UKC

Research carried out by surveyor Alan Dawson, owner of the Grahams hill list, suggests that Ordnance Survey maps - and the hundreds of guidebooks and websites that rely on them - list the wrong summit height for 'over half' the mountains in Britain.

Read more

My question on this is, given there appears to be a universal appetite to produce accurate and consistent mapping, will the survey results be fed into to OS? This being as they are the most widely used maps by hill-goers.

That'll be erosion then. 

the answer is in the article...

"Where evidence of an appropriate quality survey is provided, we accept data from third party surveyors so that we can update our mapping and the National Geographic Database."

That certainly shows willing from OS.

I guess the difficulty there is knowing the veracity of any new data being fed into the OS.  I mean if one well meaning amateur says a hill is 451m and another says its 443, what do you do ? Do you replete the contours near 450m or not ?

Personally I'd prefer OS resources were used to get cycle paths properly mapped!

The very wrong spot height west of Beinn Fhada was definitely changed by the os on the most recent 1:50000

We climbed Carnan Eoin (143m) on Colonsay a few days ago and it definitely felt like a Munro

It was always common knowledge that brown numbers (estimated by photogrammetry) on OS maps had a margin of error. One of the reasons I was never keen on using feet as a unit, the rightmost digit in a height was usually meaningless.

There are several teams out there with high end kit producing some consistent results, but really only of use to those of us on the barmy end of the bagging spectrum.

Imagine the confusion you could have caused if you'd waited three days to release this... 

If you've ever read Wainwright's 'Walks on the Howgill Fells' (1970?) he describes finding an absolute clanger by the OS in their height of Uldale Head. I think he proved they were about 200 feet short using basic visual evidence on the spot (sorry!).

Perhaps a more accurate title might be that “Mountain heights are not consistent”.

To say mountain heights are wrong implies there is a singular definite answer. There is not. Height above what? and where? And when?

The OS state heights above the geoid, a surface that represents where a sea-level would be if the land wasn’t “getting in the way”. The chosen sea-level (0 height) was the mean sea level at Newlyn in the period 1915-1921. Since then sea levels have changed and the Britain Isles have moved about and tilted a little bit.

So what height is any hill and does it really matter?

Quite right.

Also: "97% are correct to within three metres."
Characterising that in the headline as "most mountain heights are wrong" - jeez, bit harsh mate. 

That's more than good enough for pretty much anybody, especially for recreational purposes.  I doubt even the nerdiest of the nerds records their day's height gain to anything like that precision (though a few may kid themselves otherwise).

You said it yourself - relative to the (WGS84) geoid, the colloquial proxy for which is “sea level”. Nobody in surveying has used the sea level approximation you describe since at least the start of the GPS era.

Pretty easy to include the relevant metadata - and anyone who can be bothered to lug a differential GPS up a hill for fun will almost certainly be keen to share this in detail. Also depends on how the info is shared - most Leica/Trimble units will spit out all the info which can be shared and would be very annoying to fake.

Life is rarely that simple. There are many CRS’s, OSGB36 is what OS use and it does indeed use Newlyn as its height datum. What your GPS measures relative to, is a different matter.

See https://www.ordnancesurvey.co.uk/blog/ostn15-new-geoid-britain

The OS (used to) require a certain number of hours of data relatively to a number of base stations. For data quality and audit purposes one imagine.

Whilst this likely a bit OTT for most purposes these requirements were well specified. Not all recreational surveyors made surveys to these exacting standards and, in these cases, the OS did not use their results. 

The headline could just as easily have been that the vast majority of mountain heights are correct.

By taking an average of these figures, it is possible to say confidently that the summit is 40,000 1/2 feet above sea level.

A stupid clickbait headline - give your sub-editor a good talking to.

As the article says: most mountain heights are wrong, 'but not by much.'

This is why it is important - sometimes you "need" to know which top is the highest.

Although (I believe) he's amateur in the sense that he's not being paid to do it, I don't think anyone would consider Alan Dawson's height surveying work as "amateur".

He's the source of a significant proportion of the changes to improve the correctness and accuracy of the DoBIH (Database of British and Irish Hills) - the master list that all mountain/hill bagging nerds (that includes me 😁) refer to.

I think that's a missing zero on your map?

I wish! Sadly, just the lack of oomph in the legs.

Interesting article, clickbait headline aside.

Overall, as much as I'm as pedantic as anyone, how amazing were the old surveyors to get so close with such rudimentary equipment? No satellites, just maths! UK mountain heights, and maps, have been incredibly good for a relatively long time now.

Teammates and I climbed and surveyed many of the highest peaks in Antarctica from 2001-2009. We carried a Trimble 5700 DGPS unit to the top of peaks and either waited there for hours or went down and had it retrieved by another pair later. We always took a minimum of one hour data (advised from the start by Geosciences Australia in order to use the AUSPOS processing system) and on Mt Vinson and most of the highest tops we had around 10hrs of data, as that was initially the limit of two sets of batteries. Within a few years we could get that out of one set.

Our height of 4892m for Vinson was only 5m below the old official USGS height of 4897m, so just within their +/- 5m noted on all their maps. We shared our data with the USGS and BAS, and the 4892m is now on most modern maps. The article here notes that some UK peaks might be out by a few metres. We found that Mt Craddock, thought to be the 5th-highest mountain in Antarctica, was actually 282m lower than previously estimated.

http://publications.americanalpineclub.org/articles/12200628600

Observations and photos from that trip, added to observations and photos from the previous season, meant that not only was the height-order of Antarctica's highest peaks changed considerably, but we discovered a whole 'new' mountain - Mt Rutford.

http://publications.americanalpineclub.org/articles/12200726102

When the Sentinel Range was first surveyed in the early 1960s, by teams on the ground (ice) and from aerial imagery, they had Vinson at 5140m - a figure you could still see on maps until quite recently. Even when they resurveyed in 1979, the teams out on the plateau struggled to sight the true top of Vinson as it's relatively flat, a bit like some of the challenges mentioned in the article. Mt Tyree on the other hand is much sharper and was easier to sight. They could barely even see Mt Shinn, so it had no set height, despite being suspected to be Antarctica's 3rd highest mountain. It was solving this (to me) big unknown that was the genesis of my decade of climbing and surveying with the Omega Foundation.

But as others have said above, all these figures are relative - to the geoid, to changing 'sea levels' and to improving technology. Our figures too will one day be proven 'wrong'. It's easy to get lost in the minutiae and forget how incredible it was that decades ago, centuries in some cases, long before satellites, people were able to analyse and understand our surroundings, to order our world, with such accuracy upon which other systems could be built - towns, maps, transport, shipping, aviation, geography...

I used to work with Alan briefly in the late 80s when he was a data scientist at the Uni of Liverpool and we went on a few walking trips together. We did a route on Milestone so he could get used to rock, to get up the Inn Pinn, towards the end of completing the Munros. Sounds like an excellent retirement project from a lifetime dedicated to geekery! I can only guess the amount of time spent on this, and something very British about it. My take is that the OS went for ‘as good as we can make it within the parameters we set for ourselves’. In some cases Alan has been able to improve on that.

A lot of Austrian huts still proudly state their heights based on the old Austro-Hungarian sea level somewhere in the Adriatic, because that's a bit lower - and therefore the huts a bit higher! - than whatever the current baseline used in the eastern Alps is.

Something you may also know is that tides affect the height of the land as well as the sea, up to tens of centimetres (I can't find how much in the UK) . So unless an average height of a summit is being taken over a full tidal cycle, it shouldn't be reported any more accurately than the nearest metre. 

 Can this guy weigh in on Three Pebble Slab grade!! 

🤣

European countries use a variety of height datums (data?!). Quite a nice comparison here:

https://evrs.bkg.bund.de/Subsites/EVRS/EN/Projects/HeightDatumRel/height-da...

You could argue the French “exaggerate” the heights of their mountains more than the Austrians. (They don’t! It’s just a choice that was probably made for pragmatic reasons over a century ago.)

I’d echo what others have said regarding the amazing job surveyors did with rudimentary tools and with poor transport networks etc. all those years ago.

I think you are referring to solid earth tides / terrestrial tides (I.e. how land masses move in response to cyclic gravitational changes).

The GPS system takes this into account and, in a sense, the Newlyn datum also takes this into account as it was an average over 6 years.

I mean fundamentally all measurements are 'wrong' because every measurement has an associated accuracy and precision which contribute to a uncertainty that is always present (not matter how small).

Does this mean all those poor Munro-baggers are going to have to get out again for one more hill? Oh dear...

...you must be fecking joking!!!

Not very much in the UK -- *except* in the case of Cornwall  youtube.com/watch?v=lCA0II1sVZA&

</begin-nerd-alert>

For Ceum na h-Aon-choise, chances are many will have gone over it (although they may have missed the absolute highest point if the path is a bit to one side) on the traverse of Beinn Fhada (Kintail).

At most it'll be classified as a Munro top.

On streetmap the 1:25,000 shows 891m but the 1:50,000 shows 925m.

Google maps 3D view clearly shows it's significantly higher than the next bits down the ridge at ~870m.

</end-nerd-alert>

Interestingly it is not direct tidal forces making the land go up and down but the weight of the water on the earth;s crust.

Yes, terrestrial tides. That's interesting. My presumption was that the Newlyn datum would take this into account quite well as it moves with the UK land mass to an extent, rather than the WGS84 ellipsoid which is static.

That sounds simple and willing in theory, but in practice reporting a mountain height to OS is a laborious process and they are not usually interested. Also, they ask for two hours of data collection (minimum of one hour) which is asking for hypothermia and not realistic when you can get just as accurate results from 15 to 20 minutes on a summit, as several replicated surveys have shown. Alan

1 hour doesn’t seem so much longer than 20 minutes for the sake of a consistent record between DoBIH and the GB national mapping agency. Especially if one is so inclined to go to the effort of removing and rebuilding cairns etc. to conduct a survey on the first place.

Should say that in Britain (I.e. excluding N.I.) we actually have  more than one height datum. Since 2015 there is: Mainland Britain (Newlyn, 1915-1921), Outer Hebrides (Stornaway, 1977), Orkney (based on levelling to Newlyn, 1977), Shetland (Lerwick, 1900), Scilly Isles (St Mary’s, 1887).
 

They don’t all align so a hill on Harris might be taller/shorter (can’t remember which way it is) than a hill of the same given “height” on the mainland...

Obviously they can’t.

However, if a survey is published but is not conducted to a standard such that the national mapping agency will not use the result (but it is generally known to be sufficiently accurate for the purposes hill height measuring) and that survey has the effect of moving a hill into another (arbitrarily defined hill) classification then things get a bit, unnecessarily, complicated for those what publish lists of hills…

That's funny!!

Owns in the sense of being the custodian of the list, improving the accuracy as better measurements become available, correcting any errors and issuing "official" updates as necessary.

This is definitive proof, finally, that the world is not round as claimed, and I'm sure someone cleverer than I will be along soon to explain how.

😉

Perhaps the height in millimetres, that is, the precise length of the route 

It’s 482 metres

As a nerd who is fascinated by surveying I found the article title very exciting… sadly the content didn’t quite live up to the hype. 

Alan, in the case of my Antarctic work the 1hr minimum stipulated was, as told to me, due to a combination of the low angle of satellites in Antarctica meaning fewer passes for data collection and that the AUSPOS system operated by triangulating ground stations - a problem in Antarctica where 98% of the ground is covered in moving ice.

https://www.ga.gov.au/scientific-topics/positioning-navigation/geodesy/ausp...

So it went off stations quite distant from each other and at quite sharp angles for the triangle, which I was told was not ideal. One station was over 3000km away on the other side of Antarctica, one was up near Rio Gallegos in Argentina (though they varied slightly over the years).

You can't even rely on the datum when the rocks of Cornwall bounce up and down twice a day.

youtube.com/watch?v=lCA0II1sVZA&

That video was already posted by CantClimbTom (https://www.ukclimbing.com/forums/ukc/most_mountain_heights_are_wrong_says_...)

And may refer to the same phenomenon highlighted earlier by Jimbo C (https://www.ukclimbing.com/forums/ukc/most_mountain_heights_are_wrong_says_...).

I replied at the time (https://www.ukclimbing.com/forums/ukc/most_mountain_heights_are_wrong_says_...) that, “The GPS system takes this into account and, in a sense, the Newlyn datum also takes this into account as it was an average over 6 years.”