Perimeter Precision

[JakobMiksch]

We computed the perimeter of all lakes in Fennoscandia. However, the perimeter can vary a lot depending on how detailed the underlying polygon is. Since we have lakes from different kinds of sources, we should check if the level of detail is comparable across all of the lakes.
Otherwise there might be a danger for wrong data.

[ninsbl]

Good point.
When working on different countries, different data sources and different levels of detail are just a reality.
What can be done is to count the number of vertices and relate that number to the perimeter. That way we get an impression of the precision of the digitization.
If the precision effect is significant would be another question (e.g. how much does perimeter vary between really complex and simple lake geometries, and between lake geometries with different digitization scale....)

[ninsbl]

This has been added to a set of geometry metrics which are computed on N16 at the moment.
Should be fixed soon!