Wednesday 29 July 2009

Coincidence

I've been working on getting the Yorkshire coastline up to scratch for a while. I've been gathering details from various sources and so on a very rainy day I thought I'd spend an hour trying to pull it together. I wanted to separate the existing coastline from the administrative boundary. The boundary had been added to the coastline as a starting point, but this is actually not right. The boundary of English counties generally lies along the mean low water mark - a completely impossible mark to discern by a visit. The best data we have that we can use is NPE, which shows the mean low water mark and high water mark for much of the coast.

On many coastlines this will be very good - easy to trace either in Potlatch or JOSM and accurate. On the Holderness coast it is easy to trace, but the accuracy can be poor because the coast has eroded so quickly that it could easily be 50 metres out and maybe more than 100 metres out in places.

I started untangling the relation for the county boundary and English region from the coastline and adding a way for the administrative boundary from the south northwards up the coast then adding the relations to the new way. Unbeknown to me Mickey (Warofdreams) was adding the beaches down the coast, also from NPE. The result is interesting and good, although it is still rendering.

The coastline way should be on the high water mark, which in many places it is not. This is leaving a gap showing water between the coast and the beach. Now the coastline is not tied to the boundary I'll have a go to nudging it into place to sort this out. It would be good to sort the coastline into its proper position before the next coastline update for Mapnik which only happens occasionally.

No comments: