Now, pay attention 007… it’s another post all about technology. This week I have been learning about Geographical Information Systems, usually known as GIS for short. A GIS is just (just!) a computer programme which is specifically designed to display and organise maps and other geographical data.
GIS are incredible powerful tools for linking together all kinds of spatial data. They are very commonly used to manage and protect archaeological remains. For example, every County Council in the UK keeps a Historic Environment Register – a list of all the known archaeology in the area – to help them respond to planning applications which might damage the archaeology. These registers, which used to be made up of card indexes and OS maps, are invariably kept on a GIS nowadays. This is one of the things that a GIS will do very well. It allows you to link textual information, such as you might get in a database, with searchable, clickable maps.
They are also very useful tools for linking lots of different kinds of maps and survey data together. Like a lot of landscape projects in archaeology we have used GIS from the beginning of the project to allow us to have a single accurate overview of all our survey data. Computers also have the power to analyse your maps in ways that mere humans can’t – they can tell you what would be visible from different places in the landscape, allow you to add and remove trees or any other kind of vegetation, rise and lower sea levels. There is a great introduction to the general potential of GIS in archaeology by Mark Gillings and Dave Wheatley called Spatial Technology and Archaeology which I have put on the reading list.
So, I’ve always known all this in a general sort of way. In practice this means that I knew what a GIS would do but if I wanted it done I had to get someone else, usually Olaf or Michelle who both have better things to do with their time, to hold my hand through the process of doing it. I also know that this level of knowledge is NOT GOOD ENOUGH and have had a generalised resolution to do something about it for ages.
In the past this has meant identifying a quiet week and sitting down with a lot of internet tutorials and manuals and trying to learn one or other of the different software packages. This always ends in tears, or at least in a disgruntled resolution to go away and do something easier instead.
Part of the problem is that GIS packages are used in lots of areas apart from archaeology. GIS are very complex beasts and they can do a lot more than the kind of analysis I am interested in. The other part of the problem is that I inevitably leave the learning until I have some urgent survey data to process. Inexplicably, there is never a tutorial called ‘Getting started in GIS by turning Rick’s survey data into a digital elevation model’. They always want you to start by learning how the software works. In my fevered imagination this invariably involves adding lakes to a map of some midwestern US state, getting the software to calculate the area of the lakes and add labels to that effect. The lake thing always takes me so long to get right that most of my week has gone, I still can’t work the programme properly and I have to find some non-GIS based solution to my original query.
This time, however, as you can see in this screenshot of part of the surface of the hill as measured by Simon and I before Christmas, I have prevailed. Until last week the project GIS software was ArcGIS 10. This is made by ESRI and is probably the most powerful and widely used GIS software in the world. It is used by lots of archaeological organisations and it works very well for the sort of analysis we want to do. But, it’s expensive and even with an educational licence we couldn’t afford to have multiple copies on the university network to allow students to access and work on the data.
I have been wondering for a while about switching over to an open source GIS. Previously I had been nervous about this as I don’t speak C++ and, as you have seen, have enough problems with computers without getting involved with anything that gives me the freedom to write my own code. However, Mike has been very successfully using the open source Quantum GIS (usually known as qGIS) developed by the Open Source Geospatial Foundation for his work on the Bronze Age Barrows of the Ribble Valley.
With this encouragement I have spent the week transferring all the mapping and data across from ArcGIS 10. After the obligatory tutorial about lakes (Alaska this time), I managed to get everything up and running on my own. I have to say that in comparison to previous experiences qGIS is a dream to learn. Certainly I managed to stick with it long enough to get onto downloading and processing the data from our survey work.
Of course, now we can see the results of our contour survey, or at least part of them, we can see that there are some problems with the data we collected. In particular, we were not keeping the aerial of the GPS rover stable enough. We also need to change the settings so that the machine has longer to work out where it is before it records each point. However, these are all fixable and part of the process of learning to use the new kit. We are going out again towards the end of the month to gather a lot more data and we hope to get the problems solved then.