Your correspondence about Ordnance Survey (OS) contours was passed to me
as an interested party, but no longer an employee (I used to be a manager
at OS and very much concerned with the accuracy of maps, an important
subject as you imply; I am also interested in climatic change and its
past and future effects). My answer is as follows - mainly off the top
of my head although I am referring to one document which I happen to have
handy. I don’t have time to research this properly just now.
1. 50 cm per century sea level rise sounds rather a lot and is unlikely
to be more than one of many opinions. Contrary suggestions are made in
(4) and (9) below.
2. Although Hillary refers to large scale maps, in the OS context
1:25000 and 1:50000 are usually referred to as small scales, the basic
(large) scales in Great Britain being 1:1250, 1:2500 or 1:10000 for
urban, rural and mountain areas respectively.
3. OS contours today are based on Ordnance Datum Newlyn (ODN). This is
defined as mean sea level determined at Newlyn in Cornwall (South west
England) during the period 1915 to 1921. Previously all OS heights were
related to tidal observations at Liverpool (NW England) in an earlier
period. OS also used a tide gauge at Dunbar (Scotland) long ago.
4. I happen to know that local mean sea level at Newlyn in 1974 has been
quoted as 0.086m above ODN (i.e indicating a rise of 8.6 centimetres in
just over half a century). I do not know the accuracy of these
determinations but it is extremely unlikely to be as good as the
millimetres, or even centimetre, implied here).
5. All OS contours (at 5m or 10m intervals) were resurveyed from 1969
onwards, on the 1:10000 scale maps, when the decision was made to
metricate. Previously contours had been surveyed at intervals
denominated in feet (generally 25 feet). The metric contours for the
whole country were completed in the 1980s, and have almost certainly by
now filtered through to all recent editions of 1:25000 and 1:50000
sheets. The height standard error of most of these modern contours (as
surveyed photogrammetrically) was of the order of 1 to 2 metres. Sea
level rise obviously represents a systematic error which at present and
over a century or so is considerably smaller than this random error in
the surveyed contours.
6. OS also surveys - and revises - the high and low watermarks of
ordinary tides (ordinary spring tides in Scotland). However it is
usually extreme meteorological events which are more relevant to threats
to the coastline.
7. You rightly indicated that the accuracy of contours can also be
discussed in terms of horizontal position. However the form of the
ground as indicated by contours will not change except in the case of
catastrophic coastal events (such as the Scarborough landslip in the
early 1990s) or man-made activity such as the re-landscaping of former
opencast mine workings. In these cases OS will always resurvey and
republish the contours within a reasonable period.
8. If one accepts that the height value of contours will always be
related to an arbitrary datum one could make an appropriate height
correction to every contour value if this were to become necessary after
some centuries of sea level change.
9. More interesting is the question of whether the land is tilting
relative to the sea level surface in which case a single correction would
not be appropriate. Students of physical geography in the 1960s were
taught that there was considerable evidence of isostatic (post ice age)
uplift and tilting of Great Britain such that sea levels were appearing
to fall, rather more in the north of the country than the south. I
believe that considerably more research is required, and that much of the
alarm over sea level rise due to global warming may be misplaced. It
will be geodetic levelling and tidal observations, not contouring, which
will assist improved understanding of these matters.
10. I would be interested to hear from the north American correspondent
of the extent to which this kind of detail - and accuracy of survey data
- is available there.
Paul Newby
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|