Digital Spatial Technologies in Archaeology

The aim of the talk is as described in the abstract below; to cover in at least some depth the various inter-related strands pertaining to the use of spatial technologies in archaeology.

Digital Spatial Technologies have become central to modern archaeological practice. There are a number of interrelated strands to this which can be broadly categorised as capture, management, analysis/interpretation, visualisation and dissemination.

Techniques and technologies used to capture spatial data include: Total Station Theodolites (TST), Terrestrial Laser Scanners (TLS), Airborne Laser Scanners (ALS) and Global Navigation Satellite Systems (GNSS); computational photography including Reflectance Transformation Imaging (RTI), Structure from Motion (SfM) and photogrammetry.

Allied with this are tools and techniques to support management, analysis, visualisation and dissemination including more robust, ontologically driven, semantically enabled data models and Archaeological Information Systems (AIS) to handle both spatial and spatially referenced digital data and all manner of visualisations and interfaces (2D, 3D, graphs, web, portals, etc) for resource discovery, analysis and dissemination.

Digital resources are being made accessible like never before, with spatiality forming a key component, opening up new potential with platforms such as Google Earth and Spatial Data Infrastructures (SDI) for research, public access and heritage management either here, now, or becoming possible, drawing on and breathing new life into archaeological archives and indices.

All of this combines to help us as archaeologists create richer, multi-vocal, data driven narratives and theoretical frameworks and ultimately better understand the past and convey this to a wider audience. Drawing on experiences from one of the UKs largest archaeological units as well as ongoing projects across the heritage domain in which I have participated or observed, this talk aims to give a personal view on where we as a discipline are at and some ideas for where we can go next.

Highlight of the talk were tweeted by Graeme Earl whilst I was presenting and through the discussion that followed. This storify doesn’t go into everything I said and which was discussed, rather provides a bit more detail about the tweets to support the presentation on Slideshare.

Archaeological Computing Research Group (ACRG) Seminar: Digital Spatial Technologies in ArchaeologyDigital Spatial Technologies have become central to modern archaeological practice. There are a number of interrelated strands to this wh…

@pauljcripps talking in @ArchCRG Talking about spatial technologies and commercial archaeologyGraeme Earl

RT @GraemeEarl: @pauljcripps talking in @ArchCRG Talking about spatial technologies and commercial archaeologySarah May

It’s true. Since the last time I spoke to the ACRG in this context back in 2010, things have moved on a bit so whilst the general themes are pretty much the same (that’s kind of what I do from a research/professional standpoint; it was unlikely I would be talking about osteology all of a sudden).

@pauljcripps talking about http://www.pointools.com/ as tool for integrating survey data @ArchCRGGraeme Earl

@GraemeEarl @ArchCRG Also Kubit’s Pointcloud software: http://www.kubit-software.com/CAD/Products/PointCloud_Laser_Scanner_AutoCAD/laser_scanner_AutoCAD.phppauljcripps

Yep, Pointools is great. Use it for much of my visualisation and data handling work with pointclouds.
Also Kubit’s Pointcloud software which does things slightly differently and allows for the manipulation of pointclouds, imagery and CAD data all within AutoCAD. Great for digitising and producing 2D/3D vectors.

@pauljcripps discussing the potential of the stratigraphic matrix as user interface @ArchCRGGraeme Earl

This is an idea that has been kicking around since the days of Revelation, an EH project led by Sarah May. The idea being that rather than being am output of the archaeological process, as is the case with many digital archaeological information systems, we can put the matrix back at the heart of coming to understand the stratigraphic relationships on a site. Once the matrix becomes a dynamic graph of whatever we feed into it, we can start to use it to visualise different components, facets if you will, of the total knowledge base through the recorded stratigraphy.

@pauljcripps discussing need for spatial data infrastructure to minimise duplicated effort e.g. in desk based assessments @ArchCRGGraeme Earl

@pauljcripps massive need for this spatial data infrastructure in the planning process @ArchCRGGraeme Earl

Oh yes indeed. And this was apparently the most contentious bit of the talk, being compared to Kevin Costner’s Field of Dreams.It is true, it’s a bit chicken and egg: Use will come through infrastructure and there is little use case without infrastructure. However, as I have discussed previously, we waste a lot of time and money in the heritage sector just handling data. SDI approaches can help to solve many of the problems relating to access, exchange, licensing, etc.

@pauljcripps talking about gyrocopter survey of archaeological sites @ArchCRGGraeme Earl

Watch this space: this summer, I will hopefully be running a project to record vaste swathes of countryside using UAVs (fixed wing actually) fitted with RGB cameras and multi-spectral sensors. Products will include some seriously detailed otho-images and a Digital Surface Model, produced using automated photogrammetry.

@pauljcripps talking about salisbury plain graffiti recording project in collaboration with @ArchCRGGraeme Earl

Yes, one of the more random projects I’ve been involved with, albeit tangentially, but challenging nonetheless. Chantel Summerfield studies tree graffiti so Wessex Archaeology and members of the Archaeological Computing Research Group (ACRG) at the University of Southampton went up on to Salisbury Plain and scanned some of the trees for her research project.

@pauljcripps @ArchCRG @GraemeEarl Thanks for the great seminar yesterday! Tom Brughmans

@tombrughmans @ArchCRG @GraemeEarl No probs, glad it was useful pauljcripps

And the follow up…

The joys of twitter: when something is in your mind, you can just tweet…
We then went on to have a bit of a rambling discussion around all things spatial with some other lovely folks like @portableant joining the fun.

In workshop on 4D archaeological data in gloriously sunny and rather beautiful Ghent http://bit.ly/KFbrlfGraeme Earl

Wish I was there…Unfortunately, the commercial world doesn’t pay for attendance at such academic events

Interesting INSPIRE document on spatial data interoperability http://bit.ly/JBfg1v . Of relevance to your @ArchCRG seminar @pauljcrippsGraeme Earl

@GraemeEarl that is indeed an informative doc, summarises most of the key points (it’s certainly where I got lots of references from pauljcripps

This really is a very cool document. Not only does it set out a stall for SDI very clearly but it has loads of useful to pointers to more info. Anyone interested in SDI should have a look at this.

@GraemeEarl this sounds really cool. Any ideas on how to express and manage uncertainty, or only talk about it? #4dghentEleonora Gandolfi

@eleonorag1 @GraemeEarl Temporal uncertainty can be described using CIDOC CRM concepts & event based modelspauljcripps

Indeed it can and that is the subject of one of my papers for CAA 2012, which needs finishing as the deadline looms…

Interesting discussion at #4dghent spatial & textual ontologies. We need formal spatial ontologies? e.g. the form of a door @pauljcrippsGraeme Earl

@GraemeEarl indeed, there are more formal spatial ontologies emerging. Saw great talk on such at @ADS_Update a while back incl. GeoSPARQLpauljcripps

RT @pauljcripps: @GraemeEarl Possible to use existing/emerging spatial ontologies to extend core ontologies such as CIDOC CRM. Sharing/reuse/nesting = goodSarah May

Can’t find my notes on this but basically involved plugging in existing spatial ontologies to extend CIDOC CRM. Recycling and reuse is the key to this. No point reinventing the wheel and by nesting ontologies each bit can be looked at seperately, leveraging expert knowledge in key domains (eg getting the spatial technologists to develop the spatial ontologies which can then be linked to the core heritage ontology (CIDOC CRM) and inherited by specialisms (eg CRM-EH).

#4dghent Direct (GPS) vs indirect (scan) observs: primacy of multi-sensory fieldwork process given VR tools? @pauljcripps @Tom_FranklandGraeme Earl

@eleonorag1 Yes. Talking about spatial uncertainty, provenance and processing workflows at #4dghentGraeme Earl

Similarly, the event based models I discussed in the seminar handle things like provenance, assertion, inference and multivocality well.

Is there a faceted browser available for viewing GIS data? i.e. what, where (less significant) and when #4dghent @pauljcrippsGraeme Earl

Duh! moment – of course ArchSearch @ADS_Chatter integrates mapping with facets http://bit.ly/JxjTVH Examples for 3d spatial data ? #4dghentGraeme Earl

@GraemeEarl Faceted browsers I’ve seen are textual. But could use thematic mapping / dynamic querying to achieve similar…?pauljcripps

Indeed, many GIS based maps are in effect faceted by the virtue of the way in which they use thematic mapping. This is becoming a lot more explicit now as more and more UIs use the who/what/where/when type paradigms.

@GraemeEarl @pauljcripps do you mean like: http://beta.finds.org.uk/database/search/map/q/coin ? Sorry, jumping in.Daniel Pett

@portableant @GraemeEarl that’s pretty much exactly what I meant!pauljcripps

@portableant @pauljcripps Exactly But a 3d spatial data browser equivalent? #4dghentGraeme Earl

@GraemeEarl @portableant the GeoWeb3D peeps are heading that way, n’est ce pas…? http://www.geoweb3d.com/pauljcripps

The GeoWeb3D platform is really exciting. If you haven’t seen it already, go check it out. Unlike many of the globe viewer type platforms from Google and Esri, this is far more like a true GIS as we know and love, but in gorgeous 3D.

Trying http://bit.ly/Jxm23y CityGML viewer #4dghent @pauljcripps @portableantGraeme Earl

@GraemeEarl @portableant that looks cool! GML is a good solution to integrating broader ontology based text data with spatialnesspauljcripps

Another interesting platform. GML by its very nature integrates with ontological data very nicely, or can do with appropriate structure and a bit of robust semantics enforced. As such, tools like this built around GML have a lot of potential moving forwards.

@pauljcripps and the next problem… What about thesauri for represent & align arch. interventions other than http://bit.ly/KxgDvq #4dghentGraeme Earl

@GraemeEarl they are all Activity Types (ie events in the present) which have spatialness, temporality & result in data/physical objectspauljcripps

@GraemeEarl *lots* of thesauri / word lists for describing features/finds, could use some (descriptive not prescriptive) alignment though…pauljcripps

It’s a tricky one. There are different ways of approaching excavation and other forms of fieldwork. One way is to reduce each recording tradition to basic spatio-temporal building blocks, ie the context, be it a single context or a spit or a block or whatever. But there does really need to be more work in this area so when we talk about building truly interoperable datasets through ontologically mediated search tools (eg those produced by the STAR and STELLAR projects) there is semantic clarity between datasets produced through different recording traditions.

Found Time Manager #QGIS plugin http://bit.ly/K4aZ4P but only limited evidence of archaeological applications @pauljcrippsGraeme Earl

@GraemeEarl #ArcGIS supports time from v10 too. Easy to pull start/end dates from db & assoc with spatial entities for display. Nice.pauljcripps

The joy of time. Not only is GIS heading off into the 3rd dimension, it’s rapidly moving into the 4th. Being able to specify time attributes to GIS data is amazingly powerful. It’s early days yet and can be a bit limiting but the only way is up. The Esri approach for example uses a basic temporal model where a Feature has spatial + temporal extents, each one being bounded independently. As such, there is no way to represent a Feature that changes through time, it has to represented as a series of features each with a set of distinct spatial and temporal bounds.

And so, forwards…

Discussions such as these are only the start. I’m sure there was lots of amazing work presented in Ghent which I look forward to catching up on in due course.