How accurate should the co-oridinates be ?

Post by **caughtatwork** » 18 June 04 1:14 pm

Let me ask a question from a newbie.

I've hidden a cache which has been found by 2 and not found by 3.
I checked the cache location this morning and the cache is still there.

When I took the GPSr to the location this morning, I got a slight variation from the co-ords as listed at geocaching.com

ie.
Listed: 37 45.017
Today 37 45.014

I figure that the original co-ords might be 'out' by around 4 meters (if that) which I would have thought were pretty much within the normal range for a GPSr to be.

Should I update the cache co-ords or is this kind of thing pretty normal and I'm worrying for nothing.

caughtatwork.

tolmh · Post by **tolmh** » 18 June 04 1:58 pm

How long did you wait at the spot before taking the coordinates? If you just dash up, drop the cache, glance at the GPSr then wander off, the reading probably won't be terribly accurate. Best to let the GPS settle down and average for at least a few minutes.

Post by **caughtatwork** » 18 June 04 2:01 pm

Both times I gave it around 5 minutes to settle down.
Originally the reading was taken with a Garmin 38 and a EPE of 11 meters.
I now have a Garmin eTrex and I have an EPE of 4 meters.

Sorry, should have mentioned that before.
So, am I worrying for nothing or should I change it?

GEK · Post by **GEK** » 18 June 04 2:12 pm

A variation of .003 of a degree is negligible. If a cache-hunter can't cope with this sort of variation then they shouldn't be in the game.

I've been to many caches where I thought the co-ords were well out, sometimes by up to 20m, and other visitors have agreed. But it hasn't spoilt my enjoyment of the cache at all. In my opinion accurate co-ords are nowhere near as important as having a well-presented cache in a top location.

If you are really concerned about the cache being difficult to find then I would suggest providing better hints rather than fiddling with the co-ords.

GEK

EcoTeam · Post by **EcoTeam** » 18 June 04 8:05 pm

If you are unsure as to whether the co-ords are accurate or not, try asking on the cache page for others to post the co-ords they get, so then others can benefit. This is quite common practice, and is actually encouraged. It is also why to posting system on GC.com has the ability for finders to enter co-ords.
In practice, 5m is considered the norm, and 10m is generally considered "a bit out" and many cachers in this case will post a correction to help others, esp if there have been several no-finds.

There is nothing I find more dissapointing (in a traditional "easy" cache) than doing the hard work getting to a cache location only to leave empty handed. The more help I have the better for traditionals

That is what the hints are for too, many cachers enjoy trying on their own first but if they get stuck it's good to know you can decode the hints and get a good clue.

Most cachers should be able to handle up to 10m out, depending on the location of course.
I recently did a cache that had the cache deliberately placed 20m + away from the coords, needless to say we didn't find it because we were too caught up thinking that it's "never" that far out!

EcoDave

Kerry · Post by **Kerry** » 18 June 04 8:52 pm

4 metres is well within the system accuracy and really not an issue.

As for averaging, well these days (without SA) averaging for the period of time that most are prepared to average is quite frankly a waste of time.

By far, what is more important is actually being at a particualr location to take most advantage of the best conditions and this can basically be predicted in advance.

Cheers, Kerry.

HMAS MB · Post by **HMAS MB** » 18 June 04 10:51 pm

The agency that controls The GPS constellation, will only garuntee the accuary of the network to 100m. You have to remember that your GPSr only "guesses " where it is based on 50000 mathimatical calcutions per sec. Mind you these "guesses" usally are very accurate, and EPE does mean "Estimated" positional error.

Kerry · Post by **Kerry** » 18 June 04 11:06 pm

HMAS MB wrote:The agency that controls The GPS constellation, will only garuntee the accuary of the network to 100m.

100 metres? now that's a little dated isn't it? Perhaps try a 13 metres 95% of the time or in the worst case scenario 36 metres 95% of the time. The other 5% well pick a number any number will do. Mind you they don't actually guarantee anything "on the ground" but Signal-In-Space (SIS) figures.

You have to remember that your GPSr only "guesses " where it is based on 50000 mathimatical calcutions per sec. Mind you these "guesses" usally are very accurate, and EPE does mean "Estimated" positional error.

Yes "Estimated Position Error" is really nothing more than what the software code is made to display as if there was any commonality or standard in EPE then all units should just about show the same, but of course they don't.

50000 mathematical calculations per second? that's an interesting one as if the signal takes around 0.07 seconds to reach a receiver then really one shouldn't expect more than about 14 position solutions per second and even then the receiver would need some fairry heavy CPU power.

The best receiver I've seen was rated at 20Kz, 20 updates per second.

Cheers, Kerry.

HMAS MB · Post by **HMAS MB** » 18 June 04 11:15 pm

Nope thats the current figure quoted by the US Air Force....which in its self is odd because the US Navy always controled Navstar and Navstar 2.

I use the term " Mathimatical Calulations" loosely as the system actully does 50000 Interations a sec which are a combination of receive and process. Given that the first part is to do a Fast Fourier transfomation to determine the doppler shift on each of the up 12 channels then apply the Kalman filter track algorythym to the result, to say calculation is probably more accurate than any other term I can come up with that Non techs will understand.

Kerry · Post by **Kerry** » 19 June 04 11:14 am

HMAS MB wrote:Nope thats the current figure quoted by the US Air Force....which in its self is odd because the US Navy always controled Navstar and Navstar 2.

I use the term " Mathimatical Calulations" loosely as the system actully does 50000 Interations a sec which are a combination of receive and process. Given that the first part is to do a Fast Fourier transfomation to determine the doppler shift on each of the up 12 channels then apply the Kalman filter track algorythym to the result, to say calculation is probably more accurate than any other term I can come up with that Non techs will understand.

Is there any available reference to that figure then? because I'd like to see a handheld that was capable of that.

As for the US Navy, I don;t thing so but USNO have some input into the time keeping side of the system but overall management of GPS comes under the Interagency GPS Executive Board with the US Airforce responsible for system mangagement and with the 50th Space Wing, 2nd Space Operations Squadron being responsible for operating and controlling the system. The space wings all come under the US Airforce.

Oh and don't worry about the non tech stuff

might even clarify what your trying to say.

Cheers, Kerry.

HMAS MB · Post by **HMAS MB** » 19 June 04 3:07 pm

When I did my first Magnavox Transit/GPS course in 1985 the US Navy controlled it. It was designed so the the Boomers- Ballistic Nuclear submarines could get an accurate targeting fix prior to lauching. Altough under Transit the mast had to stay above the surface for at least 40 mins. When GPS as in the Navstar constellation was first activated, as long as your DR position was within 25 NM of your position the reciever would start to fix within 3 mins...a far better propostion if you are a submarine.

I am trying to find the reference that the US Air Farce put out with the 100 M accuracy. I am aware that NOW it is under Joint control.

As so as I find that ref I will post it.

"M"

Kerry · Post by **Kerry** » 19 June 04 3:27 pm

Transit? or NAVSAT (Navy Navigation Satellite System) now yes that was Navy through and through and actually Transit was still active until 1996 and controlled all through its operation period by the Naval Space Operations Center in California and yes submarines were its primary customers.

Just as a matter of interest the person who commands Schriver and the GPS ops is a female and is USAF all the way.

well that's today's useless bit of info.

Cheers, Kerry.

HMAS MB · Post by **HMAS MB** » 19 June 04 4:19 pm

Navsat is the system Transit was the Constellation.

http://tycho.usno.navy.mil/gpsinfo.html, but as you pointed out it is old.

GeoJo · Post by **GeoJo** » 19 June 04 8:45 pm

I found it today, so now you've got a 50/50 result!

maccamob · Post by **maccamob** » 19 June 04 9:03 pm

EcoTeam wrote:Most cachers should be able to handle up to 10m out, depending on the location of course.

I agree. If my GPSr reads under 10m after I've been standing at the cache for a while, filling in the log, I'm happy. Similarly, I take several averaged readings while placing a cache. The more those readings vary from each other, the more readings I take before averaging the averages to give the final result.