This is just a brief update to outline some of the latest developments to the project’s GIS element.
Web mapping: constructing an Open Source geostack
John, Xin and I have been thinking this month about how we might go about constructing the web-based mapping element seen as a vital component of the final website that we will produce as part of the project. Xin constructed a good initial plan and then I found a very useful tutorial and software package on OpenGeo. This OpenGeo Suite is well-documented and seems to function well as a route to provide webpage-embedded interactive maps.
The team installed the Suite on the OeRC’s server and used it to feed data to a test webpage (written based upon the OpenLayers API). The results were very promising and we hope to use this software / process as part of the construction of our final website output, where appropriate. What data and results the final website will map is still very much an open question!
Converting Ordnance Survey grid references
Whilst looking at and importing AIP data for field systems into the project’s GIS database, I wrote a small piece of Python code to convert Ordnance Survey National Grid References (NGRs) to numeric x and y coordinates for GIS implementation (in metres, not latitude / longitude). The code should be available for download here. It runs in Python (2.7) and would need expanding to include the capture of NGRs from a source file and the writing of output x y coordinates to a results file, as it is only really of use for bulk conversion purposes. Apologies for the quality of my coding…
My previous (brief) online guide to this subject can be found here. Incidentally, my main tip for doing this manually is: make sure you do not confuse your easting and your northing!
Processing NMP data for GIS analysis
I have also been working on processing the NMP data for the south west and south east received from English Heritage.
It was found to be useful to copy the raster layers into a Raster Catalog object in ArcGIS. This makes possible various analytical methods that would otherwise be closed to raster data (such as the Select by Location tool), and also makes it easier to handle multiple tiles in one batch (including controlling the display of blank vector tiles at inappropriate scales for raster rendering, e.g. when zoomed out too far to show any detail). The Raster Catalog will be updated as further data is received.
With the vector data (in CAD format), it was found to be useful to convert the files into a single Geodatabase (.gdb) object in ArcGIS. This, again, makes it easier to handle multiple tiles at once (including maintaining symbology across tiles) and also makes it easier to output composites of multiple tiles to other formats (such as shapefiles). Again, this Geodatabase will be updated as further data is received.
The results look good and should form a strong basis for the research undertaken by the project (in conjunction with our other data sources, of course).