Originally Posted by yes
The OP also said that he is using software to determine speed. The software has elevation data in it, and could potentially include that into the speed calculation. The gps data that he is keeping is probably points and time rather than points and speed. In either case, the computer could easily correct for elevation.
I think the key word in your statement is 'potentially'. Even if the computer's software does account for elevation, there are other issues that come into play, making a GPS a less than perfect speedometer.
Most modern GPS units do measure altitude, as long as there is reception from four or more satellites. However their software, (in the unit), does not take this altitude into account when calculating speed. This is why the Garmin unit that they recently released, specifically designed for use on a bicycle, uses a wheel magnet to calculate speed.
A hard wired speedometer that measures rolling distance will always be more accurate than a GPS. Civilian GPS units can only display your whereabouts within 3 meters. And to attain that much accuracy, the unit has to be WAAS-enabled and average a stationary reading for a minimum of 120 seconds. For maximum accuracy, using a GPS, the user would be stopping every 3 meters to let the GPS average its location. This doesn't even take into account reception problems that can and do arise from cloud cover and foliage.
Using a GPS to measure speed is fine for most recreational cyclists. But, as I mentioned above, it sounds like the OP wants something more accurate.
