GPS off by 36 feet

[jjmcgo]

I don't want to hijack the Lost Franconia Hiker thread further but when I posted that my GPS is usually off by 36 feet, I got two responses.
It's a brand-new Lowrance i350c. I was wondering if the problem arose because I inputted my home location from my living room rather than the street in front of my house, approximately a distance of 36 feet.
I come up to an intersection, stop at the sign and it says I'm 36 feet from the crossing street.
I thought it was because these things had an accuracy of "plus or minus 30 feet" because I've seen people make references to being at the true summit of Owl's Head while their GPS says its 30 feet away.

 

[Pig Pen]

36 feet is close enough for use with nuclear weapons. Not close enough for semiconductor manufacture. A little marginal for steering big ships under bridges.

 

[Chip]

You'll get more comprehensive responses, but it doesn't have anything to do with where you started. There is a plus/minus with GPS that has to do with the unit and the number of Sats it's employing. I think +/- 15 feet is about standard, which is why it would be dangerous to trust your GPS on the Knife Edge (or some such) in a white-out.

 

[Neil]

My gps is accurate to one centimeter.

That is, when I download my tracklog onto the computer and bring up the map on the screen. The log is usually within 1cm (at a 1:24000 scale) of where I actually was.

Hope this helps.

 

[NewHampshire]

On my eTrex Legends satallite page the unit actually tells me the level of accuracy. Usually its anywheres from about 15 feet to the highest I ever saw at 38 feet (it was a very overcast day and I only had a lock on 3 satellites I believe.) While I am sure more expensive gets you better reception (and thus smaller margin of error) the fact is GPS units are not meant for pinpoint accuracy (+/- inches for example.) The accuracy you are getting should be more than acceptable for "reasonable" navigation.

Brian

 

[Pete S]

GPS Accuracy-

"Problems" with the position that a GPS gives--lat/long, grid,or "X" on a electronic map-- often stem from the Datum that the GPS is using being different from the Datum used to create the map being used. The Datum setting determines how the unit converts the raw GPS information into positions. Many GPS units are shipped set to the WGS 84 datum, a datum commonly used making modern maps. Many topographic maps are from earlier surveys, and use WGS 72, or even earlier datums, such as NAD 27. Making sure that your GPS unit is set to the same datum that the map you are using was created with will result in more accurate position fixes. Having the GPS unit on a different Datam than the map can easily result in positions on the "wrong" side of a stream, or show you walking in the lake instead of on the trail!

 

[kmorgan]

If your system supports it you should also turn on the WAAS option. Quote: This is the Wide Area Augmentation System. A satellite system that improves the location tracking accuracy of the GPS system from 20 meters to three meters. Deployed in 2003, WAAS comprises a group of earth stations and two satellites that cover North America. The earth stations track the GPS satellites and send correction signals to the WAAS satellites, which are picked up by WAAS-enabled navigation systems. End Quote.

On my Garmin units when WAAS is enabled the satellite #'s are 33 and above and signal bars for sat reception show a "D" when they are corrected.

More information here: (among other sites) http://en.wikipedia.org/wiki/Wide_Area_Augmentation_System

Kevin

 

[Chip]

My gps is accurate to one centimeter.

That is, when I download my tracklog onto the computer and bring up the map on the screen. The log is usually within 1cm (at a 1:24000 scale) of where I actually was.

Hope this helps.

Does that mean your GPS is accurate to .78 feet ? That'd be pretty darn good. What am I missing here ?

 

[sardog1]

Does that mean your GPS is accurate to .78 feet ? That'd be pretty darn good. What am I missing here ?

A deeper appreciation for Neil's wicked ways.

EDIT: OK, now that I've had my fun, it's time to get back on topic. jjmcgo, please tell us more about what "off" means in this instance. Are you referring to a discrepancy between coordinates displayed on your GPS receiver and a map position? If so, then a datum mismatch is a good bet. If not, I'm intrigued by your reference to "inputted my home location[.]" What specifically does that mean? Lowrance apparently has some "initialize from map" feature -- is this what's going on here?

 

[Kevin Rooney]

Hope this helps.

It does, Neil - clears everything right up. Thanks for your insight.

(I am pinching your head ...)

 

[Neil]

If my tracklog is off by 1cm at 1:24,000 then am I off by ¼ kilometer, 5/32 mile, 1.2 furlongs or 120 chains? Or is it the datum? I generally use the Hu-tzu-shan datum.

(Anyway, 30 feet sounds close enough to me.)

 

[jjmcgo]

"Here" means two things, Sardog

I can plan a route by inputting the address where I start and inputting the address where I'm going and that will give me a route.
Or, I can just input where I want to go and "start" is understood as "here," where ever I am.
I was sitting on my couch when I inputted our home address as opposed to sitting in the car in front of my house so maybe everything is off by 36 feet for that reason.
I was just joking on the other thread when I said my GPS would lead me into the water beside the pier instead of onto the pier because it's off by 36 feet. I think I'll put more effort and reading into the manual this weekend while I'm traveling.

 

[MTNRUNR]

not being a technical person...

Guess I'd wonder most about the Knife Edge statement.

When on a hike...say, going over the knife edge or on the pier and you set a waypoint there. On your return, being in a white out for instance, if you "goto" that waypoint...will it guide you back to that same spot? or?

I'm guessing it would depend on your satellite reception each time?

 

[Papa Bear]

not being a technical person...

Guess I'd wonder most about the Knife Edge statement.

When on a hike...say, going over the knife edge or on the pier and you set a waypoint there. On your return, being in a white out for instance, if you "goto" that waypoint...will it guide you back to that same spot? or?

I'm guessing it would depend on your satellite reception each time?

In principle you can retrace your route using a GPS, but the following points would apply to doing this on the knife edge:

1)You may lose signal altogether in bad weather.
2) the intrisic accuracy of the GPS is too gross to keep you on the exact edge

But ...

3) Unlike the top of Washington or Moosilauke (where you are going from cairn to cairn), the knife edge is well defined. In other words it is next to impossible to get lost there (I guess you could get turned around and go the wrong way).
4) The problems like icy rocks or snowed over ledges make it dangerous but they have nothing to do with route finding.

In other words, if you're crossing the knife edge watch your step, go slow and you will never need to even look at the GPS.

OTOH if you are on Washington or on the Tableland in a white out and have your GPS (and you are getting reception) you would have a useful tool. Sort of like map and compass in one little handheld. Until the batteries stop working ...

 

[Neil]

I believe you are talking about using the tracklog for backtracking but here's a similar function.

Using a gps in order to remain on an unmarked trail, say for example a snowed over herd path in open woods in the ADK's is one of the most practical uses of the instrument.

When you go off the trail which is easy to do by following the wrong opening, you check your screen and compare that to the map you printed out at home. The herd trail is printed right on that map because you "stole" it from the internet or a friend gave it to you or you recorded it yourself previously. You compare the shape of your track with what the map indicates it should be and bingo, you know whether to turn right or left to regain the trail.

You might think it simpler to just copy the tracklog right into the gps but I prefer the former method.
I'll have some waypoints for the herd trail in the gps but don't like having the entire log cluttering the screen. You have to reduce the tracklog from it's hundreds of trackpoints down to something more manageable. There are several ways to do that either in Topo!, Topofusion or GPSu.

You can extend the principle to going up or down a trail, getting caught in a whiteout and deciding/needing to continue on your way.

 

[DougPaul]

It's a brand-new Lowrance i350c. I was wondering if the problem arose because I inputted my home location from my living room rather than the street in front of my house, approximately a distance of 36 feet.

No.

I come up to an intersection, stop at the sign and it says I'm 36 feet from the crossing street.
I thought it was because these things had an accuracy of "plus or minus 30 feet" because I've seen people make references to being at the true summit of Owl's Head while their GPS says its 30 feet away.

There are two types of error (from the viewpoint of the user). Systematic and random.

A systematic error would always be the same--eg the indicated position might always be 350 ft NW of the actual location. Some reasons for GPS systematic errors:
* wrong datum (the datum is the reference for stating a location--ie where is 0,0 and which physical direction does a bearing of 000 deg point). The main datums used in the White Mtns are NAD27 and WGS84. They differ by ~220 meters in the Whites.
* bug in the internal software for doing the internal datum conversion.
* other bugs in the GPS.
* errors in your map. Consumer GPSes have an accuracy of about 10 meters (30 ft). (This means that the indicated position is within 10 meters 95% of the time in perfect signal conditions.) Many maps are off by more than 10 meters.
* you can only accurately say that a GPS is off by some distance only if you have an accurately surveyed location to compare to. Profession survey GPSes operated by experts can have errors less than 1 cm.

Random errors change from measurement to measurement. One way that random errors show up is that the location shown by a stationary GPS will wander around in a small zone. Typical sources are random errors in satellite clocks are satellite clock errors, satellite orbit errors, random changes in the ionosphere, and multipath (reflected) signals. Random errors account for most of the 10 meter figure and can sometimes be reduced by averaging several independent measurements.

More info on GPS errors: http://www.edu-observatory.org/gps/gps_accuracy.html

My guess is that you are seeing a map or a "new operator" error.

Doug

 

[DougPaul]

My gps is accurate to one centimeter.

No low cost GPS is that accurate.

That is, when I download my tracklog onto the computer and bring up the map on the screen. The log is usually within 1cm (at a 1:24000 scale) of where I actually was.

Accuracy is measured in real-world scales.

If I am allowed to choose the scale on which to measure my errors, I can find a set of coordinates that are within 1 cm (on my computer screen) of any point on the planet.

Doug

 

[DougPaul]

On my eTrex Legends satallite page the unit actually tells me the level of accuracy.

Those accuracy numbers* are estimates which miss some very important error sources. You can view them as a lower bound on the accuracy. (ie if it says 20 ft, it means that the correct number can be anything larger.) Rigorously speaking, it is meaningless.

* The typical accuracy number means that the actual error is this number or less 95% time. This also means the actual error is larger 5% of the time.

The only way to accurately determine the actual error is to place the GPS at a location that is known to a significantly greater than the accuracy of the GPS under test.

Doug

 

[DougPaul]

If your system supports it you should also turn on the WAAS option.

WAAS is an FAA system designed to improve accuracy for flying aircraft.

It can, under certain conditions, actually decrease accuracy. The current WAAS satellite constellation is currently being changed and reception on the ground is poor in the NE.

IMO, it is of little benefit to a hiker. 10 meter accuracy is quite adequate. I generally leave it turned off.

Doug

 

[DougPaul]

Does that mean your GPS is accurate to .78 feet ? That'd be pretty darn good. What am I missing here ?

That Neil is either playing a joke or he is far more confused than you.

See my earlier post responding to his post.

Doug