One of the things which kept me from doing the update was the rather complicated python-scripts combined with using the arcpy-library. The whole process was in it self complex. I dreaded sticking my head down into the python/arcpy/manual processing soup.
Through the last year or so I have started using FME more and more. Once you know a tool well you start thinking – what if… So I thought – what if I instead of going one more round with python and arcpy tried solving this challenge using FME?
So I made a new version…
…and it was easier said than done. But not by much. I started working on this task one evening and two evenings later I had the production line ready. The rest was restricted by my computers CPU – quite much it seems.
The former release had some shortcomings. The area calculation of level 4 squares was missing or had the wrong number. The documentation was not what it should be. The squares only covered a few countries, mostly in Africa. The Area calculation could be better. And so on. I believe I managed to fix most of them.
The QDGC grid system is relevant for visualization purposes. It can also serve as a framework for spatial analysis. There are several issues to consider when using a degree based grid cell system like this one. The varying area of the squares along the latitude is one of them. Other area adjustments could involve adjusting density analysis based on species accessibility. The latter would of course fall into considerations of the researcher(s) involved.
This and much more is covered in a paper I wrote some years ago. Read more about it here:
The new version is not crammed with new features. As a standardized product it should in theory be a more accurate version of the former one. A few Highlights of the new version
- Production line using FME
- QDGC levels 1-4 covering 175 countries
- QDGC levels 1-2 on a continental scale
- QDGC 1-5 for African continent countries
- QDGC 1-6 for Uganda, Tanzania and Ghana
- Delivered as shapefiles
Each zipped file contains a readme-file with more information about the QDGC-files. The readme gives more background on the grids.
Since the production line relies on FME it is an easy job to deliver the files using a file geodatabase or more. Any requests for formats will be considered.
One of my main achievements in this work was to be able to add python-code to the procedures within FME. In hindsight it is nor really very hard. But the documentation on the procedures from FME could be better.
The code has been made available using GitHub:
As mentioned in the beginning the processing put some demand on my computers processing powers. Churning out millions and millions of small squares placing them in shapefiles is hard work. That said it looks like FME does this faster than than the arcpy/Python combo.
I would love to make an open source only production line. For that to happen either I will have to look at the graphic modeler in QGIS, or FME will have to go open source. I am certainly not going back to python coding without using a supporting framework like FME for the programming.
This is where you will find the files:
This posting will be updated over the next week as more more levels are a calculated. Special requests with regards to level depth for smaller areas or specific countries will be considered.