LooseCrew-JeffO: GPS accuracy


Ramblings of an adventurous guy living in Denver and playing in the mountains.
For my trail adventures, visit my Trail Bum blog

Thursday, February 07, 2008

GPS accuracy

My father and two brothers are engineers. My initial college study was science (geology, oceanography, sociology, and physics).
My hobby was designing the ultimate rally driving machine, which got me into bizarre mathematics like the Reynolds Number for wind-tunnel testing of scale models. (I tried this at home - don't.)
I'm tech support for a couple of auditors.
My point is, I'm no stranger to the Law of Averages.

If the military wants to add error to civilian GPS, and they decide to make all errors exactly 30 feet west and north, then it would be easy for the civilian market to simply correct by that static amount and be just as accurate as military GPS.
So I'm guessing the military threw in some sort of guarantee of randomness with regard to both how far off and which direction. This was called SA (Selective Availability).

If I was a civilian entreprenuer and wanted to legally circumvent most of this error, I would take multiple simultaneous readings and average them. The faster your travel, the more problematic this would be. The computer computations would have to be fast, frequent, and complex.
To shorten this blog entry, I won't go into all the conciderations. Suffice it to say that it is possible, even with SA, to average multiple readings of a slow-moving GPS unit, and end up with a probabilty-of-accuracy around 5 feet.

But that's just with the ability to average-out the government-imposed errors. You also have to consider error due to the insturment, whether it's an expensive aviation unit or a runner's GPS watch, all units and situations add error.

My only point is, that in an ideal, controlled situation, I don't see why it wouldn't be possible to calculate-out the imposed errors down very close to what the military enjoys.

While "possible", I don't know that these methods are implemented by civilian GPS manufacturers. The faster the unit samples and computes, and the more signals available, the better the posted position.
Wikipedia has a great explanation of GPS error.

Last night was a hellacious wind-storm in Denver. I was going to run 8 but instead all I did was walk to and from the grocery store. If there had been leaves on the trees, limbs would've been ripped off all over the Front-Range. If I'd gone running, it would've knocked me down every 10 minutes or so. I swear winds must've been up to 50mph.
And just last week there was a lightening storm with loud cracks and booms. That's the first severe lightening I've witnessed during a snowstorm in my life.

Tonight, I miss more running. My friend, Chris (Paperback) Reiter decided to get old on my trail-running night. Can you believe that? Why would he turn old on a Thursday? No doubt just to irritate me. So consider me irritated.
But I wouldn't miss his 35th - no way!
HAPPY BIRTHDAY, you frickin' geezer! I'll let you borrow some Geritol. If you run out of Depends, I got some extras.


At 9:03 AM, Anonymous Anonymous said...


Thanks so much for coming to my birthday dinner. I know what it means for you to skip a trail running opportunity :-)



Post a Comment

<< Home