore more accurate number to average ?
Think of tossing a coin twice - in theory 50/50 - in practice very likely to 100/0.
If you do it 100 times far far less likley to be 100/0 & far more likely to be 50/50 .
I woudl bet it is all irrelavant as the screen you look at may not be able to update as quick as the actual GPS ' position' - especially the chart plotting ones. OH and yes I can see the GPS at 50knts ( when its calm enough for me to get there ! )
I think you are wrong!
My GPS updates itself at the same rate regardless of speed (as far as I can tell). The rate at which the boat recieves information from the satellite has nothing to do with its speed. The satelite doesn't even know the boat exists.
Your %age error in speed doesn't make sense. The GPS doesn't measure speed, it measures position. It calculates speed based on two positions at a known time interval. You are correct though at low speed the GPS will have bigger %age errors on the speed but not for the reason you said.
If you are stationary the GPS software knows this and uses repreated positions to average that out. On my Lowrance handheld unit you can clearly see that the EPE (estimated position error) gets lower (better) when it sits in one spot for a little while. [Now given that no two readings are actually identical - how does it know if you are stationary or just moving really slowly? there is a setting somewhere in the software which tells it to treat two measurements within a certain range as being stationary]
For your coin tossing hypothesis - the probabiliy of getting two heads, or two tails is only 50%. It depends if you call that "very likely" or not. Its exactly the same chance as getting one head and one tail.
Finally A good plotter will have a screen refresh rate of 15 fps or better (15 fps is almost a smooth cursor movement). Which would be perfectly adequate for keeping up with a 50 knot speed on a sensible scale.