GPS accuracy test

[+BulldogBlitz]

Not sure the nail is necisary. Point one would act the same. I think that it would be a good exersise just to help you visualize the limitations of your GPSr. So far, I have been using benchmarks to do a similar test. I have been recording the accuracy at over 50 benchmarks. Sometimes it is way off (10m), but usualy 3m. Even 3m gives you a GZ of 6m (3m hider + 3m finder = 6m).

 

 

accuracy...consistency....precision... etc.

 

reading through this thread, it would appear that there is some false assumptions by some posters that once a cache has been reviewed, those coordinates are "correct". the nail is just another point... but it serves a purpose, the starting point.

 

doing this at an event, would put everyone participating relative to that first stab in the dark. i like that cheyenne GC, too bad it is so far from me.

[Andronicus]

Not sure the nail is necisary. Point one would act the same. I think that it would be a good exersise just to help you visualize the limitations of your GPSr. So far, I have been using benchmarks to do a similar test. I have been recording the accuracy at over 50 benchmarks. Sometimes it is way off (10m), but usualy 3m. Even 3m gives you a GZ of 6m (3m hider + 3m finder = 6m).

Just re-thinking my math. I think that would be a 6m radius which of couse results in a 12m diamiter (6m in any direction). Then, if your GPS is acting funny (and you may or may not know), or the hider's GPS sucks, that could easly be opend up to a 13 or even 20m radius (26-40m diamiter).

 

All that said, once I have verified my GPSr's zero point, rarly is the cache more than 5m away.

[+BulldogBlitz]

All that said, once I have verified my GPSr's zero point, rarly is the cache more than 5m away.

 

grr....now i gotta find my manual to even know if i can check this "zero point"....

 

[+sbell111]

You guys do that I will go out and find some caches. Seems like a waste of time to me.

 

Scubasonic

Buh-bye.

 

Wouldn't want to waste your time, since I see you're busy reading and posting such helpful stuff here.

 

When you get a minute between caches fix the spelling error in your sig line...

 

"This is just to fun !!!"

You sure do get your panties wadded easily.

I appreciate your concern about my underwear. You've always taken a particular interest in me, it seems, but I did not realize it was quite that personal. I'm really not interested in sharing the state, color or design of my undergarments with you, but I'm flattered by your interest.

You are mistaken in your belief that I take any particular interest in you. Many of your positions demand comment. That's as far as it goes.

 

Regarding your frillies, I similarly have no interest. I merely commented on that issue because you seem to be attacking people with little provocation.

 

While you no doubt believe that all of your ideas are winners, the rest of us are free to point out those that are stinkers. Put on your big boy shorts and accept that everyone isn't required to flatter you every time that you touch your keyboard.

[+Knight2000]

When you get a minute between caches fix the spelling error in your sig line...

 

"This is just to fun !!!"

That is kind of funny.

[+TripCyclone]

What y'all are ignoring is that if you know where true GZ is, be it a pin or a set of coordinates that you recorded, then you will likely, consciously or not, try to affect the test. You are less likely to place a marker where the GPS tells you is GZ when you know that the real GZ is over there a ways.

TAR, I think your experiment is well thought-out and would be quite useful to give people a sense of how their GPS devices really work. And I am a Real Scientist, even. I am really amazed at the responses you have gotten. Apparently science is not taught very well in the schools any more.

 

At any rate, I think that your original experiment (with the nail) is quite good.

 

Fizzy, I'm guessing you'll be happy to know that as a science teacher, I use this very exercise in my own 8th grade classroom. And we teach it as both consistency and accuracy. I mark three locations. One is a single measured waypoint out in the open. One is a single measured waypoint in the nearby woods. And one is an averaged waypoint in the woods. Students (grouped together due to limited amount of GPSr's), then navigate as close as possible to where their GPSr says I measured the point, and plant a flag. We talk about how averaging can improve the accuracy of the measurement, and how depending on factors such as brush, satellite positions (time of day of measured vs revisited), cloudy vs clear weather, different units, etc, can all affect the accuracy of the signal. Then we talk about how all of the devices seem to consistently veer towards one direction from the original spot during each class. Flags are left up over the course of the day and by the end, each spot has around 30 flags set up.

[+bunkerdave]

You might find this virtual cache to be of interest:

 

GC3997

[+NYPaddleCacher]

You might find this virtual cache to be of interest:

 

GC3997

 

This earth cache is kind of interesting as well: GC13FZ7

 

It's located on a high sand dune in Nags Head, NC. Finders are asked to post a waypoint of the "highest spot on the dune". Since the dune is always changing the coordinates vary widely. I thought it was a really interesting idea for an earthcache and it was just a very cool place to be with a great view.

[GOF and Bacall]

I sure am glad these things have such a variable nature. How much fun would it be if we all had consistent accuracy to within an inch? "5 ft...4Ft...3ft...2ft...1ft...HUCKLE BUCKLE!"

[+gpsblake]

In place of the nail, you can use an adjusted benchmark, if it's convenient. You'll just have to pretend not to know where it is.

 

Yup yup, that's the true accuracy test. Your handheld should be within 6 feet of an adjusted benchmark.

 

Now to these people who claim their consumer GPS is always within a few inches of accuracy. That's impossible. I'll explain.

 

The reason is, the XX.xx.xxx format isn't that accurate to measure within inches.

 

Example:

 

N 40 00.000 W 80 00.000 is about 6 feet from N 40 00.001 W 80.00.001... So 6 feet is the most you can really be accurate to on a consumer grade GPS and that's providing your are getting perfect signal.

 

Accuracy is no longer the issue with GPS units, any GPS unit will get you to the cache site today. But not a single GPS will HELP you find that tricky hide or help with your hunting skills.

 

So if you want to really increase your finds, work on your finding skills, and stop blaming your GPS unit.

[oakenwood]

For those interested in locating benchmarks, here's a good place to start. Remember, adjusted benchmarks are the accurate ones.

[CosmicTruth]

As a surveyor I think this is interesting thread so I wanted to add my 2 cents...

GPS is not accurate and varies through the day as the GPS Constellation changes so does the PDOP (Position Dilution of Precision) and HDOP (Horizontal Dilution of Precision), I study the Constellation to find when is the best time to get the position I want (either horizontal or vertical) even when using dual frequency equipment and Differential GPS techniques for GPS Surveying. So this test may not be good enough, we may want to get a position each hour for a month or so to get the idea of how accurate GPS is. But I can tell you with 1 unit it’s off by less than 30' most of the time and is usually off by about 12 feet, and occasionally is within 3 feet, but is never perfect. check out the scatter charts at http://pnt.gov/public/sa/diagram.shtml

also I should add that ocasionally there are not enough sats to even get a position.

 

Cosmic

Edited May 2, 2010 by CosmicTruth

[+TheAlabamaRambler]

The week has passed and no one posted any results, so I suppose this test wasn't as interesting as I thought.

[+Ed56]

The week has passed and no one posted any results, so I suppose this test wasn't as interesting as I thought.

Your test is quite interesting but since I don't have any yard at all I did my test a different way. Several years ago I hiked up a local 6512-foot peak where there is a USGS first order survey medallion or "Brass Cap". I placed my GPS unit on top of the marker and let the unit average its' readings for a while then saved its' reading. Later at home I compared my reading to the UTM coordinates published on the USGS site. According to the site my units' reading was off about 9-feet in longitude and 7-feet in latitude. Of course a typical geocache will not be located in such a nearly ideal spot. I probably should go back and take additional reading for comparison but it's a long drive and a long hike although I have taken reading at a lower order marker with reasonably good results. Happy trails.

[+loudy13]

so is the nail camo'ed does it meet caching guidelines, haha

 

cool idea for the week long test im gonna try it, "what else am i gonna do stay here and learn"

 

CACHE ON

[+22shoeless]

so is the nail camo'ed does it meet caching guidelines, haha

 

cool idea for the week long test im gonna try it, "what else am i gonna do stay here and learn"

 

CACHE ON

I am about 2 weeks into my geocaching career and very thankful I stumbled upon this thread. I honest to gawd believed until now that if my GPSr said I was within 1 meter of the cache, then I was. Doh! Game changing revelation.

 

Thanks!

[+TheAlabamaRambler]

so is the nail camo'ed does it meet caching guidelines, haha

 

cool idea for the week long test im gonna try it, "what else am i gonna do stay here and learn"

 

CACHE ON

I am about 2 weeks into my geocaching career and very thankful I stumbled upon this thread. I honest to gawd believed until now that if my GPSr said I was within 1 meter of the cache, then I was. Doh! Game changing revelation.

 

Thanks!

Welcome to the game obsession!

 

Every GPS and user is different, but I will offer a suggestion that may help now that you know that all GPS receivers are liars!

 

When I first approach GZ (Ground Zero, the place where my GPS believes the cache to be located) I try to do so in as direct a line as I can. Once the GPS gets below 20' I pretty much put it in my pocket and look around.

 

If I don't find it right off I don't chase the GPS in circles at GZ, I walk off a ways (50 or 60'), note the direction and distance the GPS reads and look in that direction and at that distance for some sort of landmark. Then I make a 90° turn and walk 50-60' and do the same thing. The point where the two sight lines intersect is very likely to be where the cache is. In a fuzzy sort of way that's called 'triangulation' and is very helpful in urban environments where the cache is close to a wall or wilderness caches with tree cover or in a canyon where GPS signals tend to bounce.

 

Good luck, have fun!

[4wheelin_fool]

so is the nail camo'ed does it meet caching guidelines, haha

 

cool idea for the week long test im gonna try it, "what else am i gonna do stay here and learn"

 

CACHE ON

I am about 2 weeks into my geocaching career and very thankful I stumbled upon this thread. I honest to gawd believed until now that if my GPSr said I was within 1 meter of the cache, then I was. Doh! Game changing revelation.

 

Thanks!

 

Yes, this is a good thread because that's often a common mistake among new cachers. I usually display the accuracy in one of the fields on the GPS when I'm searching, but even that is not completely accurate.

 

Another test is to take 2 or more GPS units, turn them off and drive to a predetermined set of coords. Then turn them on at the same time and watch to see which one zeros in quicker.

[knowschad]

The week has passed and no one posted any results, so I suppose this test wasn't as interesting as I thought.

Well, at one point, I thought that I was interested, but then SBell111 came along and pointed out to me that I really wasn't.

[+smithdk]

I taught a GPS class yesterday and decided to try this with the people attending the class. I hid the nail in the ground the previous day with my Vista HcX. It was an open grassy area with wooded hills on either side about 100-150 feet away. The students were instructed to go to the waypoint and put in a flag at GZ. Each of them used an Etrex Legend H. I used a tape measure to determine the distance of the flags from the nail. 2 flags were less than 4 feet from the nail. 1 flag was a little over 6 feet from the nail and a group of 4 flags were about 8 feet from the nail.

[+TheAlabamaRambler]

I taught a GPS class yesterday and decided to try this with the people attending the class. I hid the nail in the ground the previous day with my Vista HcX. It was an open grassy area with wooded hills on either side about 100-150 feet away. The students were instructed to go to the waypoint and put in a flag at GZ. Each of them used an Etrex Legend H. I used a tape measure to determine the distance of the flags from the nail. 2 flags were less than 4 feet from the nail. 1 flag was a little over 6 feet from the nail and a group of 4 flags were about 8 feet from the nail.

Excellent! Thank you for reporting your results!

[+drfred]

I appreciate the issue you are raising, TAR, and think that the game is improved by people understanding the limitations of the technology.

 

Let me throw in one more term, just to confuse the picture even more: precision

 

People have mentioned "consistency" but I think they really mean "precision". This has to do with the amount of error in any particular reading by your device and the repeatability of multiple measurements.

 

Once you have determined if your particular unit is precise, then you can try to decide whether your unit is accurate as well.

 

From years of reading reviews of different units and their strengths and weaknesses, a few points have emerged. In terms of importance IMHO:

1) User interface is probably more important than the actual precision and accuracy of modern devices

2) The algorithm used for averaging the fluctuating position information (how stable is your reading)

3) As far as technical differences, the chipset used plays an important role in the time to satellite acquisition and maintenance of a steady reading

 

Everyone seems to focus on #3 above, but again, IMHO, that is only #3

[Narnian Rockhound]

I agree this is an worthy topic - users who understand the limits of GPS are better contributors to the game as a whole. And though I can appreciate the debate of accuracy/consistency/precision, let's not lose sight of the larger issue and get bogged down in technicalities. Most folks will understand this issue as "accuracy", even though that's technically incorrect.

 

There seems, in my experience, to be a need to understand this especially from the point of view of those who hide a cache. Some folks seem to take a quick snapshot for coords, displaying an attitude that getting the best coords. possible isn't very important. I would prefer to see clearer guidelines on the site for how people should go about measuring the positions for their hides.

 

Several people have mentioned some critical steps to take, but unless I've missed it, there are a few important points that haven't really been well flushed out.

1) The number of satellites you are hearing is obviously critical, but also important is whether you are getting a WAAS signal in addition to the regular birds. WAAS improves your position calculations tremendously, but it can take some time for the signal to be received and used to correct the position, sometimes up to 15 minutes after turning the unit on. And since the WAAS satellites are in fixed positions in the southern sky for us N. Am. folks, knowing this can help a user make better/more informed choices when marking positions.

2) A few folks have mentioned within-unit averaging, which is also really important - to let the unit sit there at the location for ~5 minutes is a tremendous benefit when marking a point, as most units will continuously average over time their location calculation if the unit is not moving.

 

Compare, for example, the difference between firing up a unit, getting 4-6 birds/no WAAS in 2 minutes and grabbing a point quickly, to letting the unit warm up, get all 12 channels cranking w/ WAAS corrections, and let the unit sit and average for 5 minutes after that. There is a tremendous difference in the "correctness" of these two ways of doing it.

 

3) Taking several measurements over the course of a single day or several days is also really useful as the satellites move across the sky. The question becomes, though, how does the ordinary user go about "averaging" several lat./long. sets of coords? Most folks probably wouldn't know how to approach this very well - but, if you follow good procedure and get ~4 really good measurements, they probably won't be too terribly different and a visible pattern would emerge. Plotting the results in googlemaps or similar might help people see how their measurements are displaced from one another.

 

Other corrections, like differential GPS or post-processing, can also improve positions tremendously down to the sub-meter range, but these are not going to be accessible to common GPS users.

[+22shoeless]

I agree this is an worthy topic - users who understand the limits of GPS are better contributors to the game as a whole. And though I can appreciate the debate of accuracy/consistency/precision, let's not lose sight of the larger issue and get bogged down in technicalities. Most folks will understand this issue as "accuracy", even though that's technically incorrect.

 

There seems, in my experience, to be a need to understand this especially from the point of view of those who hide a cache. Some folks seem to take a quick snapshot for coords, displaying an attitude that getting the best coords. possible isn't very important. I would prefer to see clearer guidelines on the site for how people should go about measuring the positions for their hides.

 

Several people have mentioned some critical steps to take, but unless I've missed it, there are a few important points that haven't really been well flushed out.

1) The number of satellites you are hearing is obviously critical, but also important is whether you are getting a WAAS signal in addition to the regular birds. WAAS improves your position calculations tremendously, but it can take some time for the signal to be received and used to correct the position, sometimes up to 15 minutes after turning the unit on. And since the WAAS satellites are in fixed positions in the southern sky for us N. Am. folks, knowing this can help a user make better/more informed choices when marking positions.

2) A few folks have mentioned within-unit averaging, which is also really important - to let the unit sit there at the location for ~5 minutes is a tremendous benefit when marking a point, as most units will continuously average over time their location calculation if the unit is not moving.

 

Compare, for example, the difference between firing up a unit, getting 4-6 birds/no WAAS in 2 minutes and grabbing a point quickly, to letting the unit warm up, get all 12 channels cranking w/ WAAS corrections, and let the unit sit and average for 5 minutes after that. There is a tremendous difference in the "correctness" of these two ways of doing it.

 

3) Taking several measurements over the course of a single day or several days is also really useful as the satellites move across the sky. The question becomes, though, how does the ordinary user go about "averaging" several lat./long. sets of coords? Most folks probably wouldn't know how to approach this very well - but, if you follow good procedure and get ~4 really good measurements, they probably won't be too terribly different and a visible pattern would emerge. Plotting the results in googlemaps or similar might help people see how their measurements are displaced from one another.

 

Other corrections, like differential GPS or post-processing, can also improve positions tremendously down to the sub-meter range, but these are not going to be accessible to common GPS users.

 

Narnian - a nice simple set of guidelines for capturing the position of your new cache - much appreciated. I now understand why my chum was 15 meters off down a cliff on my first experimental cache. Ugh!

[+fizzymagic]

There seems, in my experience, to be a need to understand this especially from the point of view of those who hide a cache. Some folks seem to take a quick snapshot for coords, displaying an attitude that getting the best coords. possible isn't very important.

You're so right. This attitude is reinforced by people who don't understand the problem who say "well, if the seeker's receiver isn't accurate then the hider's coordinates don't need to be, either."

 

The asymmetry of the situation is not well appreciated. The fact is, if the hider does not get good coordinates then it doesn't matter how good the seeker's GPS is, since the coordinates it is seeking will not be very close to the actual cache location. That is why it is so important for hiders to get good coordinates.

[+WRITE SHOP ROBERT]

Incorrect.

 

The nail allows the simulation of geocaching, where finders attempt to locate something hidden by a person who then took coordinates with a GPS. It needs to be actually, not "sort-of" hidden in order for the experiment to work properly.

But the experiment wasn't to simulate geocaching. It was to measure GPS accuracy. You're re-writing the objective to fit the process.

 

The nail was an attempt to do a double-blind test. But unless the nail is known to actually be at the target coordinates (which would require survey-grade equipment) it doesn't serve its purpose.

This is why having a clearly marked spot that is KNOWN to ba as accurate as we can get, and then recording the relative spots where your GPS says "Zero" would be a more accurate scientific test. See the link I posted in my first post.

[+WRITE SHOP ROBERT]

 

doing this at an event, would put everyone participating relative to that first stab in the dark. i like that cheyenne GC, too bad it is so far from me.

I'll be looking for another similar benchmark in my area and setting up some sort of puzzle using that as the reference point. There are similar BMs all over with the same degree of accuracy.