15.6.6. Accuracy ( or Geometric Accuracy)

There are three measures of accuracy that can be applied to a mapped feature:

• Accuracy of Shape (Geometric Fidelity)
• Relative Accuracy
• Absolute (Positional) Accuracy

15.6.6 1. Accuracy of Shape

The accuracy of feature representation depends on the scale of capture; its associated limitations and if any degree of generalisation has been used. The principle of accuracy of shape is 'that any real world alignment must be accurately replicated in Land-Line', for example:

• Features that are square, rectangular or straight in the real world must be square, rectangular or straight in Land-Line;
• Adjacent features must be correctly separated, orientated and aligned to each other;
• Subject to established tolerances and the limitations of the source survey scale, there should be no discrepancy between the measured and scaled distances between any two (well-defined) adjacent features.
  

15.6.6 2. Relative Accuracy

Relative accuracy is a measure of the positional accuracy of a data point in relation to other local points of detail. OS is continually sampling data and testing the accuracy of well-defined points of detail. Random samples are taken in different geographical areas for the various source survey scales and survey methods.

For current relative accuracy levels see Ordnance Survey Landline Users Guide Page 165 at http://www.ordnancesurvey.co.uk/oswebsite/products/landline/pdf/lluserguide.pdf

15.6.6 3. Absolute Accuracy

This is a measure of the positional accuracy of a point of detail compared to its 'true' national grid position and is determined by precise surveying techniques.

For current absolute accuracy levels see Ordnance Survey Landline Users Guide Page 166 at http://www.ordnancesurvey.co.uk/oswebsite/products/landline/pdf/lluserguide.pdf

15.6.7 OS Data Capture Flow Lines (and typical timings)