Using OS maps

Jon Maynard Boundaries Ltd, Boundary Demarcation and Disputes, Rights of Way, Expert Witness, Chartered Land Surveyor

BOUNDARY
PROBLEMS

An Internet resource provided by Jon Maynard Boundaries Ltd

Using Ordnance Survey maps

  You are here:    Boundary Problems  |   Investigating Boundaries  |  Using Ordnance Survey maps
OS maps generally used in boundary demarcation
Scale, accuracy and perception
What is accuracy?
Relative Accuracy table for OS maps
Currency of OS maps
Selection and Generalisation of ground features shown on OS maps
Map interpretation
Field numbers
Administrative boundaries and property boundaries
Use Ordnance Survey maps with care
  FREQUENTLY ASKED QUESTIONS
Boundary definitions: Does the OS map define my boundaries?

Boundary definitions: Can I scale from an Ordnance Survey map the true position of my boundary?

Use of OS maps in conveyancing and land registration

Land Registry title plans are based on Ordnance Survey maps. It was common practice for conveyance plans and some transfer plans to be based on Ordnance Survey maps, frequently as tracings made from the original Ordnance Survey maps.

The OS maps generally encountered in boundary disputes are:

1:2500 scale County Series maps (published between the 1840's and 1940's);

Example at right (not to scale) is an extract from a 1909 edition 1:2500 scale County Series map
1:1250 and 1:2500 scale National Grid Series maps (published between 1945 and 1990);

Example at right (not to scale) is an extract from a 1975 edition 1:2500 scale National Grid series map
1:1250 and 1:2500 scale digital maps, essentially the National Grid Series, made available as either computer readable data files (available since the late 1970's) or as plots generated by computer from data files (available since 1991). Names for these products include Land-Line, Superplan and Siteplan.

Example at right (not to scale) is a plot made from Ordnance Survey Siteplan DXF, a digital map file purchased in 2006.

 

  FREQUENTLY ASKED QUESTION  
   
Boundary definitions: Does the OS map define my boundaries? I say it does; my neighbour says it doesn't. Which of us is right?

The Ordnance Survey map is a map of the physical features encountered on the land by Ordnance Survey's surveyors. It is a map made without any enquiry as to the positions of property boundaries. It cannot therefore be a definitive map of property boundaries. Land Registry is careful to point out that its title plans, based as they are on Ordnance Survey mapping, show only the general positions of the boundaries of registered land.

Your neighbour is right.

See also http://www.ordnancesurvey.co.uk/oswebsite/site/contact/boundaries.html.
 

 

Scale, accuracy and perception

Whilst Ordnance Survey maps enjoy a worldwide reputation for accuracy, it is very unwise to attempt to measure distances from them in order to scale those distances up and to set out on the ground the theoretical position of the boundary. One reason for this is that any error you make in measuring distances on the map is magnified by the scale of the map when you set it out on the ground. So if you make an error of half a millimetre (one-fiftieth of an inch) in measuring on a 1:1250 scale map, you will produce an error of 625 mm (about 2 feet) on the ground.

Another effect of map scale is that objects can appear to be much wider on a map than they are on the ground. For example, consider a thin line representing a fence:

  • Most Land Registry title plans are produced at 1:1250 scale. The thickness of the line on the map actually represents 10 inches (250 mm) on the ground.
  • Other title plans are either produced at 1:2500 scale (rural areas) or are enlarged to 1:1250 from 1:2500 scale. On both of these the thickness of the line on the map actually represents 20 inches (500 mm) on the ground.

Compare those line thicknesses with the average larch-lap panel fence, supported on fence posts only 75 mms (3 inches) thick!

 

What is accuracy?

More important than this, you have to consider the accuracy of the map you are using. Accuracy can mean a number of different things:

  • absolute accuracy: is a feature shown in its true position on the surface of the earth?
  • relative accuracy: is the position of any feature on the map shown in its true relationship to other nearby features?
  • currency: does the map show all of the things that have been built, and exclude all of the things that have been demolished, at the date the map claims to portray (the publication date)?
  • selection: does the scale of the map require certain features to be omitted for the sake of clarity?
  • generalisation: does the map scale require certain features to be shown out of their correct position for the sake of clarity?

The absolute accuracy of a map is not an issue in boundary disputes. This is because a high level of absolute accuracy assures us that the map shows a Hertfordshire mansion in its correct position relative to the owner's holiday cottage in Cornwall. With boundary disputes we are concerned only with the Hertfordshire mansion and the immediately adjacent land (or perhaps with the Cornish cottage and its immediate neighbour).

Relative accuracy is a major concern in boundary disputes. Is the fence shown in its correct position relative to the houses either side of it? Ordnance Survey has over many years tested both the absolute and the relative accuracy of a large sample of their published National Grid Series of maps. As a result, in about 1997 they published an accuracy statement for their large scale maps. The following is an extract of what Ordnance Survey says about the relative accuracy of its large scales maps.

Achieved relative accuracy of Ordnance Survey map data.

The following table shows the expected relative accuracy values for well defined points within each accuracy category. The values apply up to the stated maximum measured distances quoted in the table.

Table 3 - Relative Accuracy
 
Scale and method of original survey
Expected Relative Accuracy at differing confidence levels
Maximum measured distance
68%
95%
99%
1:1250 scale Maps surveyed at 1:1250 scale using instrumental methods to provide a framework of controlling detail.
±0.4m
±0.8m
±1.0m
60.0m
1:2500 scale resurvey/ reformed Maps surveyed at 1:2500 scale using instrumental methods to provide a framework of controlling detail.
±0.9m
±1.8m
±2.3m
100.0m
1:2500 scale overhaul Maps originally recompiled from pre-1946 County Series mapping.
±1.2m
±2.3m
±3.0m
200.0m
1:10,000 scale * Maps surveyed at 1:10000 scale using instrumental methods to provide a framework of controlling detail.
±3.5m
±6.7m
±8.8m
500.0m
 
* Some generalisation of detail does occur for cartographic reasons on 1:10,000 mapping therefore some points of detail may appear to be less accurate than these standards.

 

The accuracy statement will tell you that even in the best of circumstances (a 1:1250 scale map), whilst some of the lines will be spot on, you can expect the position of any line on the map to be shown up to 400 mm (16 inches) out of position relative to other lines within 60 metres (185 feet) of it. For a "1:2500 scale Overhaul" map the comparable relative accuracy of the lines on the map is 1.2 metres (4 feet) relative to other lines within 200 metres (656 feet). I have come across a few lines on some maps that are much further out of position than this.

 

  FREQUENTLY ASKED QUESTION  
   
Boundary definitions: Can I scale from an Ordnance Survey map the true position of my boundary?

No. Just take a look at Ordnance Survey's Relative Accuracy table, above, and judge for yourself.

See also the "Example of confusion arising from map generalisation", below.

See also the Frequently Asked Question (Acacia Road) on the Title Plans page.
 

 

In recent years Ordnance Survey has undertaken a Positional Accuracy Improvement programme. The purpose of this is to make their maps more compatible with the positions obtained from GPS equipment. The main effect of this is to move the grid lines relative to the rest of the map detail, with little effect on the relative accuracy of the map. If you are working from a conveyance plan that is based upon, say, a 1970 edition OS map then you are not helped - and may be a little puzzled by the grid lines - if you compare that conveyance plan with a title plan based on a recent, positional accuracy improved, OS map.

Remember also that the relative accuracy of County Series 1:2500 scale maps will be worse than that quoted for the National Grid 1:2500 scale overhaul maps in the table above.

Currency issues relate to whether we can trust the OS map to show all of the physical features (relevant to the map specification) that were on the ground at the publication date and to not show any features that had already been demolished or otherwise removed from the ground.

County Series maps were “periodically” revised. In other words, when it was decided that the map was out of date it was brought fully up to date and republished. With periodic revision one should have confidence that what was shown on the map was actually on the ground at the date of publication. If you have, say, the 1911 edition of a County Series sheet then you can be reasonably sure that everything that was on the ground and should have been mapped is shown on that map, and that any demolished buildings have been removed from it.

National Grid maps present a more confusing scenario. These were “continuously revised” (although it would have been grammatically more correct to call them continually revised) and made available to the public as “first editions”, “interim editions” and “fully revised” editions.

  • First editions (distinguished by an edition letter that was always “A”) could be relied on for currency to the same extent as a County Series map.
  • Fully revised editions (distinguished by an edition letter that is alphabetically later than A) could also be relied on for currency to the same extent as a County Series map.
  • Interim editions (distinguished by an edition letter that is suffixed by a number of bars and/or stars) contain the latest available survey information. This means that the map might contain revision in one street but omit new ground detail resulting from either construction or demolition in an adjacent street. Interim editions cannot therefore be fully relied upon for the currency of the information they contain.

Digital maps should be treated for currency in the same way as interim editions of National Grid maps. Since the cessation of printed 'new editions' in about 1990 Ordnance Survey's large scale maps may show the "latest available survey information" (and this might include revision added only 48 hours previously), which is not the same as saying that the map is up to date in all respects. Ordnance Survey has made strenuous efforts to ensure that significant ground detail finds its way onto its maps within a specified number of months. However, their revision programme is driven by an intelligence system, and new ground detail gets added to a map only if the local surveyor becomes aware that it has appeared on the ground. It is perfectly possible for a surveyor to map a new house built as infill development in one street, and miss another house in an adjacent street.

If you are trying to use the presence or absence of a feature on a particular version of a map to prove the age of a ground feature (such as a fence) then you have to be absolutely certain that the map was current (ie. complete and up to date) at its publication date.

Selection: there is not sufficient space, even at the relatively large scales of 1:1250 and 1:2500, to show on the map everything that occurs on the ground. Examples of selection include:

  • where a hedge and ditch run alongside each other it is the hedge that is shown and the ditch that is omitted;
  • where a hedge separating two fields has a fence on either side (presumably to isolate livestock in one field from livestock in the other field) then the hedge will be shown and the fences omitted;
  • buildings smaller than 8 square meters in extent are usually omitted;
  • bay windows and porches are omitted.

Generalisation: Sometimes features that are too important to omit from the map stand too close together to be shown in their correct positions on the map. The rule is that parallel features either have to be shown at a minimum distance from each other (1 metre on 1:1250 scale maps and 2 metres on 1:2500 scale maps) or they have to be merged into a single feature. For example:

  • where both features are too important to omit, then it may be necessary to displace one of them in order to maintain a minimum distance (for reasons of clarity and tidiness) on the map: thus a hedge and a stream or drain less than 2 metres apart may be shown 2 metres apart on a 1;2500 scale map;
  • two detached houses with a gap between them of between 1 and 2 metres will be drawn at 2 metres apart on a 1:2500 scale map, whilst two detached houses with a gap of less than 1 metre between them will drawn as if they were semi-detached houses on a 1:2500 scale map.
  • on a flight of steps, individual steps will be drawn at 1 m horizontal spacing on a 1:1250 scale map, even though the step treads will be narrower than this.

Example of confusion arising from map generalisation

On the basis of Fig. 1, an Ordnance Survey County Series 1:2500 scale map, the owner of 'A' told the owner of 'B' that their common boundary ran from the north-east corner of B's house perpendicular to the front hedge (at right). A claimed that some of his land had been taken into B's garden, and A wanted it back.

Fig. 2 results from an accurate survey. It shows that more than 0.5 metre (nearly 2 ft) separates the two houses that Ordnance Survey has shown joined together. The survey also reveals that the long-established hedge between the two front gardens (right hand side of Fig. 2) had been drawn by Ordnance Survey at about 1.3 m (4 ft 6 in) to the south of its true position. Both the closing of the gap between the two buildings and the mis-positioning of the hedge, whilst misleading to the naïve map reader, are within specification for a County Series 1:2500 scale map.

Fig. 3 overlays the map onto the accurate survey to illustrate the relationship between the Ordnance Survey map and reality.

Incidentally, it did not occur to the owner of 'A' that if he followed through with his own logic then he would have to voluntarily give up part of his rear garden to the owner of 'B'.

 

Map interpretation:

Certain line features shown on OS large scales maps may be readily interpreted according to context:

  • Buildings are usually shown by a solid lines forming an outline that is filled with stipple (black dots);
  • Railways are evident from the smooth and parallel pairs of closely spaced lines representing the two rails of each line of track;
  • Water courses may have text or symbols associated with the lines that represent their centres or their two banks, such as the name of a river, the annotation “stream”, a nearby arrow indicating the direction of flow, or the annotations “issues” or “sinks” where the line starts or ends.

A convention that assists interpretation is the one by which lines that represent features that would be an impediment to a pedestrian are shown solid, whilst those that are not are shown dashed. Dashed lines are used to show features that rise or fall by, or that represent a change of ground surface level of, less than 0.3 m (1 ft). Thus:

  • a fence, hedge, wall, bank or ditch will be shown by a solid line:
  • road edge kerbing, and a change of surface (with no associated change of level) from a paved footpath to a grass surface, will be shown by a dashed line;
  • overhead features, and underground features are also shown dashed.

Whilst the foregoing may seem straightforward, complications can arise because published (as opposed to digital) Ordnance Survey maps use the same line symbol for a wall, fence, hedge, bank, ditch and stream, making it sometimes difficult to know just what the line represents.

Digital map feature coding (also known as "layering") can add information that was not available on printed monochrome maps. If you have CAD (computer aided design) software or a GIS (geographic information system) then you will be able to access the feature code that is attached to every line, symbol or text string that makes up the digital map. In fact, it is these feature codes that allow CAD and GIS systems to display line maps with different colours to represent different feature types.

Above: monochrome plot, not to scale, from Siteplan DXF


Right: the same extract, at larger scale and in colour, with feature codes (layers) identified.

In the plot at above right the following codes have been used to differentiate between solid lines that would have been indistinguishable in a printed monochrome map:

  • G8030001 Building outlines, internal building divisions - for buildings of at least 50 sq m
  • G8030015 Railway (standard gauge) - each rail is shown by a separate line
  • G8030030 General line or minor building detail - fence, hedge, wall, antiquity above ground level, aqueduct, pier, viaduct, etc - in the above example this code has been applied to fences, walls, a minor building and the barrier at a level crossing
  • G8030059 Water detail - bank of wider drain, canal, stream or river; centre line of narrow drain or stream; basin, dock, lock or moat (water filled); Shore of lake, loch, pond or reservoir

and the following codes have been used to differentiate between dashed lines:

  • G8030021 Road (public) edge of metalling - the limit of public road surface where not defined by a solid feature such as a building
  • G8030031 Closing links (inferred detail) - This layer does not represent a real-world feature. Closing links (or inferred links) are usedto form a sensible division of property. It is not a legal representation of boundaries.
  • G8030032 General ground level or minor overhead detail - This layer includes many linear aspects of the landscape that do not present an obstacle to a pedestrian - in the above example it is used to show the back edge of the footpath beside the public road and the edge of the road carriageway as it crosses the rails at a level crossing.
  • G8030036 Vegetation or landform limit.

To understand the layering system and the way it relates to the terrain encountered in the real world you should consult Ordnance Survey's OS Sitemap User Guide and Technical Specification which can be found at
http://www.ordnancesurvey.co.uk/oswebsite/products/ossitemap/pdf/OS%20Sitemap_userguide.pdf.

The depiction of vegetation requires comment.

  • Where a continuous area of ground cover (usually vegetation but sometimes marsh, rock or sand) occurs it is signified by a symbol.
  • The limits of an area of vegetation may be shown as coincident with other physical features such as fences or road edges. Where the limit of vegetation cover is unfenced, the approximate limit will be shown as a dashed line.
  • Hedges are without exception shown by a single, solid line that represents the centre, or root line, of the hedge.
  • First edition County Series maps identified hedgerow types by the addition of tree symbols.
  • Avenues of trees are shown by a series of tree symbols placed for cartographic effect – no attempt is made to place one tree symbol for each tree that exists on the ground.
  • Landmark trees are shown by a tree symbol with a dot beneath it to represent the actual position of the tree. Examples of such positioned trees are rare on OS maps: they are most usually found where the tree is used as a marker along a county boundary.

 

Field numbers

Field numbers are often used as a tool to identify individual fields when conveyancing rural land. Field numbers (also known as parcel numbers) derive from the requirement that was placed on Ordnance Survey to measure the area of every county, borough, district, ward and parish in England, and their equivalents in Wales and Scotland. Ordnance Survey's method of operating was to measure, on the face of the map, the area of every field or other convenient parcel of land and to sum the areas of all the land parcels that made up each parish.

Field numbers were published only on 1:2500 scale maps from the County and the National Grid series of maps: they were dropped from the digital maps that succeeded the National Grid series after 1990. Field numbers appear in one of two formats. On County Series maps the fields were numbered sequentially within the parish, starting at 1 and running into the hundreds. The sequence was then repeated in the next parish.

National Grid maps adopted the convention of identifying a point roughly in the centre of the field and then naming it according to an abbreviated version of its 10 m grid reference. The kilometres portion of the grid reference may be inferred from the number of the map sheet on which the field is located. The first two digits of the field number represent the hundreds and tens of metres of easting, and the third and fourth digits represent the hundreds and tens of metres of northing.

The interest in field numbers, from a boundary demarcation perspective, comes not from the areas values that are attached to them, but from the inference as to the location for the boundary that may be implied by identifying the field using its Ordnance Survey field number.

Imagine that a field has been sold and that it has been identified by reference to its field number shown on the Ordnance Survey map. On three sides the field is unambiguously bounded by a fence. On the fourth side the field is bounded by a hedge, and beyond this hedge there is an associated ditch. At a later date a boundary dispute arises between the then owner of the field and his neighbour who owns the land beyond the hedge and ditch. You may think that the hedge and ditch presumption applies, and there is plenty of case law to support that view. But there is also case law that holds that the boundary must follow the centre, or root line, of the hedge. The reasoning behind this case law is that the field number specifically relates to the line that bounds the measured field, and that line represents the centre of the hedge and excludes the ditch beyond.

 

Administrative boundaries and property boundaries

The Ordnance Survey Act 1841 instructed

"that the Boundaries of the several Counties in England and Scotland, and of Berwick upon Tweed and of the Isle of Man, should be ascertained and marked out".

A book, now out of print, that is the modern authority on administrative boundaries is J R S Booth's "Public Boundaries and Ordnance Survey 1840 - 1980", Ordnance Survey, Southampton 1980. In it, Booth tells us that

"Ordnance Survey does not concern itself with property boundary as such and no such boundary is marked on plans. However, much public boundary was defined by the property boundary existing at the time of the order making new boundary".

In amplification of the process Booth writes:

"The O.S. does not concern itself with property right further than is necessary to obtain the mereing for a public boundary. It frequently happens that an administrative boundary follows a feature which defines a boundary between privately owned properties. Where this is the case it is necessary to ascertain a 'property right' in order to position the public boundary which for reasons of administrative convenience should be made to coincide with the private property boundary."

Booth then states, seemingly illogically,

"Public boundaries shown on OS maps are not evidence of the position of private property limits"

and he cites Section 12 of the Ordnance Survey Act 1841 as his authority for this statement.

In another book, also out of print, J B Harley's "Ordnance Survey Maps: a descriptive manual", Ordnance Survey, Southampton, 1975, we learn:

"The legal status of boundaries on Ordnance Survey maps is a matter of practical importance. The basic position is that the publication of such administrative boundaries on various map series has never invested them with any common law or statutory significance in evidence ... The development of case law * has, however, altered the situation ... and the 1:1250 and 1:2500 maps carry a prima facie indication to a Court of Law that a boundary existed at the map position shown, at the time of survey or revision."
  *     Especially Fisher v Winch 2 All ER 144 (Ct. of App. 1939 1 KB 666) and Davey v Harrow Corporation, 1957 2 All ER 305.

Harley goes on to explain the process.

"Because [administrative] boundaries are invisible and cannot be surveyed by direct methods, their precise location in relation to visible ground features is recorded by perambulating the boundary line and 'mereing' it to those features ... The term mereing has also been extended to apply to the written statement indicating the precise relationship of a boundary to the adjacent detail (for example, 4 ft RH = 4 ft from root of hedge)."

It follows that where a dispute arises as to the true position of a property boundary that has also been used to define an administrative boundary, the mereing recorded on the Ordnance Survey map for the administrative boundary is also evidence as to the position of the property boundary at the date at which the administrative boundary order was made. For convenience this is taken as the date of publication or of revision of a published County Series or National Grid map, but the date may be less easy to establish for a post-1990 digital map.

And what are the mereings that may apply as evidence of the position of a property boundary?

CH = Centre of Hedge
0.91m RH = 0.91m from Root of Hedge
CW = Centre of Wall
FF = Face of Fence
FW = Face of Wall
BB - Base of Bank
TB = Top of Bank
EK = Edge of Kerb
A range of distances from a linnear feature such as a Root of Hedge or Top of Bank may be applied, 0.91 m (3 ft), 1.22m (4 ft) and 1.52 m (5 ft) being the most common.

 

Use Ordnance Survey maps with care

The Ordnance Survey Act 1841 does not grant Ordnance Survey the power to ascertain or alter property boundaries. On the face of it, Ordnance Survey maps should form no part of an investigation into the position of a property boundary. However, many conveyance plans are based upon Ordnance Survey maps, as are all of Land Registry's title plans. Moreover, old Ordnance Survey maps are a primary source of information on the physical features that existed at various dates in the past. Further, case law allows the use of Ordnance Survey maps as evidence. All things considered, it is inevitable that Ordnance Survey maps will be used during any investigation into the position of a property boundary.

When using Ordnance Survey maps, exercise the greatest of care!

Correct interpretation of the map can only be achieved by taking the map onto the site and comparing it with the features on the ground to decide:

  • what has been shown on the map;
  • what has been omitted from it;
  • what has been deliberately moved out of its correct position for the sake of clarity;
  • and what has been shown in the wrong position because of inaccuracies in the survey that produced the map.

A correct interpretation of the map usually concludes that the line on the map identifies the physical feature to which the boundary is related, and it is the position of the actual feature on the ground - not the position of the line on the map - that determines the position of the boundary that follows the feature. Even then, you must first make sure that the legal boundary does actually follow the physical feature, and that you are not dealing with a physical feature that merely represents a general boundary that relates somehow to a nearby legal boundary.

 

  Return to Top of Page