Accuracy of Topo Map Altitudes

[iluka]

A question regarding topo map altitudes. I was on a recent trip in the Sierra Nevada. We climbed Mt. Barnard which is listed on the topo map as being 13,990 feet in elevation. The topo map says the vertical datum is from the year 1928. I happened to turn on my GPS on the summit for the heck of it and it read an altitude of 14,010 feet based on a 3D fix on at least 8 satellites with crystal clear skies. The unit said it was confident within 20 feet (which just overlaps with the topo elevation). Which would you believe?

[AlpineK]

I'd believe the 1928 topo data. Elevations are the least reliable data on a GPS.

[faster_than_you]

Let me get this straight, you're concerned about a 20 foot deviation? I'd hold out for 27 satllites... But that's just me.

 

I hope you can sleep tonight.

[iluka]

Let me get this straight, you're concerned about a 20 foot deviation? I'd hold out for 27 satllites... But that's just me.

 

I hope you can sleep tonight.

 

I can sleep fine, thanks. Not concerned... just curious.

 

Also, for people into climbing 14ers, the 20 feet makes a difference in the case of this and, perhaps, other peaks with elevations near 14K.

Edited May 16, 2006 by iluka

[olyclimber]

I'd bring a step ladder to summit just to be on the safeside...but seriously, GPS's are not very accurate of altitude, at least compared to a barometer. And to get down inches...well, best to go up there and check yourself to see how high it is.

[MarkMcJizzy]

The topo map all the way. For the following reasons:

 

1) The vertical dimension is a GPS units weak link, and most prone to error. Ed Seedhouse has espoused on this at length on other threads.

 

2) A summit is a high quality target for a surveyor.

 

3) The old USGS topo maps were made by craftsmen, true experts at their field

 

4) If it has a BM on it, bet your house that it is right.

[genepires]

I don't trust anything with bowel movements on it.

[Gary_Yngve]

http://erg.usgs.gov/isb/pubs/factsheets/fs04000.html#accuracy

[AlpineK]

That's a DEM Gary not a topo. The DEM is interpolated from a topo.

[ashw_justin]

GPS reads 10,778 feet on the summit of Mt. Baker, and I remember it being close to dead-on for Adams and Hood too, although I forget the exact values. Shit is scary accurate in my experience. Fun toy, not to mention a beautiful thing to have in a whiteout (if you had the foresight to record your uptrack).

[Gary_Yngve]

That's a DEM Gary not a topo. The DEM is interpolated from a topo.

 

The same applies for a topo. If the site hasn't been surveyed, then it's +/- 20 feet. Or more, if the terrain is irregular.

[still_climbin]

Maybe it grew! You know, tectonic plates colliding and all.

[gslater]

I'd believe the 1928 topo data. Elevations are the least reliable data on a GPS.

 

Elevations are indeed the least "reliable" part of the GPS system, but with a GPS receiver that does statistical averaging over lots of samples when it thinks you've stopped moving, and given sufficient time (a few minutes, at least) the GPS will be pretty damn precise. Much more precise than any decades-old survey, that's for sure.

 

But then again, I probably wouldn't spend much time worrying about it. It's all completely debateable about how accurate (as opposed to how precise) the reference datum is.

 

The main downside to GPS for elevation use in wilderness navigation is that its numbers are regarded by most people as absolutes, and they got sidetracked by how their GPS numbers don't match what's printed on the topo. The relative elevation differences from one point on a topo map to any other point on that map are pretty damn accurate, and that's just fine when you're using a barometric altimeter and frequently recalibrating it against the topo at known points. If you're using the GPS for your elevation data while trying to navigate using a topo, the difference in elevation numbers can be confusing.

[cj001f]

Much more precise than any decades-old survey, that's for sure.

Oh bollocks. Because it's old doesn't mean it's poor data - I'd venture the errorbars for their measurements were less than those on the GPS receiver

[TrogdortheBurninator]

GPS elevation works poorly with only 3 satellites due to the error in vertical triangulation. WIth many satellites registering, the accuracy increases significantly. With a clear sky, I will trust GPS way more than barometric because barometric pressure is always changing slightly from the time that you calibrated. I would guess that snowpack could be somewhat of an issue too when you are that close to a benchmark elevation.

[gslater]

Much more precise than any decades-old survey, that's for sure.

Oh bollocks. Because it's old doesn't mean it's poor data - I'd venture the errorbars for their measurements were less than those on the GPS receiver

 

Over a small area, sure, but not over the range covered by a typical topo or bigger. All those little errors add up in a hurry, and the technology back in the days of yore was good, but not where it is now.

 

Not trying to slight those who did that sort of surveying back then; my grandpa was one of them. I'm sure they did a fabulous job with what they had. But if the old methods were more precise, then why is everyone migrating to GPS-based measuring? Convenience, sure, but nobody would give up substantial precision to get it.

 

I bet someone out there has studied this issue extensively. May have to spend some time digging around for it, as the boss is out all this week...

[klenke]

I use a 12-inch acrylic ruler to measure mountains. It's very accurate. I take it to the nearest 1/16th of an inch.

[cj001f]

Over a small area, sure, but not over the range covered by a typical topo or bigger. All those little errors add up in a hurry, and the technology back in the days of yore was good, but not where it is now.

 

Not trying to slight those who did that sort of surveying back then; my grandpa was one of them. I'm sure they did a fabulous job with what they had. But if the old methods were more precise, then why is everyone migrating to GPS-based measuring? Convenience, sure, but nobody would give up substantial precision to get it.

 

I bet someone out there has studied this issue extensively. May have to spend some time digging around for it, as the boss is out all this week...

 

Given that the peak in question (Barnard) was originally surveyed to be >14k but remeasured by a USGS team and found to be 13,989 I'd give some stock. Equipments only as good as the operator.

 

I'd worry more that Thunderbolt Peak is 4 feet shorter than previously surveyed

[Ed_Seedhouse]

For summits, Take the map, by far. Not even close. GPS altitudes are generally specified at +/100 meters without WAAS correction, and +/- 10 with WAAS.

 

But even if we assume the GPS altitude is exact, there's another problem: 14,000 feet above what?

 

Above sea level? Sea level varies by several meters every day, and mean-high-tide level by several meters over the world. But your GPS doesn't know anything about sea where you are at, anyway. Oh, and on the scale of tens of meters the Earth itself isn't round.

 

What the GPS gives you is a computed height above the "geoid", which is a mathematical model of mean sea level for the whole planet that it carries in it's memory. Of course, a geoid good enough to give really useful results requires a huge slice of memory, much more than any consumer GPS has, and so the models of the geoid the GPS carries is rather crude, and contributes much to the error of a GPS altitude.

 

No satellite configuration, however good the geometry, can do anything about the crudeness of your GPS's internal geoid model.

 

The summits of mountains on maps were determined by direct survey and the map, even a very old one, has a much better fix on the shape of the geoid at any given locality than any GPS can yet carry.

 

Contours are another matter - there are many errors in countour lines on maps because contours are very hard to do. But significant summits altitudes are known very well and are usually much more accurately reported on a map than any GPS can do with it's internal geoid model.

 

My own Geko 201 GPS reports an altitude for my house, with WAAS correction in effect, that varies by as much as +/- 15 meters depending on when I look at it. I am somewhere between 75 and 100 feet above sea level and that's all I really know with any confidence.

[Dechristo]

Geoid willickers! good info.

 

I am somewhere between 75 and 100 feet above sea level and that's all I really know with any confidence.

I'm somewhere between heaven and hell and that's all I really know with any confidence.

[Ed_Seedhouse]

> I am somewhere between 75 and 100 feet above sea level

> and that's all I really know with any confidence.

 

It's a fairly interesting question here in the Northwest. Do you live above or below the most likely tsunami height when the big one hits? Well, actually I'm on Vancouver Island which, from a Canadian perspective is actually the Southwest, but you know what I mean.

 

I'd like to know if a 30m tsunami will take my house away or if it will just take my lower neighbors. Unfortunately I can't find accurate enough contours on a map to tell me. Maybe I should hire a survey crew to ease my mind...

[curtveld]

Do you live above or below the most likely tsunami height when the big one hits?

Depends on how much slop there is!

[payaso]

WAAS is virtually useless in the mountains. In fact you will get better performance by disabling WAAS and you will consume less batteries as well.

 

It is appropriate for boating or anywhere with wide clear views of the horizons.