Is Owls Head really a 4K?

[hippotone]

When I hiked Owls Head earlier in the year and arrived at the summit, the 3 GPS devices (all Garmin) that my group had all showed the elevation at about a hundred feet below what it should have been. I know that GPS units aren't always the most accurate with elevation but thought it was odd for all 3 to be off like this when they had been very accurate on other peaks earlier in the day. This was disappointing to see for some of us and raised the question whether Owls Head is a true 4K or not. I'm sure it is, so have others had similar experiences on Owls Head or on any other mountains where multiple devices show an incorrect elevation that was more than just a few feet off?

 

[bikehikeskifish]

Is it possible all three devices did not make it to the true summit? I have found there to be some discrepancy with published distances / elevation, distances of course being fractals and there not being any one correct distance, while elevation is usually pretty darn close as long as a I calibrate the GPS before leaving the trailhead. I do have a barometric altimeter in mine.

Every program I've ever plugged a track into has put the summit at over 4000 feet as well.

Tim

 

[hippotone]

I almost put in my description that it was the new summit, haha. When I looked into this a bit at first I did read up about the old and new summit. I would have been pretty bummed if I had to go out and do it again!

 

[bikehikeskifish]

4032.8' according to my most-recent Garmin track ... officially still at 4025'

Tim

 

[DougPaul]

When I hiked Owls Head earlier in the year and arrived at the summit, the 3 GPS devices (all Garmin) that my group had all showed the elevation at about a hundred feet below what it should have been. I know that GPS units aren't always the most accurate with elevation but thought it was odd for all 3 to be off like this when they had been very accurate on other peaks earlier in the day. This was disappointing to see for some of us and raised the question whether Owls Head is a true 4K or not. I'm sure it is, so have others had similar experiences on Owls Head or on any other mountains where multiple devices show an incorrect elevation that was more than just a few feet off?

You don't say which models you were using. The models and their calibration modes fall into 3 categories:
* GPS altitude (no barometric altitude sensor): The nominal accuracy for the GPS altitude is ~95% probability of being within 20 meters (66 ft). The errors are a combination of 2 types: systematic (eg satellite orbit or clock errors, variations in the ionosphere, poor satellite constellation, etc) that will affect all nearby units similarly, and random (eg noise in the receiver, differences due to the antennas being in slightly different locations, differences in the immediate surround, etc) that will affect the units independently. The systematic errors generally dominate so 3 GPSes giving essentially the same altitude doesn't mean that it is accurate. (It really only says that the GPS units are operating properly.)
* Barometric altitude (Garmin sensor units) with manual calibration: The key question is how is the barometric altitude calibrated and what has happened since calibration. If you calibrate all to the same "known altitude" at the same time and the known altitude is inaccurate, all will exhibit similar errors. Or if the barometric pressure is changing rapidly, then the error of all 3 will drift similarly and the more time, the more error.
* Barometric altitude (Garmin sensor units) with automatic calibration. (Automatic calibration is done by automatic comparison to the GPS altitude over an extended continuous period just prior to the measurement, preferably several hours or longer.) I haven't seen any official accuracy specs for this mode.

Garmin specs the barometric altitude accuracy (for the 60CSx, others are probably similar) as "accuracy, +/- 10 feet", subject to proper user calibration. I suspect that automatic calibration is probably best for most hikers and if you turn the GPS on at the trail head or earlier (eg for the drive to the trailhead) the accuracy should be as good as it will get by the time you reach the summit. This is what I do and I usually find the summit altitude to be within 10 feet of the published altitude. (Don't forget to correct for how high the GPS is above ground level.)

Doug

 

[Remix]

To add to Doug's point, garmin connect can be set up to ignore ALL the altitude data recorded by the device, and substitute the altitude corresponding to the position recorded by the device.

 

[DougPaul]

To add to Doug's point, garmin connect can be set up to ignore ALL the altitude data recorded by the device, and substitute the altitude corresponding to the position recorded by the device.

While the OP didn't say specifically, I suspect the measurements were taken at the summit where Garmin Connect would not be in play. But if one is examining the track on a computer after returning home, the program can show whatever it wants. (I don't use Garmin Connect so I have no info on what it actually does.)

As an addition to my earlier post, vertical accuracy for case 1 (GPS altitude) is highly dependent upon the satellite constellation--if the constellation was bad at the time of measurement, the error could have been much larger. (The +/- 20 meter number is a long-term average accuracy...)

Also, the barometric altimeter depends on an atmospheric model (ie assumed temperature and humidity profiles). If the current profile is different from the model (plus the effects of wind...), it will introduce additional errors. It also assumes no horizontal change. Thus greater the altitude and position differences between the calibration point and the measurement point the greater these errors are likely to be.

Doug

 

[skiguy]

While the OP didn't say specifically, I suspect the measurements were taken at the summit where Garmin Connect would not be in play. But if one is examining the track on a computer after returning home, the program can show whatever it wants. (I don't use Garmin Connect so I have no info on what it actually does.)

As an addition to my earlier post, vertical accuracy for case 1 (GPS altitude) is highly dependent upon the satellite constellation--if the constellation was bad at the time of measurement, the error could have been much larger. (The +/- 20 meter number is a long-term average accuracy...)

Also, the barometric altimeter depends on an atmospheric model (ie assumed temperature and humidity profiles). If the current profile is different from the model (plus the effects of wind...), it will introduce additional errors. It also assumes no horizontal change. Thus greater the altitude and position differences between the calibration point and the measurement point the greater these errors are likely to be.

Doug

Time for a DougPaul GPS bake off!

 

[TCD]

This is interesting. But haven't I seen many presentations of serious research involving very accurate measurements of altitude using the GPS system and a research instrument placed on a location (say, Everest) with altitudes being measured within fractions of inches? If we are really concerned about Owl's Head, how much would it cost to bring instrumentation like that to bear?

 

[DougPaul]

This is interesting. But haven't I seen many presentations of serious research involving very accurate measurements of altitude using the GPS system and a research instrument placed on a location (say, Everest) with altitudes being measured within fractions of inches? If we are really concerned about Owl's Head, how much would it cost to bring instrumentation like that to bear?

It just costs money to hire a trained top-grade GPS-trained backcountry surveyor with top-grade equipment (itself costing many thousands of dollars). The data collection will take hours to days and will require post-processing. (Tree clearing or a sturdy tower reaching above the trees may also be necessary.) Accuracies of ~.5 cm are possible by these techniques.

One must also establish the geoid (mean sea level) in the area to a similar accuracy. (Altitude is the height above the geoid.)

Is the altitude of a minor peak in the Whites really worth it?


Everest is interesting, not only because it is the highest (above sea level) peak in the world, but because it is rising due to tectonic movements. Mauna Loa is higher above its base (~33K ft) and Chimborazo (at 1 deg S) is higher above the planet's center because the earth has a larger radius at the equator than elsewhere. http://geology.com/records/highest-mountain-in-the-world.shtml


There are research arrays of high accuracy GPSes placed in active tectonic, earthquake, and volcanic zones. Having GPSes in place on both sides of an earthquake fault can be very informative...

Doug

 

[TJsName]

The GPS I use (my smart phone) currently reads about 100' low - I need to re-calibrate it. For instance, it currently has me at -13 feet, despite being on the 2nd floor on a hill in Boston. I should probably take care of this now that I'm thinking of it...

 

[hippotone]

Thank you everyone for all of this information and reasons for what could have been the cause. One of them was a Dakota 20 but I am not sure how or when they were calibrated. I should probably get a GPS of my own..

 

[DougPaul]

Thank you everyone for all of this information and reasons for what could have been the cause. One of them was a Dakota 20 but I am not sure how or when they were calibrated. I should probably get a GPS of my own..

Consumer GPSes are marketed as being simple to use, but there is a lot more (both in equipment and operator skill) involved in getting sub-cm accuracy results.

Consumer GPSes are complex devices and can fail (ie give poor results) for a number of reasons. Unfortunately, one may have to understand the inner workings of the device to understand the failures. That said, if a few guidelines are followed (eg keep a good view of the sky), they generally work pretty well. When used properly, the nominal horizontal accuracy is 95% probability of being within 10 meters and I have described the vertical accuracy in a previous post.

Accuracies will improve when multi-frequency operation becomes operational. (Current consumer GPSes use only the L1 frequency but the newer and future-launched satellites will also support the L5 frequency. When the second frequency becomes usable, it will be possible to reduce the biggest source of error--the ionosphere. L2 is currently available, but it only supports a military signal.) I don't recall the deployment schedule off hand, but IIRC it should be within the next 10 years. You will, of course, have to buy a new GPS to take advantage of the new frequency...

Doug

 

[BISCUT]

God love ya! And to think, I returned my Garmin because I thought it was a distraction! You guys are well above me and I do appreciate the read and the time you have into it.

 

[ChrisB]

God love ya! And to think, I returned my Garmin because I thought it was a distraction! You guys are well above me and I do appreciate the read and the time you have into it.

LIDAR technology might resolve the issue of summit heights once and for all. "It is part of a national project announced last year by the Obama administration to map the entire country using LiDAR. The goal is to have it completed in eight years..."

See this article for details...

cb

 

[Remix]

What ChrisB said....