Another GPS question

[Hikin_Jim]

Jim, you're thinking of Selective Availability. This "futzing" was the deliberate degradation of the satellite signals in a manner (the addition of a pseudorandom signal) that only the military could characterize (via knowledge of the seed of the pseudorandom sequence) and reverse.

Selective Availability is no longer active. And will likely not be active again for the reason you note. The "averaging" you refer to is actually the technique known as Differential GPS (DGPS).

As far as I know the only factors currently limiting the accuracy of civilian GPS receivers are the inherent limitations of the GPS system. But there are some good solutions to addressing these limitations. Those civilian users willing to pay for units employing these techniques (geologists, surveyors, precision farmers) can enjoy centimeter-level accuracies.

Zip, unless you're both really motivated and really well funded, these units are generally too heavy and expensive for outdoor use. But there are some indications that this may soon change.

Wow. Thank you.

So, based on the above links, it looks like the "normal" accuracy of a GPS is +/- ten feet -- assuming you can get a signal from at least four satellites simultaneously. If you had less than four satellites, there would presumably be some degradation of the accuracy.

HJ

[lilbitmo]

There's also WAAS that no one has mentioned, some of the units can get an additional signal from ground based "Pre-determined" locations that also send out a signal, since they are a fixed location they add to the accuracy of your units reading as far as your current location where as the satellites are always moving and so are you so your units have to make adjustments, thou minor for those movements, the WAAS or so they say is supposed to eliminate part of that inaccuracy and give better results overall.

They are supposed to help reduce the accuracy of your location down to some 20 feet and my Garmin 60-CSx had that function but I had to turn it on and off as it ate up more battery life.

You can read about how that is supposed to help Here

[turtle]

If you had less than four satellites, there would presumably be some degradation of the accuracy.

Actually, with less than four satellites, it's not merely less accurate. Rather, there is truly no unique solution for your position.

GPS is essentially a trilateration calculation. You need one distance measurement (that is, one satellite) for each of the three spatial dimensions along which you are trying to resolve position.

The fourth satellite is needed because you don't actually measure distance to the satellites directly, but rather time of flight of the signal from the satellite. The fourth measurement allows for a solution in time as well (a fourth dimension of sorts). Without the extra satellite, each individual GPS unit would need an incredibly accurate clock that is synchronized to a clock in each of the satellites.

Adopting this approach has an interesting and beneficial consequence... your GPS unit (by virtue of making the fourth measurement) is also a very, very accurate clock.

[jfr]

Here are a couple of examples of the accuracy and repeatability of gps tracks:


This is a shot of our track last August in northern Yosemite, where the PCT goes past Miller Lake.
No, I did NOT walk in the lake, and I also hiked directly on the trail.



This set of tracks is located in Tuolumne Meadows, just east of the campground entrance, where highway 120 crosses the Tuolumne River.
No, I did NOT jump off the bridge into the water, and I walked on that nice trail toward Lembert Dome, not on the grass.

ACME Mapper locations of the photos

Zip: These shots should give you some scale of the accuracy. If those distances are too far for your worries about getting cliffed-out and terrain-trapped, then be sure to leave something memorable near the end of the tricky sections, and let the gps get you close enough. Making waypoints when you leave a marker might help you find the tough spots before it's too late.

[Hikin_Jim]

That photo is pretty helpful. So, you've got accuracy, but not quite pinpoint accuracy.

HJ

[zippetydude]

I've been in and out of town on an emergency so sorry if I abruptly drop threads.

Thanks to all for the helpful replies. jfr, your post is especially helpful because it gives a pretty clear idea of what kind of accuracy I can expect. Seems to me like it will be quite sufficient. If I'm off my previous course by 20 feet and can't figure out what to do about that cliff in front of me then I shouldn't be out there in the wilderness at all! I'll probably see if I can pick one up used and give it a try as I attempt to route find up to Onion Rock. It's killin' me to look up there and see that big dome just waiting for someone to climb around on it and take some pictures to make everyone else's mouths water.

z

[Zé]

Jim, you're thinking of Selective Availability. This "futzing" was the deliberate degradation of the satellite signals in a manner (the addition of a pseudorandom signal) that only the military could characterize (via knowledge of the seed of the pseudorandom sequence) and reverse.
.

Ooooh some DSP dirty talk I like it!

Z, my 5 yr old Garmin 60CSx generally had accuracy ~ +/- 8 ft outside of deep canyons, and works quite well for a whole bunch of x-country stuff.