Annex 1B
ACCURACY AND REVISION OF PUBLIC TASK DATASETS
ACCURACY
Ordnance Survey sets and monitors standards
of accuracy for its datasets and products commensurate with the
"nominal scale" of the data.
Accuracy of geographical data may be defined
in terms of parameters relating to:
— Precision: the resolution of the
geo-reference co-ordinates of individual items contained within
the data;
— Geometric fidelity: the degree to
which the geometry of individual features within the data represent
the ground alignment and shape of the features being represented;
— Relative accuracy: the degree to
which distances between adjacent or near-adjacent features contained
within the data represent the equivalent ground distances, orientations
and physical relationships;
— Absolute accuracy: the degree to
which the geo-locations of feature(s) within the data represent
their absolute position on the surface of the earth.
Accuracy standards for the datasets and products
maintained as the Public Task will be subject to continuing dialogue
with major customers in government and business, and consumers.
Ordnance Survey's databases and products are
produced or derived from the information collected during survey
and mapping activities for the most detailed large scales datasets.
Current survey accuracy standards for these data are:
TOPOGRAPHIC DATA:
|
| Absolute accuracy
Compared with the National Grid. Absolute error—Root Mean Square Error (RMSE)
| Relative accuracy
Relative error. (Over specified distance between points taken from the map)
|
|
1:1250
(urban) | 0.5 metres
| ±0.5 metres
(60 metres)
|
1:2500 resurvey or reformed
(urban and rural)
| 1.1 metres | ±1.0 metres
(100 metres)
|
1:2500 overhaul
(urban and rural) |
2.7 metres | 1.9 metres
(200 metres)
|
1:10 000
(mountain and moor-land) |
4.1 metres | ±4.0 metres
(500 metres)
|
|
TERRAIN AND HEIGHT DATA:
|
| Digital Terrain Model
| Contours
|
|
| DTM Grid Interval
| Absolute accuracy
Absolute error—Root Mean Square Error (RMSE)
| Contour Interval | Absolute accuracy
Absolute error—Root Mean Square Error (RMSE)
|
High Resolution Data
Urban & selected floodplain areas
| 2.0 metres | ±0.5 metres
| Only supplied as
DTM | —
|
| Rural areas | 5.0 metres
| ±1.0 metres | Only supplied as
DTM
| — |
Mountain and
moor-land areas | 10.0 metres
| ±2.5 metres | Only supplied as
DTM
| — |
Standard Resolution Data
Urban areas
| 10.0 metres | ±2.5 metres
| 5.0 metres | ±1.0 metres
|
| Rural areas | 10.0 metres
| ±2.5 metres | 5.0 metres
| ±1.0 metres |
| Rural areas | 10.0 metres
| ±5.0 metres | 10.0 metres
| ±1.8 metres |
Mountain and
moor-land areas | 10.0 metres
| ±5.0 metres | 10.0 metres
| 1.8 metres |
|
REVISION
Ordnance Survey uses a combination of Continuous and Cyclic
Revision programmes to maintain the currency of the Public Task
datasets, according to the nature of the change and its importance
to users.
Continuous Revision:
— High profile developments for which user demand
will require all of the relevant information to be captured by
the date of completion/opening;
— Significant housing and industrial developments,
new communications (roads, railways etc.) and other significant
major changes to the landscape, including demolitions and changes
to addressing and naming information, which merit inclusion within
the database within 6 months of identification of the change.
Cyclic Revision:
Systematic sweeps, including the use of remote sensing techniques,
through the mapping of rural and mountain & moorland areas
at intervals of between 2 and 10 years according to the nature
of the landscape and the frequency of change, for:
— Other changes, particularly in rural and moor-land
areas to buildings, land enclosures, vegetation and other less
significant geography.
Revision policies for the datasets and products maintained
as the Public Task will be subject to continuing dialogue with
major users in the government, business and consumer arena.
| 
