Cool thing with the virtual partner and so. I use a regular GPS that tracks my location and speed. Then I download the data to the PC and analyse it using Excel (I made some fancy spreadsheets). Yours stores tracks too, doesn’t it?
I recall other threads in which the altitudes are less accurate so I want to ask, is this happening only during muni when trees or steep slopes block some satellites, or is this a general shortcoming in all riding conditions? Do we have enough gps users here to collect some reporting on the altitude accuracy?
I think this is a general problem for these kind of GPS. Those with barometers are more accurate, or so I’ve heard. If the unit only relies on GPS, you need a good signal from many satellites to get a accurate altitude. It’s not that strange if you think of how it works.
If I recall correctly there already is a thread addressing these issues. Might be worth to do a small search.
I’ve not heard of units with barometers, I’ll have to look into that.
Actually I think the topic is scattered in a few threads, I am hoping this would be an opportunity for a lot of gps users to repeat their experiences here so it’s all in one thread. If I get some time I might actually try a search, collect all I can find and post it here.
I did what I could, I hope that if anyone has additional information about the latest crop of handheld gps units they’ll post it here. Legtod, I hope I haven’t mucked up your thread too much, I’ll go away for a while and (I hope) let others add information to this.
Tried comparing the measured kilometer on my vetta cycle computer and speed.
Both the vetta and forerunner 201 gave same speed and distance.
The Forerunner was much easier to read because my vetta is mounted on my T7 and the Forerunner was velcroed to my KH gloves.
I was able to create custom views on the unit to show distance elevation avg speed. Not too shabby.
I tried downloading the data from the unit and experiment with the graphs and data. Pretty cool stuff. I really like the history option so I can compare ride and data from a ride on my 29" and same ride on my 36".
Anyway, the unit has a 14hr-15 battery life before recharges. Hopefully thats long enough for RTL requirements.
I have some experience in this area. Lack of altitude accuracy is a weakness with terrestrial GPS in general. There are two big reasons:
First, you can never have an ideal satellite geometry for altitude determination. This is because you can never have lock on a GPS satellite below the horizon when taking a reading from the Earth’s surface, which is required to have a vertical position accuracy matching the horizontal position accuracy. Since the vertical accuracy can never be close to optimal, altitude determination is “noisier” than the horizontal determination, which has easier (not impossible) geometric requirements for optimal positioning from the surface. An overhead canopy of leaves can disrupt all aspects position accuracy, but the vertical accuracy will always be worse.
Second, GPS uses an approximation of sea level based upon the WGS84 geoid survey model of the Earth’s surface. This model is globally accurate to about +/- 1 meter, but locally can stray by tens of meters. Defining “mean sea level” for every location on Earth is quite difficult for a number of reasons.
I have an old model Garmin Vista with barometric altimeter. The advantage of the barometer is that it has a better resolution than the GPS altimeter, so if you go up a few metres the barometer will sense that, whereas such a small elevation gain would fall within the noise of the GPS altimeter. However, the barometer is affected by weather (low or high barometer), whereas the GPS gives more of an ‘absolute’ altitude. However, as entropy excellently explained, the GPS altimeter is plagued by inaccuracy, which boils down to the reading jumping up and down around the true value.
The old Vista (and probably also the newer models that hansc posted links to) combines the two altimeters together in a clever way. The GPS altitude is used to calibrate the barometric altimeter, but kind of in a slow way (hyperbolic approximation). Effectively, you have the best of both worlds: the resolution of the barometer with the absolute accuracy of the GPS altimeter. (Not talking about the deviations of the WGS84 datum level to actual mean sea level.) By the way, you can set the old Vista to ignore the barometer and rely on the GPS altimeter only.
However, even with the two altimeters combined, there still is noise in the data. Especially in unstable weather conditions (e.g. with gusty winds, or showers), there are random variations in air pressure that the Vista interprets as going up or down. In the tally that is kept internally to show total elevation gain, these have the effect of overestimating Total Ascent.
In my Excel processing of the Vista data, I have experimented with a threshold level, such that only elevation gains above the threshold get summed up. This would kind of eliminate the elevation noise that is not genuine. This is not very real of course, since a small but genuine elevation gain would also be dismissed, whereas the noise on a bigger elevation gain would count as real elevation gain. But it’s the best I could do since I have no independent handle on the noise of the combined altimeter output. I can set the threshold such that I believe the resulting total elevation gain.
I have a Garmin Foretrex 101 and 201, very similar GPS to the one in the OP. I like to upload my data to Bones In Motion at http://bimactive.com/ which does a really nice job of extracting information from the GPS data, getting altitude data from topographic maps, and showing you your ride on a variety of types of maps.