"zero-in" Variables

[Exfireguy]

I'm sure all of us have found a site when the GPSr said you were off by some measure. Maybe as much as 50 ft. I didn't attribute that error to anything other than tolerance in the circuitry of my machine compared to another. But, recently I was showing someone what we do and my own receiver had the same degree of error while searching for my own site. Tell me if I'm thinking correctly. Depending on the time of day, the sats that are being used are different than the original ones used for reference. Or, some are waxing and some are waning which puts them in a different aspect than when first used. Just don't tell me that I better get a better receiver! Thanks for any explainations.

[+Miragee]

What kind of GPSr do you have? Was the location near buildings or trees? Was it overcast?

 

My GPSr got 7' accuracy when I was in Joshua Tree National monument, but on an overcast day in a canyon when I was placing a cache, the best I could get was 35'.

 

The other day in a residential area around lots of trees, the best accuracy I could get was about 24'.

[+miles58]

I have been amazed at the consistency in accuracy I have found. Almost all the caches I have done have been inside 25 feet. That is just mind boggling considering the variables involved. The user might not let the unit settle and that will affect the posit. The sattelite geometry can affect accuracy. The variation between units of the same make and model can affec the posit. Variation between different manufacturers can affect posit. Variation in the method used to establish the posit will have an effect. Variation in the overhead cover, weather, clouds, rain, snow, or leaves can have an effect.

 

Considering that about the best you can get is going to be averaged posits that will be accurate to a few feet when repeated by one individual with one GPSr it says a lot that maybe half of the caches I've done when the unit says I am there, I am within three/four feet of the cache.

 

Watch the position on your GPSr and compare it to the EPE. I find that a shakey position does not correlate very well with high EPE. I can have a very high EPE and very stable (and accurate) position, and a low/moderate EPE with jumpy position. Some days the sats and the weather all line up well and you can literally see it when you shift hands with the unit.

[+CompassCollector]

As I understand it (and someone more knowledgable on the actual algorithms used should feel free to correct me) a large part of the variability of accuracy depends on the angle between the sattelites that are 'visible' to the GPSr at the moment. The larger the angle (lower in the sky) then the more accurate. If all of the visible sats are close to zenith (such as in a canyon, or in tree cover), then the calculations are less accurate.

[+jimmyreno]

I've gone caching with a number of other people & all our GPSr give the same reading to within 8' when there is a clear view of the sky.

Garmin, Magellin, doesn't matter.

 

Under a tree cover or near several tall building the signals are either blocked

or reflected so their timing is off. Then you can get all sorts of weird readings.

 

Those "people" who place caches under tree cover should say so in their cache description.

[Exfireguy]

Thanks for the replies. I do understand the relationships between sats that are in an x, y, and z position to my receiver. Once, on a mountaintop I had 10 strong signal bars. I presume that the accuracy would be max at that time. What I'm not sure about is why my same machine would read differently at the same site on different days/times. BTW It is a Magellan eXplorist 200. The only squawks I have about it are short battery life (4 hrs) and the fact that I had to download and print the manual.

[+DaveA]

The satellites are in a non geosynchronous 12 hour orbit. So yes, you can get a different list of sats at different times. see this link

 

Also as the sats orbit their position relative to your GPSr changes.

 

The system was designed to ensure that every spot on the globe would always have a minimum of 4 sats visible, but sometimes you will get many more.

 

As an aside the 16 channel receivers hitting the market are kind of silly. There are, I believe 27 or 29 sats. Unless the number is increased nobody is ever going to track 16 sats at once.

 

I recently did a cache where ground zero moved 80' in nearly opposite directions repeatedly by my simply walking 10'. It was odd and doesn't happen often. I had a mostly clear view of the sky too. Go figure.

[+Miragee]

That happened to us one day. I had my eTrex Vista, my friend had her Garmin Map60CS. We were on top of a very rocky hill in a mostly- residential area with a clear view of the sky. Neither of our GPSrs would show us the way with the arrow.

 

We finally had to just watch the numbers go down as we walked around in circles.

 

It was very strange . . . I have no idea why that hill affected our GPSrs that way.

[+Team Cotati]

There is no way on God's green earth to verify the accuracy of a GPSr unit by using a movable cache container. On the other hand, if you can locate a fixed object placed by an entity for the purpose of establishing a known position on the surface of the earth and which has been recently verified to be accurate by an entity responsible for such verification, then you have found a point on the face of the earth where you can certify your handheld GPSr.

 

Why you would think that doing so is important.......remains a mystery. Well except just for the fun of it, I might even do that. If it was convenient and I just happened to be in the neighborhood. I used to think that USGS benchmarks were useful for such a purpose. Used to that is until I observed a highly intelligent and techinical "discussion" in here a while back that convinced me to the contrary. However since I remain able to locate caches on a regular basis without experiencing significant difficulties, I have concluded that verification of my GPSr is not required. Although I gotta tell ya that it is mighty tempting to go over to my nearby friendly benchmark ever once in a while and check it out. You know, just for the fun of it.

 

Now that I think on it fer a moment or two, "recent" is a relative term...ain't it?

Edited July 31, 2005 by Team Cotati

[+geognerd]

I was just doing some maintenance on a couple of my caches. One cache was actually 40ft off the trail, but my GPSr initially insisted that the cache was on the trail. Once I went off the trail and found the cache's hiding place, I let the GPS sit for a while. After 3 mins, it told me I was 4ft from the cache, which is acceptable. This makes sense, because when I hid the cache originally, I let my receiver sit for a few minutes before taking the coordinates. This has happened when searching for other peoples' caches. I'd find the cache while the GPSr pointed 20ft away and say "Man, these guys put up some sucky coordinates!" But after setting my receiver down, signing the log, and exchanging swag, I'd look at the coordinates on the receiver and they'd be within 0.002 of what the cache owners posted.

[+Team Cotati]

I was just doing some maintenance on a couple of my caches. One cache was actually 40ft off the trail, but my GPSr initially insisted that the cache was on the trail. Once I went off the trail and found the cache's hiding place, I let the GPS sit for a while. After 3 mins, it told me I was 4ft from the cache, which is acceptable. This makes sense, because when I hid the cache originally, I let my receiver sit for a few minutes before taking the coordinates. This has happened when searching for other peoples' caches. I'd find the cache while the GPSr pointed 20ft away and say "Man, these guys put up some sucky coordinates!" But after setting my receiver down, signing the log, and exchanging swag, I'd look at the coordinates on the receiver and they'd be within 0.002 of what the cache owners posted.

Ya, I have had the exact thing happen to me a few times too. I wonder what expalins it. Do you use a Magellan Meridian also?

[+TeamRJJO]

I'm sure all of us have found a site when the GPSr said you were off by some measure. Maybe as much as 50 ft. I didn't attribute that error to anything other than tolerance in the circuitry of my machine compared to another. But, recently I was showing someone what we do and my own receiver had the same degree of error while searching for my own site. Tell me if I'm thinking correctly. Depending on the time of day, the sats that are being used are different than the original ones used for reference. Or, some are waxing and some are waning which puts them in a different aspect than when first used. Just don't tell me that I better get a better receiver! Thanks for any explainations.

Position error fromthe satellites is given as follows:

 

Error = RMS URE x GDOP

 

where RMS URE is the root mean square average of the navigation error coming from each individual satellite and GDOP is dillution of precision as a function of satellite geometry.

 

In general, the URE for a satellite gets higher, thr longer it has been since the satellite was given a fresh navigation upload (usually, an upload is done once every 24 hours). GDOP changes constantly as the satellites move in their orbits. "Ideal geometry" is one satellite directly overhead, and three on the horizon spaced at 120 degrees apart.

 

So, in theory, it's possible to have satellites with fresh uploads, but bad geometry, and get worse position error than satellites with old uploads, but great geometry.

 

You won't have a whole lot of insight into URE, but there are several utilities out there that will allow you to plot DOP over the course of the day using the satellite almanacs as input.

 

There are other contributors to position error as well, such as interference, multipath effects, etc. But all else being equal, geometry and URE are the main contributors.

 

Peace,

TeamRJJO

[+Team Cotati]

And to think, people use this technology to guide aircraft to landings and to navigate cross country thousands of times per hour everday. They must have WAAS enabled, eh? Or is it disabled?

Edited July 31, 2005 by Team Cotati

[+TeamRJJO]

And to think, people use this technology to guide aircraft to landings and to navigate cross country thousands of time per hour everday. They must have WAAS enabled, eh? Or is it disabled?

I'm not really sure how many commercial aircraft use WAAS. Many use LAAS -- local-area augmentation systems. Quite a few airposts use pseudolites to augment GPS. Pseudolites are basically beacons putting out GPS-like signals to augment the coverage provided by the satellite system.

 

Peace,

TeamRJJO

[+Team Cotati]

And to think, people use this technology to guide aircraft to landings and to navigate cross country thousands of time per hour everday. They must have WAAS enabled, eh? Or is it disabled? 

I'm not really sure how many commercial aircraft use WAAS. Many use LAAS -- local-area augmentation systems. Quite a few airposts use pseudolites to augment GPS. Pseudolites are basically beacons putting out GPS-like signals to augment the coverage provided by the satellite system.

 

Peace,

TeamRJJO

Yeah I agree, they must have WAAS disabled and LAAS enabled. Hopefully LAAS is more reliable than WAAS, eh?

[+Team Cotati]

And to think, people use this technology to guide aircraft to landings and to navigate cross country thousands of time per hour everday. They must have WAAS enabled, eh? Or is it disabled? 

I'm not really sure how many commercial aircraft use WAAS. Many use LAAS -- local-area augmentation systems. Quite a few airposts use pseudolites to augment GPS. Pseudolites are basically beacons putting out GPS-like signals to augment the coverage provided by the satellite system.

 

Peace,

TeamRJJO

Yeah I agree, they must have WAAS disabled and LAAS enabled. Hopefully LAAS is more reliable than WAAS, eh?

"Use of GPS pseudolites now makes it possible to apply GPS navigation in an indoor environment. A pseudolite is a signal generator that transmits GPS-like signals to nearby users. In the United States, interest in pseudolite-based indoor navigation stems in part from imminent implementation of new automatic location capability in mobile cellular phones - so-called Enhanced 911 or E911 - required by a new Federal Communications Commission (FCC) rule.

With this motivation, in 1999 the Seoul National University GPS Lab (SNUGL) developed a centimeter-accuracy indoor navigation system using asynchronous pseudolites. The system has been upgraded in the last year to include carrier phase cycle-slip recovery and automatic cycle ambiguity-resolution functions. Using this system as a position and attitude sensor, SNUGL implemented a vehicle control system and obtained 1-2 centimeter control errors. These results demonstrated that, if pseudolites are used, GPS navigation is possible in indoor environments, such as a large factory, indoor amusement park, or anywhere GPS signals are blocked. We are convinced that our research results and use of this approach will help pioneer the extension of GPS navigation into the indoor world. This article describes the pseudolite-based positioning techniques and their application to track and control a miniature vehicle operating on an indoor track.

 

Conclusions and Future Work: SNUGL developed an indoor navigation system using a constellation of ceiling-mounted pseudolites. Experimental analysis showed that the horizontal point-positioning error was 0.14 centimeter (RMS) in the static case and 0.79 centimeter (RMS) in the dynamic case. For the purpose of using this system as a position and attitude sensor for vehicle control, we installed two antennas on the roof of a small test vehicle. Static error analysis showed that the horizontal positioning error was less than one centimeter (RMS) and that the attitude error was less than one degree (RMS). We also implemented cycle-slip recovery and automatic cycle-ambiguity resolution functions to make our system more robust and practical.

 

Indoor airports?

[Exfireguy]

Wow, some of you repliers are off my scope! I think that rather than worry about an error of 25 ft or so, I'll just do what I did a few days ago. While I was waiting for a lock on (which I didn't get in the canyon that I was in), my wife used the clue and found it. Nothing like hearing "Oh Honey, it's over here. Neener, neener!". Great discussion though. I did learn from it. Thanks.

[+Miragee]

You were lucky there was a clue . . . sometimes there isn't, or the clue is a "useless one." Check out the long-running thread about those. It makes for entertaining reading.

[+Team Neos]

The one thing that I can control that seems to make the most difference in the accuracy of my GPS is to have it turned on well before I want to try to find a cache--fifteen minutes before at a minimum, if not a half an hour.

 

Does that make it more accurate? I dunno, but I know that my unit shows a smaller degree of accuracy (usually about 15 feet) after it is on for at least a half an hour, compared to about 30 feet ten minutes after it is turned on.

 

When I locate caches, the coords showing on my unit are usually within 8-10 feet of the posted coords.

Edited August 1, 2005 by Team Neos

[+crake]

I'm amazed by the accuracy of my Meridian. However it needs to average for a few seconds, and I experience something like "look ahead" error - in that when I'm walking the GPS position seems to lag slightly. This is probably averaging (the reported location is behind me).

 

Now, sometimes I wonder how accurate the person who stashed the cache was. Folks are suppose to average on multiple days, right?

[+Miragee]

Nope. I wouldn't want to hike back up to my most recent series of five caches to take another reading on another HOT, HOT, HOT day. I took a couple of readings on the way up and on the way back as I placed the series. The coords have been right on for the first finders.

[+TeamRJJO]

The one thing that I can control that seems to make the most difference in the accuracy of my GPS is to have it turned on well before I want to try to find a cache--fifteen minutes before at a minimum, if not a half an hour.

 

Does that make it more accurate?

Probably a bit, at least. Having the receiver on gives the internal clock a chance to flywheel off the GPS signal for longer, reducing position error.

 

Peace,

TeamRJJO

[+Team Cotati]

Flywheeling of the ole GPSr are we? Hope that the diffilator doesn't break, that usually throws the whole thing off balance. Quite a mess really.