Jump to content

A simple method to improve position accuracy


tick54

Recommended Posts

Using a Garmin 12XL with readings of an EPE of 20 feet or less, against known GPS points, I find a maximum error of a single reading to be 30 feet. To increase accuracy, I average the readings with a simple method which may be different than what others use when they “average” a reading. My method is to record a range of readings and simply average the highest and lowest reading obtained, regardless of the most common reading encountered.

 

My reasoning goes like this. A single reading tells me that the actual location is somewhere within a 30 foot radius circle where I am standing. Other readings at the same location will have the same meaning. As the range of the readings increase, the error of my true location decreases.

 

gpserror.jpg

 

In practice, I normally get a range of 5 units (one unit = .001 minutes) or 30 feet of latitude within 5-10 minutes. The average between the high and low readings will then be reliably within 15 feet of the true location. If I get a range of 10 units or 60 feet of latitude, my error will be zero.

 

This may take more time than most geocachers care to spend, but if higher accuracy is desired, it is a method that can be used. After testing and practice, I find that very accurate readings can be obtained with off the shelf, inexpensive, basic units. I am eagerly awaiting the arrival of an etrex-H to do some more testing.

Link to comment

I guess that would be assuming the float was to the opposite side. It could have been off and then further off in any direction.

 

I think there is a larger circle that encompasses that limit of the accuracy at that location and if you took enough readings you would simply fill that circle. Taking just two, or three, or four readings wouldn't be more accurate.

 

I haven't seen much benefit to averaging. Personally I just take a few readings and see which one gets me closest to my cache. That one was the accurate reading when I took it.

 

edited for clarification

Edited by BlueDeuce
Link to comment

There are certainly cachers that are more qualified to discuss your theory than I am, but I'm not sure that your assumption that your'true location' must be within 30 feet when your GPSr has an epeof 30 feet. Since EPE means 'estimated positioning error', one would have to assume that while you have some confidence that your actual location is within 30 feet of those coordinates, you cannot be completely sure that your actual location is within 30 feet of those coords.

Link to comment

Although the OP isn't 100% correct, this does give a good example of why it is always better to average several readings instead of taking just one reading or taking the reading with the best EPE. With just a few readings you can get to very small area in which the true location is most likely to be, essentally improving your EPE far beyond what you can get from any one reading.

Link to comment

My method is to use my gps' builtin averaging capability by setting the unit on the cache location for a few minutes. After I get the coords, I walk off a couple hundred feet in several directions and if I am directed back to the cache, those are the coords I use. If not, I repeat the process.

 

But I think I will add in the lunch step. I can see how that could be very beneficial to the process. :P

Link to comment

I think any averaging, unless your unit automatically does it for you, is a waste of time.

 

You take all that time, for what? So your coords are 6 feet off instead of 12? Big deal. 6 feet is not going

to make any difference to the searcher.

 

I averaged for my first 10 hides. Haven't for the next 190+. There was no little difference

in accuracy between the caches I averaged and those I didn't. In fact one of the few caches I own that

consistently had complaints about coordinates was one where I averaged.

 

If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

 

If you have a lousy signal you're just averaging bad data.

Edited by briansnat
Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

I believe the settle part was settled.

Link to comment

I think any averaging, unless your unit automatically does it for you, is a waste of time.

 

You take all that time, for what? So your coords are 6 feet off instead of 12? Big deal. 6 feet is not going

to make any difference to the searcher.

 

I averaged for my first 10 hides. Haven't for the next 190+. There was no little difference

in accuracy between the caches I averaged and those I didn't. In fact one of the few caches I own that

consistently had complaints about coordinates was one where I averaged.

 

If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

 

With 104 hides, I take one waypoint reading, and it has worked fine for me.

Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

I believe the settle part was settled.

Takes more than a minute or two for either of mine to settle. :P

Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

I believe the settle part was settled.

Takes more than a minute or two for either of mine to settle. :P

Buy a Garmin or bring a larger lunch. :D

Link to comment
I think any averaging, unless your unit automatically does it for you, is a waste of time.

 

You take all that time, for what? So your coords are 6 feet off instead of 12? Big deal. 6 feet is not going

to make any difference to the searcher.

 

I averaged for my first 10 hides. Haven't for the next 190+. There was no little difference

in accuracy between the caches I averaged and those I didn't. In fact one of the few caches I own that

consistently had complaints about coordinates was one where I averaged.

 

If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

I agree. I think good signal, settling and fresh batteries is the most important factors. My average reading seldom ever changes whenever I've done those there things.
Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

I believe the settle part was settled.
Takes more than a minute or two for either of mine to settle. :P
Buy a Garmin or bring a larger lunch. :D

A new Garmin is on the list!

Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

I believe the settle part was settled.
Takes more than a minute or two for either of mine to settle. :P
Buy a Garmin or bring a larger lunch. :P

A new Garmin is on the list!

Not a new Triton with National Geographic topos? :D
Link to comment
If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

Depends on the unit. You'd get wild readings with a SporTrak doing that. You need to let it settle.

 

I set the unit over the cache and let it settle for a few minutes. Then break the averaging and put it back. Wait 15 to 30 minutes depending of the cover and satellite configuration, then mark. I've achieved .001 repeatability with this method.

I believe the settle part was settled.
Takes more than a minute or two for either of mine to settle. :P
Buy a Garmin or bring a larger lunch. :P
A new Garmin is on the list!
Not a new Triton with National Geographic topos? :D

Touch screens? I'm leery enough with a $300 PDA in my pack. I need something tough.

 

But, who knows. By the time the disposable cash becomes available my mind may have changed.

Link to comment
My reasoning goes like this. A single reading tells me that the actual location is somewhere within a 30 foot radius circle where I am standing.
Not so. It's telling you there's a 50% chance that the actual location is within that circle. Details here. The intersection scheme, therefore, will not give better information about the location than just seeking the zero-distance point.
Link to comment

The most accurate method is to take a reading under the best conditions. Your mark would still be subject to all those other conditions that might, and usually do occur later.

 

Sorry if that's seems obvious, but when using the 'hold a standard gps and take a reading' method there's no math trick available other than making the effort for getting a good mark.

Link to comment

Using a Garmin 12XL with readings of an EPE of 20 feet or less, against known GPS points, I find a maximum error of a single reading to be 30 feet. To increase accuracy, I average the readings with a simple method which may be different than what others use when they “average” a reading. My method is to record a range of readings and simply average the highest and lowest reading obtained, regardless of the most common reading encountered.

 

My reasoning goes like this. A single reading tells me that the actual location is somewhere within a 30 foot radius circle where I am standing. Other readings at the same location will have the same meaning. As the range of the readings increase, the error of my true location decreases.

 

In practice, I normally get a range of 5 units (one unit = .001 minutes) or 30 feet of latitude within 5-10 minutes. The average between the high and low readings will then be reliably within 15 feet of the true location. If I get a range of 10 units or 60 feet of latitude, my error will be zero.

 

This may take more time than most geocachers care to spend, but if higher accuracy is desired, it is a method that can be used. After testing and practice, I find that very accurate readings can be obtained with off the shelf, inexpensive, basic units. I am eagerly awaiting the arrival of an etrex-H to do some more testing.

 

Regardless of your reasoning, what you're doing is averaging together the two data points that are most likely to be wrong. Averaging bad data doesn't make it better.

Link to comment

Does anyone have a list of the units which have this automatic averaging built in? I have an eXplorist 400, but don't seem to have that function. How long does it take to do this?

I don't have a list, but my Lowrance will average 100 readings in about 2 minutes. I have several logs stating that the coords were dead on.

Link to comment

Most of the time I'll get a very poor signal at the exact spot where I'm hiding a cache (due to tree cover). If I can find a spot thats clear either 6 or 12 feet away from the spot, and that I'm sure is exactly North, South, East, or West, I'll move to that spot and then get a reading and then add/subtract it from the final. Six feet away would affect the coords by .001 and 12 feet is .002 on average. I'll then walk away about 100 feet and turn around and approach the area to make sure that I did it right. It's far from a perfect method but it does help to improve the accuracy a bit.

Link to comment

I think any averaging, unless your unit automatically does it for you, is a waste of time.

 

You take all that time, for what? So your coords are 6 feet off instead of 12? Big deal. 6 feet is not going

to make any difference to the searcher.

 

I averaged for my first 10 hides. Haven't for the next 190+. There was no little difference

in accuracy between the caches I averaged and those I didn't. In fact one of the few caches I own that

consistently had complaints about coordinates was one where I averaged.

 

If you have a good signal, let your GPS settle for a minute or two, mark your coords and you're good to go.

 

With 104 hides, I take one waypoint reading, and it has worked fine for me.

 

You used the same waypoint reading for all 104 caches????? wow, thats amazing. :lol:

 

 

 

 

 

 

 

 

 

(sorry I just had to, Ill go back to my corner now)

Link to comment

When searching a hide out I don't let my Gps settle in. I just walk up to the coords and start looking. When I hide one I use the same strategy. I found two caches the owners mentioned letting their GPSrs settle in for 20 minutes or so and found their coords off on one 75 feet and the other 120 feet. Both have changed their listing coords to match the loggers coords. Don't know if there is a real answer to this but have to agree with the posters that think it is just a waste of time to worry about it too much. My 76CSX gets me close enough to the ones I'm looking for and finders close enough to my hides to find them. :o

Link to comment

"The Garmin 76CSx has terrific reception!"

 

Good reception may mean better accuracy but not always.

High sensitivity receivers sometimes pick up poor satellite information that the less sensitive receivers will filter out.

 

That being said, the high sensitive receivers are better overall and an improvement over the older models

Link to comment

Using a Garmin 12XL with readings of an EPE of 20 feet or less, against known GPS points, I find a maximum error of a single reading to be 30 feet. To increase accuracy, I average the readings with a simple method which may be different than what others use when they “average” a reading. My method is to record a range of readings and simply average the highest and lowest reading obtained, regardless of the most common reading encountered.

 

My reasoning goes like this. A single reading tells me that the actual location is somewhere within a 30 foot radius circle where I am standing. Other readings at the same location will have the same meaning. As the range of the readings increase, the error of my true location decreases.

 

gpserror.jpg

 

In practice, I normally get a range of 5 units (one unit = .001 minutes) or 30 feet of latitude within 5-10 minutes. The average between the high and low readings will then be reliably within 15 feet of the true location. If I get a range of 10 units or 60 feet of latitude, my error will be zero.

 

This may take more time than most geocachers care to spend, but if higher accuracy is desired, it is a method that can be used. After testing and practice, I find that very accurate readings can be obtained with off the shelf, inexpensive, basic units. I am eagerly awaiting the arrival of an etrex-H to do some more testing.

Regrettably, it's not so simple. Think of the EPE on your GPSr as the Circular Error Probable (CEP) of your missile dropping out of the sky onto your GPSr. There's a 50/50 chance that it's going to be inside the circle or outside the circle. You can mitigate the chance of the GPSr not being destroyed by making the warhead in the missile large enough for the destructive blast radius to incorporate the CEP.

 

Thus, regrettably, the chances of a location residing within the intersection of the two circles you indicated are actually pretty tiny. In reality, you would be unlikely to get such a small intersection. Where GPSr reception is poor, giving rise to widely spaced readings, or apparent position drift, you are dealing with larger circles, not smaller intersections.

 

You can think of your apparent position as a classical bell curve of distribution points. If you have a great unobstructed signal from a great many satellites (on a boat in the sea, for example) you've got a very steep sided distribution curve and your EPE will be small. On the other hand, if you are in the woods or an urban canyon, most probably you'll be within range of a limited number of satellites within a limited area of the sky (typically overhead - you can see their distribution on your GPS satellite location plot) and you'll have a very flat distribution curve and your EPE will be large. If you are in the woods, the steepness of the distribution curve will be influenced by the extent of the foliage and will thus vary seasonally.... It does not take much to figure that the best time of year to place caches will be when the foliage is at its minimum!

 

Clearly, the number and positions of the GPS satellites at any time also has a bearing on the accuracy of a given reading. There's no point in taking many readings over a short period of time as your measurements will all be based on the same subset of satellites in more or less the same place in the sky. If you want to compare multiple readings, go back at different times, hours or days apart.

 

It's often easier to obtain good readings in open space away from a cache. Take your compass with you and get yourself due east (or west) and north (or south) to get readings with a low EPE for use as the latitude and longitude respectively. Don't rely on your GPSr compass feature: regrettably I have found the top named brand batteries I use to be magnetic and therefore have distorting influence on the GPSr.

 

Lastly, check your cache coordinates in Google Earth. This has super accuracy in areas with high definition photography and people will not curse you as much when they go looking for your urban nano.

 

When I adopted and relocated my cache "Bartlow Hills" I could not get a decent set of coordinates, full stop. Thus, I described the "get yourself due west" technique in the cache listing and suggested people pace their way to the trailhead and then pace their way back to the cache. This works a treat compared to struggling with the GPSr reception in the woods.

 

When seeking caches, just appreciate the quirkiness of your GPSr and remember it's a part of the challenge :o

 

Cheers, Sandvika

Link to comment

Too complicated for me. I like to take 3-5 readings from where I intend to hide a cache, then shut down the unit and walk away about 100 feet. Repower the unit, then see which readings get me back closest. Compare the best return trips, and compare. Make a choice from that.

 

There will always be a little varience between different units, different days, different circumstances. This is part of the fun, the hunt!

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...