class: title-slide, center, middle # GIS in Archaeology ## 04 - Georeferencing ### Martin Hinz #### Institut für Archäologische Wissenschaften, Universität Bern 16/10/24 .footnote[ .right[ .tiny[ You can download a [pdf of this presentation](gis_in_archaeology_04.pdf). ] ] ] --- ## Why georeferencing information is necessary? + Many spatial informations are already available in digital form -- + Archaeology often deals with legacy data (before the 1990s) -- + Vital information about site distribution or excavations are available only as printed maps ins books -- + Even for later publications, most of the time no digital data are available -- + For reevaluation, analysis or combining with other information, it is necessary to bring them to a GIS --- ## Things to consider + Every map comes with its own projection. Sometimes known, more often not -- + Every mapping results in errors or imprecision. Every digitalisation adds errors up to that. -- + Scale matters: digitised information on large scale might be precise enough, on small scale not. It is necessary to keep meta-data to know how the data were digitised (on what scale) to be able to understand on what scale they might be used. --- # Projections + Every projection distorts some part of your map  [FlowingData](http://flowingdata.com/2014/01/13/map-projections-illustrated-with-a-face/) --- ## Things we will do today + Georeference a map -- + Digitise the information on a map --- ## Georeference a map We use the example of the Early Bronze Age burial ground of Miskovice in the Czech Republic. General location: 50°09'26.0"N 14°32'30.0"E (Degree, Minutes, Seconds) 50.157222, 14.541667 (Decimal Degree) https://goo.gl/maps/ZRNGjJZVchzGwbHS6 --- ## How the spatial information come in the publication + an overview map + more detailed map of the burials + no information on CRS are available + no reference coordinates are given   --- ## Let's Get Started 1. [Click this link](https://github.com/BernCoDALab/gia/raw/main/lectures/04/data/miskovice_overview.png) and download the 'Overview' map 2. [Click this link](https://github.com/BernCoDALab/gia/raw/main/lectures/04/data/miskovice_a-d.png) and download the detailed map 3. Open QGIS 4. Start a new project 5. Add a background layer with 'manmade features', eg. positron --- ## Locate the general area North-West of Prag + 50.157222, 14.541667 (Decimal Degree) EPSG 4326 + 1618620.5, 6472956.1 (EPSG 3857) **Check your Project CRS!** Use EPSG 3857, as this is the native projection for our background reference map (Positron).   --- ## Locate the specific mapped area + compare the overview map with the background reference map   --- ## Start georeferencing + Open the Georeferencing Tool + In there, Open the raster image    --- ## Identify and mark congruent points + Click on 'Add Point' in the Georeferencing tool + Identify in both maps points that represent the same feature + Click on that point in the raster image + In the dialog, either add known coordinates, or click on 'From map view' + Click in 'map view' on the corresponding point + Click on 'OK' + Repeat, until you have at least 5 points + **Try to spread the points over the whole image!**    --- ## Start the transformation (1) .pull-left[ + Click on 'Transfomation Settings' (the cogwheel icon) + Transformation types: + The *Linear algorithm* does not actually transform the raster. This algorithm likely won’t be sufficient if you are dealing with scanned material. + The *Helmert transformation* performs simple scaling and rotation transformations. + The *Polynomial algorithms 1-3* are among the most widely used algorithms. + The *Thin Plate Spline (TPS) algorithm* is useful when very low quality originals are being georeferenced. + The *Projective transformation* is a linear rotation and translation of coordinates. ] .pull-right[  **We use either Thin Plate Spline or Projective Transformation.** ] --- ## Start the transformation (2) + Define the Target CSR matching our input CSR (EPSG 3857). + Select the output raster file to create + Select 'Add to QGIS when ready' + Click on 'OK' + Test different Transformation algorithms   --- ## Georeferencing the detailed map + Your turn: + Open the detailed map in the georeferencing tool + Georeference it based on the already georeferenced overview map + *Tip:* You can use the corners of the map...   --- ## Result .pull-left[ + Now you should have the maps georeferenced, i.e. their orientation and extend are aligned with the reference map. + You can activate and deactivate the layer to compare + Every pixel in the map should be placed in space as best as possible + You can start digitising features of the map into spatial data (eg. as shapefile) ] .pull-right[  ] --- ## Digitising Points Now we can start recording the position of map features as spatial data. Remember the types of vector data: + Point + Line + Polygon At first, we digitise the burials as points. To do so, we add a new Shapefile layer. --- ## Adding a point layer + Click on 'Layer > Create Layer > New Shapefile Layer' + As geometry type, you can select 'Point' + Select the storage location and name of your new layer + You can add more fields (of different data tye) to the shapefile layer as needed. + (We stick to the ID field as only field for now) + Click on 'OK'.    --- ## Adding point features to a shapefile layer (1) + First, enable Editing Mode + Than, select 'Add Point Object'   --- ## Adding point features to a shapefile layer (2) + Click in the center of a burial + In the dialog, add the burial number as id + Repeat for all burials... + Click on 'Save Changes' and on 'Switch Edit Mode' to store your changes in the file.   --- ## Adding polygon features to a shapefile layer (1) + If you need to record the extend of objects, you might like to consider digitising them as polygon + Add new layer, select 'Polygon' as Geometry type + Add fields as needed, select storage location, mark name with '_polyon' or similar to know later the geometry type of that layer.   --- ## Adding polygon features to a shapefile layer (2) + First, enable Editing Mode + Than, select 'Add Polygon Object' + Left click on a point of the burial border + Add more nodes of the polygon by left clicking on further points of the burial border + Right click to finish adding points   --- ## Adding polygon features to a shapefile layer (3) + Repeat for all burials + Click on 'Save changes' and 'Switch Edit Mode' + Your new layer contains now the spatial information for all burials    --- ## Inspect the final result + You now can deactivate the georeferenced map to inspect the result on the background reference map + You can also turn on labeling to see the burial numbers + Now, you will have many different files in your output folder...    --- ## Finally + Now you can use your data for mapping and for further analysis + Keep in Mind: + Every mapping results in errors or imprecision. Every digitalisation adds errors up to that. + Scale matters: digitised information on large scale might be precise enough, on small scale not. It is necessary to keep meta-data to know how the data were digitised (on what scale) to be able to understand on what scale they might be used. + **Be as accurate as possible and neccesary** ### Homework + Select an arbitrary map from archaeological literature (individual burial ground to europe wide mapping). + Digitise the information (point or polygon) + Send me the resulting shapefile and the (georeferenced) original map. --- class: inverse, middle, center # Any questions?  .caption[Source: https://www.instagram.com/sadtopographies] .footnote[ .right[ .tiny[ You might find the course material (including the presentations) at https://github.com/BernCoDALab/gia You can see the rendered presentations at https://berncodalab.github.io/gia You can contact me at <a href="mailto:martin.hinz@unibe.ch">martin.hinz@unibe.ch</a> ] ] ]