Compiling Maps from Datasets using QGIS

Maps as images can work well to illustrate patterns and processes as well as allowing the combination of different features to illustrate concepts. This can be seen in the maps on this site, in which various datasets have been used, all of which are freely available and in the public domain, except those collected specifically by me for these maps. The maps are built up in layers which can each be ‘switched’ on or off to emphasise different views. Various data conversions and assimilations were required to enable a visually satisfying set of maps, that convey the meanings intended, to be produced.

The Maps on this site have been compiled principally from data made available in the public domain. The various datasets have been imported to QGIS and the displayed properties manipulated to make them attractive and meaningfully viewable. The core software used has been QGIS and the QGIS2Web plugin without which the work involved would have been enormous.

Without question the single biggest contribution, besides the QGIS software, to creating interactive web maps, is the application QGIS2Web developed by Tom Chadwin. See the GitHub page here (opens in a new tab)

Also incorporated into most of the maps, principally for locating known sites and identifying areas, are roads and towns.

One of the principle aims in assembling these maps initially was to increase the elevation data coverage. Ordnance Survey of Ireland produce elevation data but charge quite high prices for it, presumably because of the amount of work they represent. This map uses freely available satellite elevation data, the Japanese ALOS satellite producing the most accurate freely available data. This is produced as a series of data/image files – geotiff (Geographic Tagged Image File Format); contours are created by a program that reads the elevation data on each pixel of the image and calculates where the contour should go. This does not always align with other data, such as drainage data. In addition, the accuracy of all satellite data has a limit – in the case of the JAXA ALOS this is 30 m horizontally, but subject to reflection responses of the surface vertically. River data was obtained from two sources, neither being complete. The two do not agree and some adjustment is required. There are some erroneous features in the map as a result of these minor data inconsitencies. However, once we are aware of these, they can be corrected as necessary. Both Photogrammetric and LIDAR data are available but are expensive.

Using a GIS software also allows access to the many data sets that are made available by various projects- such as BRITICE - and allows enquiries to be made of the data, and the matching and comparison of different datasets in layers. Thus the glacial features, deglaciation landforms, Quaternary sediments, Bedrock Geology when combined with topography and drainage enhance understanding of the relationships within the landscape. This can then be combined with field observation, providing very useful and informative background and enabling the relating of visible features with underlying structure.

Manipulating Data

Some polygon manipulation was required to ensure a tight fit between the GSI coastline which is precise and accurate, and the less precise EMODNET coastline. This work was made much easier by the QGIS functions that are available.

The larger scale maps of the individual sites have incorporated GPS tracking to establish small variations in land surface height. The data was obtained using two GPS receivers operating in tandem, one positioned stationary at a point of known (mapped by OSI (Ordnance Survey of Ireland)) height, the other taken across the terrain. Both were connected to HP EliteBook8470p laptops. The point data captured in this way was corrected for temporal fluctuations in the GPS signal, using comparison between the reference height of the stationary receiver with the received height of the moving receiver, and the resulting points converted to line data.

Careful checking for errors or mistakes has been undertaken, but some may exist. Please notify the website owner of any inconsistencies or problems that you feel need to be addressed.

Datasets used include:

Ringforts have been downloaded from the archaeology.ie historic environment viewer website.

Roads - Open Street Map data - http://download.geofabrik.de/europe/ireland-and-northern-ireland-latest-free.shp.zip

River network and lake data from https://data.gov.ie/ and http://gis.epa.ie/GetData/Download.

Bathymetry data

This proved problematical due to it being of very variable relevance to what I wanted to show. EMODNet was the best source. Unfortunately some of the files were very large and were themselves a problem.

EMODNet (European Marine Observation and Data Network) bathymetry data and offshore geological data has been used for the continental shelf geology maps. A contour process was run against the RGB TIFF files to extract depth contours as a shapefile.

Bedrock and Quaternary Data

In some larger scale maps the geology has been reduced to the offshore scale of detail, i.e. grouped largely by geological periods, which was straightforward for the BGS (British Geological Survey) geology dataset of the UK. However the GSI (Geological Survey of Ireland) do not appear to have a stratigraphical lexicon available for use. Reference to the different maps and books was necessary for the onshore geology for Ireland. This was fairly time consuming and involved both correlation of formations across various publications, investigation to ensure accurate stratigraphical sequencing, and paper map scanning to capture colours for the formations and ensure consistency.

Elevation Data

Satellite data GEOTIFF images have been used to generate contours for some maps. The contour dataset is available from OSI, but at a high price. However GEOTIFF files are available to the public from satellite imagery from the GPS (US), Japanese and European satellite programs. Generating 10m contours, and in some cases 2m contours from these TIFF files has been used to determine the accuracy. In most cases the Japanese data was found to be most accurate and detailed. Even so, it has been necessary to edit the contours that were produced in some cases where the satellite data had been misrepresented for various reasons. It must be borne in mind that both the satellite image analysis processes as well as the contour creation algorithm cannot be guaranteed to 100% accuracy, and slight errors can be exaggerated by following processes.

Licensing

OSI, GSI and EPA open data is covered by Creative Commons licence (https://creativecommons.org/licenses/by/4.0/legalcode).

BGS data is made available under the Open Government Licence - Contains British Geological Survey materials ©UKRI 2020.

Data obtained from the archaeology.ie historic environment viewer webpages is covered by the following statement - 'http://creativecommons.org/licenses/by/4.0/', 'Copyright Government of Ireland. This dataset was created by National Monuments Service, Department of Culture, Heritage, Regional, Rural and Gaeltacht Affairs. This copyright material is licensed for re-use under the Creative Commons Attribution 4.0 International licence. http://creativecommons.org/licenses/by/4.0/', 'otherRestrictions'.