Benchmarks Layer


Forum » General Forums » Supporters

By TallPaul on 11/10/16 at 9:32am

Benchmarks are trig points' little brother which, until recent times, the OS still relied on for track1ing height above Ordnance Datum and there's still a lot of them scattered about Great Britain.

Here's an example: one carved into the buttress of a church tower:

Benchmark on a church

You can read more about them here and you'll notice at the bottom that they offer the complete database of about 500,000 benchmarks under an Open Government Licence. So I'm thinking about us adding them as a layer (not least for nerdy purposes of my own).

But would anyone else find that interesting? Like the water levels layer you would be able to turn it on and off and, when on, clicking on it a benchmark marker would show you information about the benchmark taken from the database.

  1. A track records where you have been and when you were there. For more information click here.

By Lancashire Lad on 11/10/16 at 1:12pm

. . . . . But would anyone else find that interesting? . . . .

Yes indeed!

In fact only yesterday I was looking at that very spreadsheet to see if it included any info on the Bowfell summit bench mark I'd photographed on my "Bowfell via Ore Gap" walk (see walk report).

Unfortunately I couldn't find anything there.

On the premise that the actual location of Bowfell's "arrowed boulder" is approximately Grid Ref NY 2446 0645, I searched the spreadsheet and looked at all results beginning with Grid Ref NY 24 (of which there are many), but none seemed to be relative to NY 24 06 (or close enough to that specific location).

Which is odd, given that the spreadsheet is stated as being the complete bench mark archive.

(I recall reading somewhere that the bench mark was put there well over a hundred years ago, but haven't yet been able to discover the specific year).

Best regards, Mike.


By Lancashire Lad on 14/10/16 at 1:43pm

. . . In fact only yesterday I was looking at that very spreadsheet to see if it included any info on the Bowfell summit bench mark I'd photographed on my "Bowfell via Ore Gap" walk (see walk report).

Unfortunately I couldn't find anything there. . . . .

As a follow on to the above, I've had some email dialogue about this particular bench mark with Ordnance Survey, and received a reply which included " . . . . we can find no record of a Benchmark on Bow Fell. It may be the mark is associated with the Principle Triangulation (as opposed the retriangulation) or levelling associated with the 1st Geodetic Levelling, i.e. pre 20th century, for which we have no digital records.
Can you please send us a photograph so we can identify and investigate this further. . . . "

I have sent them the requested photos, and have asked them to inform me of the results/conclusions of their further investigations.

I will update again here, if I hear anything further.

Regards, Mike.


By TallPaul on 14/10/16 at 2:41pm

Well my news is that we now have a working benchmark layer on our test system which I hope we can release to supporters very soon.

However it's becoming increasingly clear, having looked at what we're seeing and comparing it to what I've found on the ground that there are a lot more benchmarks out there than are shown in their database. I've been looking at some OS 25 inch to the mile (so about 1:2,500) maps from 1892 to 1905 and those show lots and lots of benchmarks and, having done a quick survey of the local ones, quite a few still exist but very few are in the OS database so I assume, based on what they told you, that these were part of the 1st Geodetic Levelling.

I can feel a project, and possibly another web site, coming on.


By Lancashire Lad on 14/10/16 at 3:10pm

Hi Paul,

I assume that you are aware of this website? - http://www.geog.port.ac.uk/webmap/thela ... kemenu.htm

which includes this page on Bench Marks: - http://www.geog.port.ac.uk/webmap/thela ... enmrkf.htm

On that page, (which does actually have two references to the Bowfell summit benchmark), is this comment - "Although you automatically think Ordnance Survey, bench marks are made by other surveyors, sometimes just for an immediate project" - so it may well be that some of the bench marks we see, and have always assumed were made by the Ordnance Survey, may well have originated elsewhere.

Regards, Mike.


By TallPaul on 14/10/16 at 4:14pm

Yup. I've come across that issue before too and indeed one the benchmarks I found while looking for local ones looked like an OS one but wasn't on the 25in to the mile map so Bob alone knows what that was.

One of the things that intrigues me in all this is that a lot of the OS benchmarks are carved into private property. Presumably they had some sort of wayleave that let them do that without having to secure the landlord's permission? ...


By TallPaul on 15/10/16 at 2:36pm

Well my news is that we now have a working benchmark layer on our test system which I hope we can release to supporters very soon.

The good news is I think we're now happy with the software so it's ready to go live however we're having to reload all 500,000 benchmarks into our database first which is going to take a few hours to do so it will probably be late today or some time tomorrow before it goes live.


By TallPaul on 16/10/16 at 8:11pm

OK, this is now live. There's a new option the menu which lets you toggle it on and off. You only see the option if you're a supporter and you only see benchmarks if you're zoomed in to 1:50K or 1:25K.

There are 517,009 benchmarks in the database so we're only displaying those in a 10km square around the centre of your map, which should be off the edge of your screen unless you have a very high resolution screen. It re-loads the benchmarks every time you move the map. If you move it a lot it may take up to five seconds before they are refreshed (to keep the load on the server down).

I've also found details of the 190 FBMs mentioned on this page which I'm in the process of importing. Once that's done we'll put those online too.

Anyway let me know what you think of it.


By Lancashire Lad on 17/10/16 at 1:48pm

Hi Paul,

I've just had a bit of a play, and it is working perfectly.

I assume that the numbers shown on the map layer, against each bench mark, refers to the "order" as noted on the description when an individual bench mark is clicked on?

To be honest, I don't know what that "order" actually means. I'd hazard a guess that you know a lot more about bench marks than I do. Just a thought, but perhaps there is some sort of basic "glossary" of bench mark terminology that might be able to be uploaded and accessed on Walk Lakes to help people who are using this new facility?

Whilst browsing I came across this by pure chance - proving that even the OS are not beyond the odd error or two ;) : -

The description for the bench mark in question says it is of the rivet type, verified in 1978, at 23.829 metres height, and -1m above ground??? Clearly incorrect, since it is located almost on the 310m contour line for that location. (and how would one verify a bench mark at minus 1 metre above ground? :shock: :D ). Obviously nothing that you can do about such things, as your markers will be picked up from the grid references given on the OS database.

Anyway, OS database errors apart, the layer is working perfectly.

Regards, Mike.


By TallPaul on 17/10/16 at 4:31pm

I've just had a bit of a play, and it is working perfectly.

Excellent.

I assume that the numbers shown on the map layer, against each bench mark, refers to the "order" as noted on the description when an individual bench mark is clicked on?

Yup. Essentially it's a hierarchy, so the fundamental benchmarks (FBMs) are considered to be error free and the rest have an accuracy based on their order and their distance, d, in km from the nearest FBM (I hope you're all paying attention at the back, I'll be asking questions later ;-) ) thus:

+-------------------+----------------+
| Type of bench mark| Maximum error  |
+-------------------+----------------+
| fundamental       |  Error free    |
| Geodetic          |  ±2mm × ?d     |
| Secondary         |  ±5mm × ?d     |
| Tertiary          | ±12mm × ?d     |
+-------------------+----------------+

For more on this see the OS document A guide to coordinate systems in Great Britain and in particular 5.3.3.

Whilst browsing I came across this by pure chance - proving that even the OS are not beyond the odd error or two ;) [...] The description for the bench mark in question says it is of the rivet type, verified in 1978, at 23.829 metres height, and -1m above ground??? Clearly incorrect, since it is located almost on the 310m contour line for that location. (and how would one verify a bench mark at minus 1 metre above ground? :shock: :D ). Obviously nothing that you can do about such things, as your markers will be picked up from the grid references given on the OS database.

Yes, I was aware of the subterranean benchmarks. They became obvious when I was analysing the data prior to loading it. There are 6,718 of them as follows:

+------------+----------+
|     Above  |    Total |
+------------+----------+
|    -3.5000 |        1 |
|    -3.0000 |        2 |
|    -2.9000 |        1 |
|    -2.8000 |        3 |
|    -2.7000 |        2 |
|    -2.6000 |        1 |
|    -2.4000 |        6 |
|    -2.3000 |        2 |
|    -2.2000 |        2 |
|    -2.1000 |       10 |
|    -2.0000 |        4 |
|    -1.9000 |        4 |
|    -1.8000 |        9 |
|    -1.7000 |       10 |
|    -1.6000 |        9 |
|    -1.5000 |       30 |
|    -1.4000 |       39 |
|    -1.3000 |       45 |
|    -1.2000 |       64 |
|    -1.1000 |       81 |
|    -1.0000 |      114 |
|    -0.9000 |      121 |
|    -0.8000 |      237 |
|    -0.7000 |      221 |
|    -0.6000 |      428 |
|    -0.5000 |      632 |
|    -0.4000 |      673 |
|    -0.3000 |      979 |
|    -0.2000 |     1600 |
|    -0.1000 |     1388 |
+------------+----------+

What this means on the ground I've no idea as yet. My plan is to pick a couple and go take a look.

As for that particular benchmark also being way off height-wise (if you right click at that location the OS data suggests it's at about 306m not 24m, which looks about right from the surrounding contours) again I'm not sure what that's all about but I did find other typos, for example "197A" for the verified date, so this dataset is clearly not completely accurate. I suspect it may have been entered from paper records and was perhaps outsourced ...


By TallPaul on 18/10/16 at 11:44am

I've also found details of the 190 FBMs mentioned on this page which I'm in the process of importing. Once that's done we'll put those online too.

And those are now being shown, with a pin of "0". Enjoy.


By Lancashire Lad on 18/10/16 at 12:11pm

. . . In fact only yesterday I was looking at that very spreadsheet to see if it included any info on the Bowfell summit bench mark I'd photographed on my "Bowfell via Ore Gap" walk (see walk report). Unfortunately I couldn't find anything there. . . . .

As a follow on to the above, I've had some email dialogue about this particular bench mark with Ordnance Survey, and received a reply which included " . . . . we can find no record of a Benchmark on Bow Fell. It may be the mark is associated with the Principle Triangulation (as opposed the retriangulation) or levelling associated with the 1st Geodetic Levelling, i.e. pre 20th century, for which we have no digital records.
Can you please send us a photograph so we can identify and investigate this further. . . . "

I have sent them the requested photos, and have asked them to inform me of the results/conclusions of their further investigations.

I will update again here, if I hear anything further.

Regards,
Mike.

Ordnance Survey received my photos, and, having further investigated the particular mark, have replied as below: -

Thank you again for your follow-up email regarding this benchmark.

We have now received feedback from the Geodetic Team which I summarise below.

First they thank you for providing further details including photographs. From the size it could be a very early benchmark but if so it is damaged (as there is no height reference point) and predates any of the information we now hold.

There are two types of cut marks for either horizontal or vertical surfaces, the most common are the latter which have a horizontal line across the top of the 'Government broad arrow' mark, which forms the 'bench' in the benchmark. For horizontal surfaces at the point of the Government broad arrow a brass rivet or bolt is installed or alternatively a small indentation which is designed to accept a 5/8" ball bearing. This latter type of mark is known as a pivot type benchmark (see attached diagram).

The mark in the photograph may be a pivot type but their guess is that it is not. In its current condition, it is clearly not suitable as a levelling reference point and appears to lack either the hole or brass bolt to which the height is actually referenced. It should also be noted that many government agencies have used the 'Government broad arrow' mark and their guess is that this may well be a boundary marker of some kind.

Not the definitive result that I'd hoped for, as it now looks highly unlikely that I'll ever discover exactly by whom, and in what year, this mark was made. But at least I have more information than previously, and have learned a bit more about Bench Marks along the way. (I have thanked those at the OS for their time and efforts in trying to answer this query - It is gratifying to know that there are still some organisations out there that do try to help when your average "Joe Bloggs" has a genuine query).

Regards, Mike.


By TallPaul on 20/10/16 at 10:33am

I have tweaked the benchmark popup to translate some of the OS shorthand descriptions of benchmark types, in particular the cryptic "FL BR" which is as Flush Bracket. I've also linked to this page which gives a good explanation of the various types, with examples.



WalkLakes recognises that hill walking, or walking in the mountains, is an activity with a danger of personal injury or death.
Participants in these activities should be aware of and accept these risks and be responsible for their own actions.