I’m discussing coastlines with a friend. We agree, and we do not agree. In part because I do not really know myself what a proper coastline is. Purpose is the word. -What is the purpose of the coastline, I recall us discussing. When you choose to include purpose into a discussion things could start to get fuzzy.
Is the coastline mapped with a purpose something different than the factual coastline? Hm… Well the purpose this time is to have information which later can be used for a sensitivity analysis.
Openstreetmap definition of the coastline is as follows: “The natural=coastline tag is used to mark the Mean high water spring which is the point of the highest tide. See natural=coastline for all the details. See also de:Coastline for much more details.” So there is a clear definition, but if the purpose trumps the standard suggestion…?
Despite the clear definition Openstreetmap mapping remains a highly subjective thing. It keeps humans in the loop – for better or worse. I am one of those people. I having traced thousands of features with thousand and thousand more vertexes I know that tracing features involves many decisions. The problem is that the purpuse is not as neutral and objective as I would like it to be. Humans will for better or worse continue to confuse the picture.
Some years ago I spent what must have been 20 hours tracing Tanzania coastline and mangrove areas. I thought I did well. Now I am not so sure.
If you map the coastline with the purpose of establishing a basis for describing the sensitive parts of a coastline, placing the coastline inside the mangroves could be a problem. Placing it outside might be a better choice for this purpose. Some times, regardless of your purpose, there is not just one right way of doing it.
I believe we need guidelines for this work which are strict. We can allways process ourselves around the issue. For the purpose of a sensitivity atlas we can “snap” the coastline to the outermost mangrove if necessary. Right?
The problem with our opensource community is that we are missing the FME part of it. Safe software is facilitating for fast conversions between formats – and they also add processing into the equation. I am starting to think that not having a sweet user interface as we can find in FME makes us lazy. When given having to make decisions for our mapping we bring purpose to the table because the transformations are too manual.