Observations On Accuracy

[+Robespierre]

Based on a few experiences with 2, 3, even 4 gps units in the group for both placing and hunting caches, I make these observations about accuracy:

A. Observation A is that, whatever the unit, it does not seem to vary more than about 12 feet from any other unit, given side-by-side operation. So I speculate that IF you put 10 different units of various age and origin on the same table on any given day, that whatever accuracy "from above" is available will be mostly uniform, and that the two readings of the two most diverse units will be within about 12'N-S, and comparable E-W/(taking into account that longitudes grow closer ).

B. Observation B is that (given that A is true...) every day you would do that, it would be mostly true, however the variation day-to-day would be much greater for the whole group, perhaps as much as 50 feet of difference between the two most diverse days. The variable factor would seem to be "above," and not the handheld technology.

 

Or stated slightly differently: The accuracy of the day/location depends on the day/location; the accuracy of the group of units in relation to each other is quite constant.

 

Am I way off base?

Clouds do NOT affect signal? Trees do? Sunspots do? Atmospheric conditions?

[+CO Admin]

Man with one GPS always knows where he is. Man with Two is never sure.

 

 

 

Clouds CAN block the signal.

[+tirediron]

Man with one GPS always knows where he is. Man with Two is never sure.

 

 

 

Clouds CAN block the signal.

...ehhhh?

[+Imajika]

Clouds, snow and rain info: link

 

From the Garmin site: GPS satellites transmit two low power radio signals, designated L1 and L2. Civilian GPS uses the L1 frequency of 1575.42 MHz in the UHF band. The signals travel by line of sight, meaning they will pass through clouds, glass and plastic but will not go through most solid objects such as buildings and mountains. link

 

I never noticed any signal loss during cloudy weather. Your mileage may vary.

[+sbukosky]

Clouds will indeed attenuate the signal. Think about water and a microwave oven. Why is a lock easier to get in the forest in the winter? No leaves. What is it about leaves? They contain water!

 

But to the topic, I find that the accuracy from day to day can exceed what the supposed error radius shows. I've frequently experimented with waypoint averaging using WAAS and getting down to a few feet accuracy. Come back a time later and one time it might be spot on and other times twenty feet off when the circle of accuracy shows to be eight feet.

 

I'd like to know if it is feasible to create a GPSR that can better the present handheld accuracy. Something getting closer to a surveyor's GPSR.

[Moun10Bike]

Clouds will indeed attenuate the signal. Think about water and a microwave oven. Why is a lock easier to get in the forest in the winter? No leaves. What is it about leaves? They contain water!

No, clouds will *NOT* attenuate signal to any degree discernible to a consumer-grade GPS receiver - see the link that Imajika posted for more info. The GPS signal frequency of about 1575mhz was chosen expressly because it is a "window" in the weather as far as signal propagation is concerned.

 

As you will learn from that link as well as others, leaves affect accuracy because they contain a coating of water, which is *NOT* the same as raindrops or clouds.

[+Renegade Knight]

Tree's are like people. Mostly water in solid form. A thin sheet of water would block a signal. A cloud won't. A leaf is a thin sheet of water. That's why canopy will kill your GPS reception.

 

My observations in using a Garmin alongside a Magellan are that they are usually within a few feet of each other.

[+CO Admin]

Clouds will indeed attenuate the signal. Think about water and a microwave oven. Why is a lock easier to get in the forest in the winter? No leaves. What is it about leaves? They contain water!

 

But to the topic, I find that the accuracy from day to day can exceed what the supposed error radius shows. I've frequently experimented with waypoint averaging using WAAS and getting down to a few feet accuracy. Come back a time later and one time it might be spot on and other times twenty feet off when the circle of accuracy shows to be eight feet.

 

I'd like to know if it is feasible to create a GPSR that can better the present handheld accuracy. Something getting closer to a surveyor's GPSR.

Yes, how much can you afford to pay?

 

 

 

And I will amend my original statement to say. Clouds can Attenuate the GPS signal.

[+Robespierre]

The links above by Imajika were very useful. However does anyone know what "atmospheric conditions" (since any normal rain is not supposed to stop signal, nor are any normal clouds) do account for a reduced accuracy. For example, I have seen days when out in the open my Vista will not promise more than a 53' accuracy. Normal-looking days, too. Usually it promises about 18'. Sometimes better. I can't always predict those good or bad days.

 

edit: I know time-of-day/satellite position is part of it. Hence the story of the cacher who made 5 trips after work to the same cache. When he finally went on a Saturday morning, he walked right to it.

Edited September 13, 2004 by Robespierre

[+Searching_ut]

I find the different observations regarding weather affecting GPS reception to be quite interesting. When I was in a USAF test squadron many years back, GPS guidance systems were one of the things we were playing with. While it wasn't part of the data we were collecting ourselves, I did get to look through data regarding atmospheric signal attenuation. As I recall, the variations from weather in the lower atmosphere were less than one percent. Nothing anyone could see in any manner from variations in their receivers performance. Upper atmosphere variations were considerably higher, enough so to be noticeable, but these aren't something you'll see by looking at the sky. As for trying to compare GPS signals to other radio frequency, different parts of the spectrum behave quite differently. What applies to one frequency doesn't necessarily apply to another...

[+SherwoodForest]

i know the GPS makers claim that rain and bad weather don't affect GPS signal....

 

but, everytime I have been out in the rain, the signal gets far worse (whether in the woods or in the clear)...

 

so i'll go with personal experience over manufacturer's claims anyday...

 

[+EraSeek]

The links above by Imajika were very useful. However does anyone know what "atmospheric conditions" (since any normal rain is not supposed to stop signal, nor are any normal clouds) do account for a reduced accuracy.

 

Here is a unit comparison: http://www.gpsinformation.net/waas/vista-waas.html

 

Yes, the ionosphere accounts for much of the error. A charged ionosphere will cause signal delay. A delay of one nanosecond equals one foot of postion error. There are other factors such as satellite geometry, clock error, multipathing and such as well.

 

No; clouds, rain, snow, sleet, fog, are miniscule factors. Layers of water on windshields, leaves, maybe a bigger factor.

 

I've used my gps in more than one whiteout. No problem.

[+EraSeek]

With good sky view and good sat geometry these are the errors you can expect and the errors that WAAS adresses:

 

Error Value

Ionosphere 4.0 meters

Clock 2.1 meters

Ephemeris 2.1 meters

Troposphere 0.7 meters

Receiver 0.5 meters

Multipath 1.0 meter

Total 10.4 meters

[+Robespierre]

Here is a unit comparison: http://www.gpsinformation.net/waas/vista-waas.html

That's a really good link, I'm just not sure that standing on a chair helped enough, even when I jumped. Gonna have to go back and spend more time.

[+EraSeek]

I think generally you are right. The bigger difference is in current reception conditions. But here is a comparison of several units in 2 areas. An open area, and a forested area.

 

Note that averaging helped very little in either case. External anntenea help only in the forested area.

 

http://www.fs.fed.us/database/gps/mtdc/gps.../Nav_3-2001.htm

 

This PDF file speaks more to the averaging and anntennea issue: http://www.fs.fed.us/database/gps/mtdc/map...smap_76_rev.pdf

Edited September 13, 2004 by EraSeek

[+Night Stalker]

I have cached with RK when he was using his GPS V and I was using my Magellan SporTrak Pro. I tseemd that consistently we right right on in the north south direction, but I was always 6 ft different (east) on the east west direction. This did not mean that one was more accurate than the other because sometimes I was closer to the cache and sometimes he was. As I remember there was only one time we actually agreed, and that time we were sure that something was wrong.

 

[+EraSeek]

I should have said averaging helps some in forested conditions, but not with an external anntenea.

[tlg]

i know the GPS makers claim that rain and bad weather don't affect GPS signal....

 

but, everytime I have been out in the rain, the signal gets far worse (whether in the woods or in the clear)...

 

so i'll go with personal experience over manufacturer's claims anyday...

 

I'm convinced too that my GPSr doesn't work as well in the rain but I believe that to be more my fault than the fault of the unit. I just don't feel like standing still trying to pick up some signals when it's pouring rain out and the rain is dripping down my sleeve and down my neck and down various other parts of the body. If it's sunny out, I don't mind standing in one spot picking up signals.

[+wvcoalcat]

I'd like to know if it is feasible to create a GPSR that can better the present handheld accuracy. Something getting closer to a surveyor's GPSR.

Actually, our survey grade receivers, when used independantly, are just about as accurate as any other receiver.

 

The accuracy is obtained when we use at least two at the same time. One is at a fixed, known point, and the other is moved to the other points needed. Then the observations are compared, resulting in the desired accuracy.

 

I know this is rather simplistic, but my co-workers keep popping in.

[+Robespierre]

I have cached with RK when he was using his GPS V and I was using my Magellan SporTrak Pro. I tseemd that consistently we right right on in the north south direction, but I was always 6 ft different (east) on the east west direction. This did not mean that one was more accurate than the other because sometimes I was closer to the cache and sometimes he was. As I remember there was only one time we actually agreed, and that time we were sure that something was wrong.

From my experience that's about what I would expect. I figure a circle with a 12' radius is about, what?, 452 square feet. That's not much to search in MOST places, so if the difference in the units is a bit less than that, great. However, it can easily be 4 or 5 times that given "atmospheric" conditions. I would also add that Garmin vs Garmin is pretty much identical to Garmin vs Maggie, in the limited experience I've seen with Maggie.

 

However, (and I'm enjoying these links) test results linked above make a greater distinction between some of the units by brand.

 

You've all seen those monthly "fishing calendars." Maybe eventually we'll get a caching calendar which marks days of really great, or really bad reception.

[+Bear_Left]

I've just been playing with what I'm calling DGPSoIP (Differential GPS over Internet Protocol), and is usually referred to as NTRIP (Networked Transport of RTCM via Internet Protocol)

 

I placed a couple of caches here in Christchurch, New Zealand last weekend, with a couple of spots that were under tree cover and giving me an EPE of 18-20m (convert that to mediaeval measurements yourself!) I hooked up the eTrex to an IBM laptop and logged onto my ISP via IRdA to a GSM phone (aren't these acronyms wonderful?), then watched as the satellite bars all started sprouting "D"s and the EPE went down to 3m!

 

In clearer areas, where I was getting the usual 'good' EPE of 5-6m (no WAAS down under...), I improved it to 1m.

 

It's too cumbersome for cache finding (but with a Palm client instead of the laptop...) but wonderful for setting caches, especially under cover.

[+CharlieP]

Or stated slightly differently:  The accuracy of the day/location depends on the day/location; the accuracy of the group of units in relation to each other is quite constant.

 

Am I way off base?

Clouds do NOT affect signal?  Trees do? Sunspots do?  Atmospheric conditions?

I think the accuracy of different units will be similar IF they are able to receive about the same number of satellites and there is no multipath error. A GPSR with a better antenna may be able to lock onto more sats than a unit with a less effective antenna, especially under trees, and this can make a big difference in accuracy. Of the variables that affect accuracy, I have found mutipath error (signal bounce) to be the most troublesome, because the GPSR may report an EPE which is flat wrong. It is frustrating to have the unit tell you you are within 10 feet of the position, with a 20 foot EPE, and then tell you you are 80 feet away, with the same EPE. When the GPS acts like this, I don't rely on the output. In places where signal bounce occurs, two GPS units side by side may differ by 60 or 80 feet, both reporting a 20 foot EPE.

 

FWIW,

CharlieP

[+Robespierre]

In clearer areas, where I was getting the usual 'good' EPE of 5-6m (no WAAS down under...), I improved it to 1m.

 

It's too cumbersome for cache finding (but with a Palm client instead of the laptop...) but wonderful for setting caches, especially under cover.

Now, this sounds really interesting. What I'm wondering is, how about when you do this at the same location every day for a week...I'm not all that technically astute, but I'm wondering about "measuring the atmospheric disturbance," getting more of a feel about the invisible variables.

 

How accurate is the BEST system out there - day to day. I presume it's a military system.

[+AtoZ]

I think you also ahe to be careful when comparing units to the antenna type, patch or helix, and the software used by the GPSr. Here is one exeracne I have had when helping some friends find a cache I have a Garmin GPSMAP76 with lastest software version Friends had a Garmin GPS 60C that uses a patch antenna. Now because I had intimate knowledge of the waypoint in question I put the coordinates into my GPSr which were the same as the one in the 60C and I there was a differance of about 20+ feet and I was closer to the waypoint then the 60C. So not sure what was going on. But was it the antenna or the software to interpert the recieved signals or what?

Cheers

[+niskibum]

I find the different observations regarding weather affecting GPS reception to be quite interesting. When I was in a USAF test squadron many years back, GPS guidance systems were one of the things we were playing with.

You didn't happen to be in the 57ags did you?

 

I for one hope that civilian gpsr's don't get any more accurate than they are right now. If they got down to centimeters this hobby would be a bore.

[+Searching_ut]

You didn't happen to be in the 57ags did you?

 

Close, it was the 431 TES, which fell under the 57FWW, even though we weren't stationed at Nellis.

[+Searching_ut]

Garmin GPS 60C that uses a patch antenna

 

Actually, the 60 C uses a quad helix antenna. As for 20 feet variation, depending on the terrrain etc that's pretty normal. I compared my two sportrak maps side by side on many occasions, and found they would disagree by more than that quite often, even though they were virtually identical hardware and firmware.

[+niskibum]

 

Close, it was the 431 TES, which fell under the 57FWW, even though we weren't stationed at Nellis.

Cool, I was in falcon amu and later viper ot&e at Nellis 87 thru 91. I have often wondered if I would ever meet any of the people I used to work with again. Was the 431 out of Mt. Home?

 

RE testing units side by side. I hear that they will interfere with each other if in close proximity. Just hearsay, so I don't know if it's true or not. From personal experience (legend and rino120) I have noticed that sometimes one performs better under some circumstances and the other shines in other circumstances. But both get me within 20ft pretty much every time, and thats close enough for caching.

[+Sputnik 57]

What I'm wondering is, how about when you do this at the same location every day for a week...I'm not all that technically astute, but I'm wondering about "measuring the atmospheric disturbance," getting more of a feel about the invisible variables.

 

How accurate is the BEST system out there - day to day.  I presume it's a military system.

Here is a recent study comparing accuracy by averaging. One test compares two GPSrs over 2 one week periods. Interesting graphs, including the plot of a 30 day test of a Garmin eMap that shows a cluster of results, most falling within a 1 meter circle, centered about 1.5 meters NE of the "true" position.

 

Edit: Trying (without much success) to make more intelligible

Edited September 16, 2004 by Sputnik 57

[+Robespierre]

Actually, the 60 C uses a quad helix antenna. As for 20 feet variation, depending on the terrrain etc that's pretty normal. I compared my two sportrak maps side by side on many occasions, and found they would disagree by more than that quite often, even though they were virtually identical hardware and firmware.

Two similar units operating side by side, and disagreeing by 20' or more....that would rather disprove my observations if others can verify.

[+Rakusan]

 

You've all seen those monthly "fishing calendars." Maybe eventually we'll get a caching calendar which marks days of really great, or really bad reception.

Aviation-grade GPS units do have that, or at least the top-of-the-line units do. It's called RAIM (Receiver Autonomous Integrity Monitoring); it's built into the box, and it determines whether there will be enough birds in the sky to navigate safely in bad weather. If you're flying in the slop, you need it.

[+EScout]

Now, this sounds really interesting. What I'm wondering is, how about when you do this at the same location every day for a week...I'm not all that technically astute, but I'm wondering about "measuring the atmospheric disturbance," getting more of a feel about the invisible variables

 

I recommend going to one of the super-accurate "adjusted" benchmarks. In the USA, we have them listed to the one-hundred thousandth of a second. I have tested my GPSrs, both a Legend and a Meridian, on various days, both sunny and rainny, WAAS on and off. It is interesting to see how "accurate" your GPSr is (how far from the known point) compared to what the EPE shows on your screen. You might want to note the number of satellites received when you read your distance.

 

With WAAS on and received while on the spot for 5 to 10 minutes, the accuracy is very good, sometimes surprisingly so. On rainny/cloudy days, it seems somewhat worse. By doing this you will get a real feel for your GPSr under different conditions of weather, satellites, etc.

 

Regarding averaging, you can do your own with a program called VisualGPS, which averages your GPSr plugged into a serial port. It has a graph and calculates error. It is freeware.

Edited November 17, 2004 by EScout

[+TrailGators]

Or stated slightly differently:  The accuracy of the day/location depends on the day/location; the accuracy of the group of units in relation to each other is quite constant.

 

Am I way off base?

Clouds do NOT affect signal?  Trees do? Sunspots do?  Atmospheric conditions?

I think the accuracy of different units will be similar IF they are able to receive about the same number of satellites and there is no multipath error. A GPSR with a better antenna may be able to lock onto more sats than a unit with a less effective antenna, especially under trees, and this can make a big difference in accuracy. Of the variables that affect accuracy, I have found mutipath error (signal bounce) to be the most troublesome, because the GPSR may report an EPE which is flat wrong. It is frustrating to have the unit tell you you are within 10 feet of the position, with a 20 foot EPE, and then tell you you are 80 feet away, with the same EPE. When the GPS acts like this, I don't rely on the output. In places where signal bounce occurs, two GPS units side by side may differ by 60 or 80 feet, both reporting a 20 foot EPE.

 

FWIW,

CharlieP

I'd like to know more about signal bounce. I typically notice large errors/eratic readings near roads or other large masses of concrete.

Edited November 17, 2004 by TrailGators

[dave and jaime]

a couple of points:

1)the epe is only a measure of error contined within the mathematicial computations of position. epe does not have any dependance on weather and only indirectly depends on sat geometry.

2)ground level weather supposedly does not affect, however disturbances in the ionosphere do affect the gps.

3)if testing units side-by-side i doubt there would be any direct rf interference due to the levels of rf noise we are dealing with.

4)as for difference in testing units side-by-side, as was said earlier the clocks in gps (quartz based) are measuring nano seconds. minor varitions, data signifant figures, and quality would account for much more than the 20-50' of 'error' we see.

6)all esitmates of location ffrom a gps are based on the ellipsoid model that the current datum is based on, this model is only an approximation of the earths surface and this modeling also introduces errors into the calculation.

 

nearly everyone here is using the gps for point position using the ca code, if results are within 20' of the known value the system has done its job. if greater acccuracy is desired differential position is necessary or using the p code.

[+AtoZ]

Clouds will indeed attenuate the signal. Think about water and a microwave oven. Why is a lock easier to get in the forest in the winter? No leaves. What is it about leaves? They contain water!

 

This statement is TOTALLY irrelevant. Microwaves are at such a wavelenght that they are purposely designed to change the rotational speed of a water molecule and thus change the kinetic energy of the water molecule and thus increase the thermal energy. The signal used by GPS DOES NOT DO THIS. Though I have or felt I have seen the accuracy of my GPSr decrease when cloud cover rolls in, this maybe subjective. Now all the stuff that is published maybe done so to snowball the DOE and any other alphabet organization into funding the GPS system. I have read all the literature and am not sure as I have not seen any imperical evidence. Ol well that is a thought but microwaves and GPS are totally differant wavelengths and do NOT interact with the same material in the same way.

cheers

[CoyoteRed]

I compared my two sportrak maps side by side on many occasions, and found they would disagree by more than that quite often, even though they were virtually identical hardware and firmware.

This quite an accurate statement, believe it or not. The wife and I carry STs, one a Map and the other is a Pro. We have differences of well in excess of 50' at times.

 

It's kind of funny that we would be going along and all of a sudden she would veer off while mine said I had another 100' to go!

 

I'm thinking it is a combination of the internal averaging of the units and the differences of the signals they are getting at any one time. The signals will be different because of the electronic shadows cast by trees and our bodies. The differences add up to the units thinking they are in different spots.

 

Knowing this, you can compensate. I generally stick my hiking stick in the ground and hang my unit from is as I hunt--all the while it is getting closer and closer to the actual electronic position the hider's GPS told him.

 

Speaking of clouds attenuating a satellite signal, they do. Period. I know, because my job in the Navy was satellite communications. We always operated at the lowest power to communicate without more than a few correctable errors a second. When it was overcast, we had to bump the power to reduce the errors. The amount of attenuation was dependent on the saturation of the cloud. A thundercloud will attenuate heavily.

 

Given this, I've even experienced problems with heavy cloud cover with an ST. Be in a borderline situation and then add heavy cloud cover--you will see the effects of clouds.

 

BTW, a leather GPS holder that covers the antenna looks cool and works great--until you get it wet. My holder wet is absolutely useless trying to use the unit, I have to remove it to get signal.

[CoyoteRed]

Ol well that is a thought but microwaves and GPS are totally differant wavelengths and do NOT interact with the same material in the same way.

Partially true statement.

 

Radio waves of differing lengths will effect materials to differing degrees. It doesn't have to be UHF frequencies to be affected by water vapor.