How accurate is the permanent way on OS Maps?

[petejones]

I've been toying with the idea of using an old OS Map (early 20th century) for a track plan and after seeing actual photos of some of the same track, the OS Map is quite off. I'm guessing that back in the days of drawing these maps by hand, they are just generalisations, especially where the permanent way is concerned.

 

Any thoughts?

[LBRJ]

Pretty much agree with you regarding a lot of railway permanent way.

Its not actually all that permanent over a longer period, particularly in sidings etc, and older OS maps tend to be drawn up from an amalgam of several years (or decades!) worth of data.

Conversely, things that  I know the dimensions of (extant houses etc) tend to be pretty much spot on with measurements 

[Wheatley]

The 25" series should be accurate for the date it was actually surveyed, but don't expect the cartographer to pick up on the difference between a diamond crossing and a slip for example.  The publication date of the map is not the date of survey,  especially for a later edition only revised for major changes. The smaller scale editions will have less detail and more compromises. 

[Joseph_Pestell]

We have been here before.

 

The standard can be a bit variable but differences will mostly be due to revisions on the ground after the survey.

 

Best is to start from the OS map to give you some idea of whether the station concerned suits the space you have. But for the fine details, at least so far as running lines are concerned, see if the Signalling Record Society has diagrams for that location and date. And, yes, as always, photographs are even better.

[Poor Old Bruce]

1 hour ago, Wheatley said:

The 25" series should be accurate for the date it was actually surveyed, but don't expect the cartographer to pick up on the difference between a diamond crossing and a slip for example.  The publication date of the map is not the date of survey,  especially for a later edition only revised for major changes. The smaller scale editions will have less detail and more compromises. 

 

Some of the larger conurbations were also mapped at 125", the relevant County Record Office or Local Studies Library should have copies from various dates. I agree with Wheatley about the 25" maps and add that occasionally a crossover may be missed off.

[IWCR]

The 2500 scale maps (and 1250 in urban areas) are the best you are going to get for any location unless you can locate the railways own detail track plans.

Each rail is shown on these along with signal posts, telegraph poles etc  but occasionally something is missed by the surveyor or misinterpreted  by the person producing the map itself. 

They are not perfect and as stated above may be a mixture of features from differing dates but as a starting point there is nothing better. 

Use them as a guide and reference photos for your specific period.

 

Pete

[Mark Saunders]

The track plan on large scale maps is only as good as the last revision for the part, even parts of quail maps especially terminals and works are not updated.

 

Find as many sources as you can!

 

Mark Saunders

[bécasse]

Crossovers being missed in the surveying work has been the biggest problem area that I have noted during years of using such maps. However this may not have always been the surveying team's fault as a lot of extra crossovers were installed during the Edwardian era as a response to the partial outlawing of rope/chain shunting, and this coincided with a major OS resurvey of the UK. The map always shows what was there at the time of the survey or partial resurvey.

As to the accurate depiction of buildings, I suggest that one should try overlaying maps of the same place from different surveys - which will quickly bring to light a lot of drawing or surveying anomalies, especially where there is photographic evidence that the actual buildings didn't change over a significant period.

[hayfield]

I have been trying to replicate a track plan from the early 1900's, certainly several things seem to be out of registrar, could it be this happens when maps are stitched together in the computer program, added to what has been said about the draughtsman's interpretation coupled with what actually happens when the track gang does on site, then I guess over time modifications occur. As said look at photo's (if available) and do your best of what evolves

[RailWest]

>>>I suggest that one should try overlaying maps of the same place from different surveys - which will quickly bring to light a lot of drawing or surveying anomalies......

 

I agree. One problem is where is a station spreads across two maps surveyed at different dates, or where part only of a map is re-surveyed at a late date.  As a typical example of a resultant problem, a signal originally surveyed and noted on the map in one location, subsequently moved to a new position further away, then surveyed and noted again on the map when that portion only was resurveyed, consequently it then appears as two separate signals on the same overall map. Same thing could happen to crossovers etc.

[Rowsley17D]

I've used the National Library of Scotland's OS map facility to plan my interpretation of Buxton Midland (Derwent Spa) in Templot which allows you to put the trackwork on top of the map. 

 

 

The site also allows you to put on top the present day satellite view which lines up with what is now left - the LNWR station, Bridge Street bridge (near C.R.) and the roads in general. Where I have facing and trailing cross-overs on the main line was in fact a scissors crossover but my turnout building skills I did not think were up to it.

[Tim V]

This has been replied to before.

 

The OS maps are not definitive for track layouts. The surveying of the track was incidental to the purpose of the survey, which was to record boundaries. Particularly the older ones (pre national grid) are suspect. But for buildings they are better.

 

They have to be used in conjunction with other sources. Caveat emptor.

[IWCR]

Crossovers,  I found one on a 2500 map I was looking at which only had one rail, the other not being shown at all.

 

Regarding Satellite views, yes very useful but again be careful, these are made from many individual photos joined together, they are tweaked and adjusted to match the adjoining ones hence buildings etc in the centre of a picture may be a good plan view, those towards the edge rather oblique, possibly not well matched to their other half on the next photo. For an overall view generally good but zoomed in for fine detail less so.

 

Pete

 

[Nearholmer]

I've noticed that some of the c1900 London 1:1056 maps have such pernickety detail of p.way that they can only have been surveyed and drawn by ex-railway chaps or by enthusiasts. Others in the same series, much more sketchy on railway detail.

[RLWP]

On 13/07/2019 at 18:22, Tim V said:

which was to record boundaries.

 

Actually, that isn't the original purpose of the Ordnance Survey maps

 

The clue is in the name

 

Richard

[Stuart]

After 43 years of working with them, I would say that you do have to treat Ordnance Survey maps with a degree of caution. That statement isn’t intended to denigrate the OS, either body or the surveyors, a number of whom I have counted as friends or former colleagues, but I do think you need a certain amount of knowledge to interpret what you see.

 

Firstly, the Ordnance Survey map is a topographical map, it is intended to show physical features – it is not, and never was, intended to show property boundaries (s12 Ordnance Survey Act 1841). It does, of course, show physical features which are property boundaries, but that isn’t it’s intended purpose.  

 

Next, the maps show physical features within the limitations of scale, and to a specified degree of tolerance. For instance, in 95% of instances on a 1:2500 scale map, a distance of 100m in the real world will be shown as between 98.1m and 101.9m on the map, so a tolerance of nearly 2%.

 

The first 1:2500 maps were surveyed on a County basis (hence Rowsley 17D’s plan above is from the Derbyshire series, map XV.13). The were originally surveyed by men with brass and wooden theodolites, and brass and steel chains. Their survey reports were then hand plotted and examined before being printed. The important thing is that they were done by individual surveyors, plotters and examiners and, no matter how stringent the instructions and examination, individuality did show through. I have seen first edition maps where, say, crossovers between running lines are nicely shown with their curved lines, and others where they are simply shown as two straight lines. But they should show the actual position of the lines with a decent amount of lineside detail. Paras 99-116 in these old instructions https://www.ordnancesurvey.co.uk/docs/ebooks/historical-instructions-to-field-examiners.pdf give some idea.

 

Revision is also a major issue to consider as RailWest says. A map may be revised because either ‘it’s time’ or because there has been some significant change in the area covered by the map. Unfortunately, the surveyor may not have noticed that the railway line has changed, especially where change is in an area difficult to see from adjacent streets etc. Or it may been more important to survey a new road or shopping centre then record the addition/removal of a siding. Even aerial photography doesn’t solve this problem – an area may be in shadow, or the photographs mis-interpreted.

 

And then there’s the issue of your chosen site crossing a sheet join (or worse still, being in the corners of four sheets) where the revision dates aren’t consistent.

If you’ve obtained a railway plan, you may not be interested in the corresponding OS map, but if you do compare them, you’ll probably find they don’t match – the OS map will be drawn on the Cassini (pre WW2) or Mercator (post WW2) projection, to the tolerances noted above whereas a large scale plan of, say, a station, will be very accurate., so they won’t exactly match.

 

BUT, having said all that, the OS is still probably one of, if not the, best mapping institution in the world.

[tractionman]

Some further discussion here:

 

 

 

[ianathompson]

On 24/06/2019 at 11:29, hayfield said:

I have been trying to replicate a track plan from the early 1900's, certainly several things seem to be out of registrar,

 

Only just seen this topic, so a late reply.

 

OS maps are a representation of a curved surface (the Earth) on a flat piece of paper.

Map makers get around this problem by distorting their map in some way. This is known as a projection.

Look in any atlas and somewhere, usually on the edge in small letters, it states the projection.

What most people would consider to be the common World Map is drawn to the Mercator projection.

 

Modern OS maps use a grid superimposed over the whole country.

The grid  distorts the map, particularly at the extremities, but not by too much.

 

In the nineteenth century it was standard practice for each individual county to be surveyed on its own unique grid, for large scale maps, which resulted in less distortion at the county's extremities.

However, when the sheets showing the boundary of one county with another are compared differences become very apparent. 

If we take the Lancashire/Yorkshire boundary as an example, the Yorkshire County series map will be at a totally different orientation to the Lancashire one, due to the use of a different grid.

This practice of creating County Series maps ceased some time in the late nineteenth/ early twentieth centuries but I cannot state when.

 

The superimposition of a map from the County Series over one from the Nationally designated grid might well account for the maps seeming to be out of register.

I suppose, as well, that a station near the county border would be more prone to distortion than one near the centre of the county.

Given the size of Yorkshire I would suspect that this would be more apparent than it would in a small county such as Rutland.

 

Hope that this helps.

 

Ian T

 

[martin_wynne]

5 minutes ago, ianathompson said:

The superimposition of a map from the County Series over one from the Nationally designated grid might well account for the maps seeming to be out of register.

 

Note that the NLS web site uses two different projections. The "Find by place" maps are direct scans of the original County Series maps.

 

On the "Georeferenced" slippy maps, those scans have been resampled to match modern georeferenced aerial images and other maps such as OpenStreetMap, so that they fit together at the boundaries and can cross-fade to a background map or aerial image.

 

What this means is that the individual scans are often found not to match the corresponding slippy map exactly. Sometimes a station platform, say, which you know to be dead straight can appear very slightly curved on one and not the other.

 

Even the modern OS maps have to be resampled for georeferencing. If you look at the OS Landranger maps on Bing Maps, on a wide-screen monitor, you will notice that the OS grid lines are slightly curved and run across the screen at an angle.

 

cheers,

 

Martin.

[Clive Mortimore]

On 13/07/2019 at 17:26, Rowsley17D said:

I've used the National Library of Scotland's OS map facility to plan my interpretation of Buxton Midland (Derwent Spa) in Templot which allows you to put the trackwork on top of the map. 

 

 

The site also allows you to put on top the present day satellite view which lines up with what is now left - the LNWR station, Bridge Street bridge (near C.R.) and the roads in general. Where I have facing and trailing cross-overs on the main line was in fact a scissors crossover but my turnout building skills I did not think were up to it.

As Jonathan has shown the old OS maps are a wonderful resource in the research for a model railway. And using Martin's Tempolt some lovely model railways could be made.

 

As others have said they might not be 100% but in many cases they are as close as we are likely to get and can all of us fit a real location in the space we have?  When used in conjunction with other information from the period one models any errors should be noticeable and rectifiable. 

 

I often just pick a location and view the images on Old Maps thinking about the next dream layout. They are fantastic.

 

 

[Rowsley17D]

And the proof of the pudding, so to speak, is what it all looks like in model form, Clive. Helicopter shot of the developing layout as mapped above.

 

[RailWest]

I hope there is going to be a trap road (as per plan) at the exit from the sidings at bottom right....:-)

[Rowsley17D]

19 hours ago, RailWest said:

I hope there is going to be a trap road (as per plan) at the exit from the sidings at bottom right....:-)

 

Hi Chris, there is one. The photo just missed it. This is a better one with 2P on it's way to Chinley.

 

Edited July 22, 2019 by Rowsley17D