serpa4
02-13-2004, 04:10 PM
Is the accuracy listed in the GPS status a diameter or radius?
Click to See Complete Forum and Search --> : Accuracy, radius or diameter?
apersson850
02-13-2004, 04:24 PM
First, it's a radius.
Second, although the exact calculation is something Garmin considers to be proprietary information, it's probably close to something known as CEP, Circular Error Probable.
Which, if we leave the mathematics behind, means that there is about a 50% probability that you are within that circle.
Double the distance, and you get something near 90%.
Triple it, and you are above 95%.
I haven't checked these figures right now, they are right out of my head. But since the exact figures for the iQue are a secret anyway, it doesn't matter if the numbers are wrong to some extent.
You get the general idea, at least. The most imortant thing to remember is that the probability that you are anywhere isn't really 100%. Ask the cat, which lurks around here, if you don't believe me.
Still, chances are good that you are within the given distance from the given position. Just remeber that it's a probability, not a fact.
Second, although the exact calculation is something Garmin considers to be proprietary information, it's probably close to something known as CEP, Circular Error Probable.
Which, if we leave the mathematics behind, means that there is about a 50% probability that you are within that circle.
Double the distance, and you get something near 90%.
Triple it, and you are above 95%.
I haven't checked these figures right now, they are right out of my head. But since the exact figures for the iQue are a secret anyway, it doesn't matter if the numbers are wrong to some extent.
You get the general idea, at least. The most imortant thing to remember is that the probability that you are anywhere isn't really 100%. Ask the cat, which lurks around here, if you don't believe me.
Still, chances are good that you are within the given distance from the given position. Just remeber that it's a probability, not a fact.
serpa4
02-13-2004, 06:35 PM
Heres a question. For geocaching, are my chances of finding what ever the greatest in the center of the circle or equal within the circle?
What I'm asking is if you have a large accuracy circle, am I most likely in the middle, inside the triangle, but could be anywhere inside the circle?
Second question, with different take on it.
If I paint a 20 foot circle on the ground around me that represents the accuracy circle, put my gps on the center dot, then look for an object within the circle, as I get further away from the center dot (gps), still within the circle, I'm less likely to find what I'm looking for as I move away from the center, but there is a chance its anywhere in the circle.
Sorry for any confussion, its hard to explain what I'm looking for.
Are my chances of finding something equal anywhere in the circle or better at the center? Or on a gradent, my chance of finding something within the circle get better the closer I am to the center.
What I'm asking is if you have a large accuracy circle, am I most likely in the middle, inside the triangle, but could be anywhere inside the circle?
Second question, with different take on it.
If I paint a 20 foot circle on the ground around me that represents the accuracy circle, put my gps on the center dot, then look for an object within the circle, as I get further away from the center dot (gps), still within the circle, I'm less likely to find what I'm looking for as I move away from the center, but there is a chance its anywhere in the circle.
Sorry for any confussion, its hard to explain what I'm looking for.
Are my chances of finding something equal anywhere in the circle or better at the center? Or on a gradent, my chance of finding something within the circle get better the closer I am to the center.
doodah
02-13-2004, 08:01 PM
I think your asking if your odds decrease as you leave the center of your circle. I don't really know, but I also don't think it matters when you consider that the person who planted the cache used a reading that was probably no more accurate than yours. If he was off by 20 feet, your odds might be better at the edges of your circle. Who knows?
Now if your hunting benchmarks, that's another matter. Those positions are pretty accurate and my experience is that if you turn on WAAS, your going to find it pretty close to the center of your circle.
Now if your hunting benchmarks, that's another matter. Those positions are pretty accurate and my experience is that if you turn on WAAS, your going to find it pretty close to the center of your circle.
apersson850
02-14-2004, 04:53 AM
The closer to the center the better.
From a mathematical point of view, you can be fooled, since the probability that the correct position is at the true center of the circle is 0%. But that's just because a point has no area, so it's nothing. But think of the probability as a cone, that flattens out to the side. Place the cone at the indicated position. The probability that you are at the correct position is proportional to the height of the cone. Still some probability rather far out, but it decreases quickly.
This is by far not correct, from a mathematical point of view, but I'm not going to get you trhough that in this forum. There's plenty of that on Wilson's pages.
Second thing:
Many people believe, that if I place a cache, when my EPE is 10 meters, and you look for it, when your EPE is 10 meters, then your real EPE, when looking for my point, is 20 meters.
It doesn't add up like that. You EPE in this case is really 14 meters. Thats sqrt(epe1^2+epe2^2).
Good old Pythagoras.
From a mathematical point of view, you can be fooled, since the probability that the correct position is at the true center of the circle is 0%. But that's just because a point has no area, so it's nothing. But think of the probability as a cone, that flattens out to the side. Place the cone at the indicated position. The probability that you are at the correct position is proportional to the height of the cone. Still some probability rather far out, but it decreases quickly.
This is by far not correct, from a mathematical point of view, but I'm not going to get you trhough that in this forum. There's plenty of that on Wilson's pages.
Second thing:
Many people believe, that if I place a cache, when my EPE is 10 meters, and you look for it, when your EPE is 10 meters, then your real EPE, when looking for my point, is 20 meters.
It doesn't add up like that. You EPE in this case is really 14 meters. Thats sqrt(epe1^2+epe2^2).
Good old Pythagoras.
tusker
02-14-2004, 06:06 AM
For anyone who wants more information on CEP, Sigma etc, this link (http://www.mercat.com/QUEST/Accuracy.htm)will probably give them all they need to know, and more besides.
apersson850
02-14-2004, 06:10 AM
Yes, and if you get really hooked, you can continue here (http://users.erols.com/dlwilson/gps.htm). For hours.
navbear
02-14-2004, 12:23 PM
I like the image of the probability cone that Anders used.
But from a practical point of view, we should not forget that the behaviour of GPS (and SBAS) is not fully modeled by probability approaches.
First, the effects that disturb GPS measurements are not really randomly distributed, but are affected by distinct influences of the atmosphere and the local environment to the antenna (among others). The typical statistical models attempt to model this by "bounding" the probability of a certain error magnitude.
This means that when you return to a certain point at the same time of day, when the positions of the GPS sats are roughly identical, you have a good chance that a large part of the errors is the same.
(Mathematically speaking, there is a correlation of errors.)
This means that you have a much better chance of finding the cache when you know the time of day the original measurement was taken.
But from a practical point of view, we should not forget that the behaviour of GPS (and SBAS) is not fully modeled by probability approaches.
First, the effects that disturb GPS measurements are not really randomly distributed, but are affected by distinct influences of the atmosphere and the local environment to the antenna (among others). The typical statistical models attempt to model this by "bounding" the probability of a certain error magnitude.
This means that when you return to a certain point at the same time of day, when the positions of the GPS sats are roughly identical, you have a good chance that a large part of the errors is the same.
(Mathematically speaking, there is a correlation of errors.)
This means that you have a much better chance of finding the cache when you know the time of day the original measurement was taken.
apersson850
02-14-2004, 02:20 PM
Not exactly. Even if the sats orbit the earth twice a day, the earth moves in the opposite direction as well. Which means that they "meet" each other, and the effective orbit time is really about 11 h 56 min. So after a while, the satellite constellation at 11 o'clock is not what it was at 11 o'clock, say, two months ago.
But otherwise you are correct. However, the mathematical theory required to cover the whole thing, is a bit too much to go through in a forum like this. That's why i try to simplfy as much as possible, without completely loosing all connection with the reality. After all, I may not even have understood it completely myself. :D
But otherwise you are correct. However, the mathematical theory required to cover the whole thing, is a bit too much to go through in a forum like this. That's why i try to simplfy as much as possible, without completely loosing all connection with the reality. After all, I may not even have understood it completely myself. :D
jonasolof
02-14-2004, 05:51 PM
You'd have to wait for a time when sat constellation is the same as it was at the irst instance + that it should be at the same time of the day for ionospheric characteristics to be similar (hopefully)
apersson850
02-14-2004, 06:35 PM
If you want to involve that too, then there is also the even more random tropospheric influence on the signals.
navbear
02-15-2004, 09:58 AM
The deeper we get into the details of GPS/SBAS positioning, the more boring this may get for the "general audience". There are other places to discuss this stuff...
Anyway, I can't resist to add a few items of errors effective for the end user:
a sidereal day, after which the GPS constellation is repeated, is of course 23h56m04s long. If however sats come or go, this is another issue...
ionosphere is the largets error source since SA went away, and SBAS helps to cope with it.
troposphere is generally modeled quite well. Because is the same for all sats, the error levels out somewhat and moves to the altitude, Unless you are experiencing a rainstorm or something alike, it should not be too bad.
Are geocachers that hardcore? :D
Anyway, I can't resist to add a few items of errors effective for the end user:
a sidereal day, after which the GPS constellation is repeated, is of course 23h56m04s long. If however sats come or go, this is another issue...
ionosphere is the largets error source since SA went away, and SBAS helps to cope with it.
troposphere is generally modeled quite well. Because is the same for all sats, the error levels out somewhat and moves to the altitude, Unless you are experiencing a rainstorm or something alike, it should not be too bad.
Are geocachers that hardcore? :D
apersson850
02-15-2004, 10:05 AM
To some extent, I agree that this may become too detailed for the "general user", whoever that might be.
But still, the iQue is (so far) unique in that aspect, that it's a PDA with an integrated GPS. As a consequence, I assume that a somewhat larger interest in the GPS part of the device ought to be present among the owners, compared to if GPS was just an accessory, as it is for most other PDA units.
I think some detailed information, with links to the enormous amount of knowledge available on the net, are appropriate.
By the way, you can't imaging how hardcore some geocachers are! :p
But still, the iQue is (so far) unique in that aspect, that it's a PDA with an integrated GPS. As a consequence, I assume that a somewhat larger interest in the GPS part of the device ought to be present among the owners, compared to if GPS was just an accessory, as it is for most other PDA units.
I think some detailed information, with links to the enormous amount of knowledge available on the net, are appropriate.
By the way, you can't imaging how hardcore some geocachers are! :p
Curious Cat
02-15-2004, 10:11 AM
Personally what I’ve found to be the biggest source of inaccuracy for me is user error. :D For those of you who are familiar with Georgia, USA I once routed myself all the way to Blairsville only to find that I needed to go to Barnesville which was hundreds of miles in the opposite direction.
f4phantomii
02-15-2004, 02:57 PM
As far as hunting for objects goes, you can do that with fair success.
My neighbor is the official County Surveyor. He frequently goes out looking for USGS markers that haven't seen the light of day for over 50 years. And references on how to locate them aren't there anymore (tree fell, road moved, etc.).
One day he asked me to go with him to a couple of local markers and bring my Garmin 12CX with the differential. We were able to locate two out of four markers using the GPS.
Keep in mind these markers are usually a half meter deep in the ground, so they aren't just laying about for you to see. One of the two we found, we hit it dead center with the probe the first time.
My neighbor now has a much more expensive surveyor type GPS. He says it saves him a ton of time since he doesn't have to dig holes all over the place looking for markers. Most of the time he can find them after three or four holes using the GPS to narrow down the search area to a reasonable size.
-Michael
My neighbor is the official County Surveyor. He frequently goes out looking for USGS markers that haven't seen the light of day for over 50 years. And references on how to locate them aren't there anymore (tree fell, road moved, etc.).
One day he asked me to go with him to a couple of local markers and bring my Garmin 12CX with the differential. We were able to locate two out of four markers using the GPS.
Keep in mind these markers are usually a half meter deep in the ground, so they aren't just laying about for you to see. One of the two we found, we hit it dead center with the probe the first time.
My neighbor now has a much more expensive surveyor type GPS. He says it saves him a ton of time since he doesn't have to dig holes all over the place looking for markers. Most of the time he can find them after three or four holes using the GPS to narrow down the search area to a reasonable size.
-Michael
apersson850
02-15-2004, 03:08 PM
One of the more basic navigational skill, expected to be mastered by a cat, is to know the difference between nose and tail.:D
My wife once drove to Linköping, then realized that Lidköping would have been better. Subtle difference, though.
My wife once drove to Linköping, then realized that Lidköping would have been better. Subtle difference, though.
jonasolof
02-15-2004, 03:08 PM
Which raises the question: Why couldn't he use the same method to find them that was used to put them there in the first place years ago if the location was so exact?
About technical level:
I think it is good to know that sat view repeats itself with the sideral day (23h56 min). This increases understanding that you can't compare results taken with totally different satellite constellations and geometry.
I also think it is good for users to know that the largest error is introduced by ionospheric difraction+ of course multipath. It helps to have an intuitive understanding of what radio signals do and go through.
About technical level:
I think it is good to know that sat view repeats itself with the sideral day (23h56 min). This increases understanding that you can't compare results taken with totally different satellite constellations and geometry.
I also think it is good for users to know that the largest error is introduced by ionospheric difraction+ of course multipath. It helps to have an intuitive understanding of what radio signals do and go through.
f4phantomii
02-15-2004, 07:42 PM
Sorry....I didn't make it clear why the USGS markers are so hard to find.
Many of them were located by reference to objects that are no longer present. The county keeps a record of the last time anyone located any particular marker. Usually it is a surveyor, and they try to update how to find it based on objects that are current.
For example, one of the ones we found had never been used or located since it had first been put in the ground in 195X. The references to it were something like....
220 degrees, 200 feet 8 inches from large oak tree.
90 degrees, 18 feet 0 inches from side of road.
27 degrees from NW corner of barn, 75 feet 2 inches.
Now following those, it should be easy to triangulate and get a pretty good search area. If it was found recently, say in the last 10 years, it is usually pretty easy to find.
In this instance, the road had been widened twice, the oak tree was gone, and the barn had burned down in 1962 and an entire neighborhood now stood in its place.
Without GPS, you would have to take references from the two or three nearest other known markers (if you can find them), and they are usually about 1/2 mile or more apart.
-Michael
Many of them were located by reference to objects that are no longer present. The county keeps a record of the last time anyone located any particular marker. Usually it is a surveyor, and they try to update how to find it based on objects that are current.
For example, one of the ones we found had never been used or located since it had first been put in the ground in 195X. The references to it were something like....
220 degrees, 200 feet 8 inches from large oak tree.
90 degrees, 18 feet 0 inches from side of road.
27 degrees from NW corner of barn, 75 feet 2 inches.
Now following those, it should be easy to triangulate and get a pretty good search area. If it was found recently, say in the last 10 years, it is usually pretty easy to find.
In this instance, the road had been widened twice, the oak tree was gone, and the barn had burned down in 1962 and an entire neighborhood now stood in its place.
Without GPS, you would have to take references from the two or three nearest other known markers (if you can find them), and they are usually about 1/2 mile or more apart.
-Michael
doodah
02-15-2004, 09:51 PM
If you're interested in hunting benchmarks (USDA markers), check this link:
http://www.geocaching.com/mark/
It's fun, educational, and often damned challenging. Also has excellent benchmarking forum with very knowledgeable participants.
http://www.geocaching.com/mark/
It's fun, educational, and often damned challenging. Also has excellent benchmarking forum with very knowledgeable participants.
Curious Cat
02-15-2004, 10:05 PM
Originally posted by apersson850
One of the more basic navigational skill, expected to be mastered by a cat, is to know the difference between nose and tail.:D
Yes I believe one smells and the other end um, well it smells too. :D
One of the more basic navigational skill, expected to be mastered by a cat, is to know the difference between nose and tail.:D
Yes I believe one smells and the other end um, well it smells too. :D
Robertp366
02-15-2004, 10:12 PM
The USGS medallions are placed with a full up survey team, doing precision triangulation with transits, steel chains (now laser DME), etc. They can be relocated the same way, but it is quite a project to rebuild a marker from the neighbors. It has been done - earthquakes tend to mess things up here in California at least on the time scale of decades. Secondary markers, which are most often the ones used for property surveying, are developed off the USGS medallions and often reverrified when major subdivisions are done. (No big developer wants to face hundreds of property boundry lawsuits because a county marker was out of place).
FWIW, the first major mapping triangulation of the US, establishing the oldest reference markers was done by a 21 year old surveyor named George Washington. His grid extended from the Potomic tol the little town of Washington Virginia, a place with a post office, four or five buildings, and a pre-Revolutionary War Inn that has the finest restaurant in America (Michelin 3 star, one of only two in the US).
Oh, and to my pilot friend, f4phantomii: white scarfs are OK, but a well placced s/w trojan horse in the right blackbox did a fine job taking out an entire Group of top-of-the-line fighters sold to the Shah and left behind in Iran when Komeni arrived..... And nobody had to die....
:p
Robert
FWIW, the first major mapping triangulation of the US, establishing the oldest reference markers was done by a 21 year old surveyor named George Washington. His grid extended from the Potomic tol the little town of Washington Virginia, a place with a post office, four or five buildings, and a pre-Revolutionary War Inn that has the finest restaurant in America (Michelin 3 star, one of only two in the US).
Oh, and to my pilot friend, f4phantomii: white scarfs are OK, but a well placced s/w trojan horse in the right blackbox did a fine job taking out an entire Group of top-of-the-line fighters sold to the Shah and left behind in Iran when Komeni arrived..... And nobody had to die....
:p
Robert
Curious Cat
02-15-2004, 10:35 PM
I’m really not trying to be a wise guy here but just out of curiosity-What happens legally if a large track of land, say 10 square miles, is shifted a couple hundred feet due to an earthquake, who owns the part of the land that used to yours but is now within your neighbors original border per GPS in California? If the boundaries don’t move with the land, couldn’t it be possible that one could lose ownership of their home?
ps Now you know how I got my name.:D
ps Now you know how I got my name.:D
Robertp366
02-15-2004, 11:01 PM
If the earthquake shifted 10 sq miles of land by 100 feet, ther would be no lawyers - or most anybody else either - living within at least 500, maybe 1000 miles. That'd be something like a magnitude 32 earthquake!!!!!
The SF quake at 7.6 moved some things a few feet but only where the soil was soft and liquified. Loma Prieta had an 18 inch vertical fault shift at the surface. Northridge was more like 10 inches at the surface.......
100 feet. FEET!!!!!
The geological record of the Yellowstone fault shows a distrubance of some 30 feet. It caused a sheet eruption some 100 million years ago that spread lava ash over 4000 miles away and essentially denuded the NA continental plate, filling the Tetheys sea.
100 FEET!!!!!!!!!!!!!!!!!!!!!!!
For the few inches of actual effects here in LA, and similarly in Loma Prieta, common law grandfathered the existing fences as 'moved'. But anything that needed permits for rebuilding or new construction had to conform to the 'true' property lines. USGS resurveyed the master markers, and the counties reverified their secondary system markers. Very few things were actually affected.
Some shakey tirvia: The largest earthquake in recorded history in the US occurred where? and moved what major river? it occurred in 18xx When??
And how many earthquakes over magnitude 6 have occurred in California? Massacheusetts? Missouri?
Robert
The SF quake at 7.6 moved some things a few feet but only where the soil was soft and liquified. Loma Prieta had an 18 inch vertical fault shift at the surface. Northridge was more like 10 inches at the surface.......
100 feet. FEET!!!!!
The geological record of the Yellowstone fault shows a distrubance of some 30 feet. It caused a sheet eruption some 100 million years ago that spread lava ash over 4000 miles away and essentially denuded the NA continental plate, filling the Tetheys sea.
100 FEET!!!!!!!!!!!!!!!!!!!!!!!
For the few inches of actual effects here in LA, and similarly in Loma Prieta, common law grandfathered the existing fences as 'moved'. But anything that needed permits for rebuilding or new construction had to conform to the 'true' property lines. USGS resurveyed the master markers, and the counties reverified their secondary system markers. Very few things were actually affected.
Some shakey tirvia: The largest earthquake in recorded history in the US occurred where? and moved what major river? it occurred in 18xx When??
And how many earthquakes over magnitude 6 have occurred in California? Massacheusetts? Missouri?
Robert
Curious Cat
02-15-2004, 11:19 PM
Did I notice a hint of enjoyment in how you wiped out all the lawyers first? :D
You said 18XX but I thought the largest recorded earthquake in the United States was a magnitude 9.2 that struck Prince William Sound, Alaska, March 28, 1964.
You said 18XX but I thought the largest recorded earthquake in the United States was a magnitude 9.2 that struck Prince William Sound, Alaska, March 28, 1964.
Robertp366
02-15-2004, 11:39 PM
Shakespear started it. King lear, Act II...."first we kill all the lawyers....."
Now you're gonna make me look it up (Like you did, didn't you?). I was refferring to the 'lower' US - specifically the 7.9 at New Madrid, Missouri, Feb 12, 1812. Two others associated with this were 7.7 Dec 16, 1811, and 7.6 on Jan 23, 1812. Caused the Missouri river to shift substantially. BTW there is a lot of Common Law about what happens to property that is affected by rivers that move their banks....
And there were several 6.x quakes in Boston area in the 17th century, and another in Manhattan Island, NY. Just to note that CA is not the shakiest part of the US.;)
Robert
Now you're gonna make me look it up (Like you did, didn't you?). I was refferring to the 'lower' US - specifically the 7.9 at New Madrid, Missouri, Feb 12, 1812. Two others associated with this were 7.7 Dec 16, 1811, and 7.6 on Jan 23, 1812. Caused the Missouri river to shift substantially. BTW there is a lot of Common Law about what happens to property that is affected by rivers that move their banks....
And there were several 6.x quakes in Boston area in the 17th century, and another in Manhattan Island, NY. Just to note that CA is not the shakiest part of the US.;)
Robert
Curious Cat
02-15-2004, 11:51 PM
Actually it was Henry VI part 2. Henry probably wasn't thinking about earthquakes.....
Robertp366
02-15-2004, 11:55 PM
Thanks. I've had that quote wrong in my notes for years.
Robert
Robert
Robertp366
02-15-2004, 11:59 PM
Did the cat just start blinking, or am I having another observer effect? (It wasn't there until I first saw it...)
Nice avitar. Seems like all the best comunity members have one. Once again I'm late and left out..... ;)
Robert
Nice avitar. Seems like all the best comunity members have one. Once again I'm late and left out..... ;)
Robert
Curious Cat
02-16-2004, 12:09 AM
Originally posted by Robertp366
Did the cat just start blinking, or am I having another observer effect? (It wasn't there until I first saw it...)
Nice avitar. Seems like all the best comunity members have one. Once again I'm late and left out..... ;)
Robert
Here you go. You’re going to have to shrink it yourself though.
Did the cat just start blinking, or am I having another observer effect? (It wasn't there until I first saw it...)
Nice avitar. Seems like all the best comunity members have one. Once again I'm late and left out..... ;)
Robert
Here you go. You’re going to have to shrink it yourself though.
doodah
02-16-2004, 08:52 AM
Originally posted by Robertp366
Now you're gonna make me look it up (Like you did, didn't you?). I was refferring to the 'lower' US - specifically the 7.9 at New Madrid, Missouri, Feb 12, 1812. Two others associated with this were 7.7 Dec 16, 1811, and 7.6 on Jan 23, 1812. Caused the Missouri river to shift substantially. I don't think so. That may have been a big earthquake, but New Madrid, MO isn't within a 100 miles of the Missouri River. That'd be one hell of a shift. :eek: I believe the river involved was the Mississippi.
Now you're gonna make me look it up (Like you did, didn't you?). I was refferring to the 'lower' US - specifically the 7.9 at New Madrid, Missouri, Feb 12, 1812. Two others associated with this were 7.7 Dec 16, 1811, and 7.6 on Jan 23, 1812. Caused the Missouri river to shift substantially. I don't think so. That may have been a big earthquake, but New Madrid, MO isn't within a 100 miles of the Missouri River. That'd be one hell of a shift. :eek: I believe the river involved was the Mississippi.
Robertp366
02-16-2004, 10:57 AM
Yep. Once again I'm missing the mark.
Robert
Robert
PDA Street
Copyright Internet.com Inc. All Rights Reserved.
Copyright Internet.com Inc. All Rights Reserved.