Positional Data Shifting
The Positional Shifting of data is required when positional accuracy of some data is known to be poor, or at least not of sufficient quality for its intended use.
Techniques used to capture spatial data have improved over the years, particularly in terms of positional accuracy. This causes problems both for data suppliers who need to update their existing data in order to use more accurate capture techniques and for data consumers if their own data becomes misaligned when the base data positioning is improved. For example, utility pipes that run within a road surface polygon within the base data, may no longer run inside the road after the data has been shifted.
Similar issues occur when performing data conflation (i.e. bringing together data from multiple sources). If the positional accuracy of the data is not the same, a straight integration can result in erroneous relative positions between features.
In order to resolve the inconsistencies caused by variations in positional accuracy, the best solution is to improve the low accuracy data to match or at least sufficiently approach the accuracy of the high accuracy data.
Shifting data to improve position is more complex than simply applying the same transformation to all features as the shifts required vary across the data. 1Integrate uses Shift Vectors to optimise how data is shifted.
Note: If a point is not inside the triangulation then it will not be shifted. The convex hull of the start point shift vectors must cover all the data to be shifted. If this is not the case, then consider creating additional shift vectors outside the data as shift points outside the extent of the data will help to avoid unshifted vertices.