Dreaming of total viewsheds on Caribbean beaches (new open access publication)

Today our new open access paper was published in Journal of Archaeological Science, introducing the concept of Visual Neighbourhood Configurations for total viewsheds. The method was conceived during a walk along the beach on the Caribbean island of Guadeloupe (I know what you’re thinking, my work is hard). From the beach it was really hard to spot boats at any great distance, let alone other islands. The beach I was walking on was very narrow and a plateau with steep edges rose up close-by. As I climbed higher and higher up the steep slopes I was offered great views of the surrounding sea and islands, “a great spot for the island’s former inhabitants to keep an eye on approaching enemies or on people using the beach” I thought. But it was also a very exposed spot, not many places to hide from approaching enemies. As I reached the plateau top the views changed dramatically. Walk just a few meters away from the plateau’s edges and you’re hidden from view to the people below and at sea. This place offered quite specific visual properties: past inhabitants had great vantage points nearby to see canoes and people approaching, whilst having their settlement hidden from view just meters away from the plateau edge. Could this explain the location of the few prehistoric settlements sitting right by the edge of the plateau?

DSC00095
Views from the plateau edge of La Désirade in Guadeloupe.

This is a phenomenon that concerns the visual properties of a small area rather than the properties of specific locations. When we explored the literature we could not find any methods for formally expressing such theories: all visibility studies in archaeology concern formal treatments of the visual properties of locations, with the properties of small areas being evaluated qualitatively. So there was some work for us to do: invent a new formal GIS method to test these archaeological theories. The first step was easy: create total viewsheds of landscapes.

Total viewsheds are awesome! They are representations of how visible each location in a landscape is from all other places. Archaeologists use them for exploring a wide range of theories about how the things past people could see affected how they behaved. But in the past it was basically impossible for me to play with total viewsheds because they involve many calculations that take a really long time. Until recently that is: present-day personal computers are so powerful that computation time is basically not prohibitive anymore. But there’s another problem: because of the limited use of total viewsheds in archaeology there are very few formal methods for me to play with. Developing new methods to work with total viewsheds is now more than ever worth our while.

Together with Mereke van Garderen and Mark Gillings, I developed the Visual Neighbourhood Configurations method to address the very limited practice of archaeologists formally expressing and testing their theories about how visibility patterns are structured. The concept recognises that such theories commonly concern not just the visibility of point locations but rather the structure of visibility in an area around focal locations. Visual Neighbourhood Configurations allow you to formally express your theory about how visibility is structured in a small area and compare it against the actual visibility as represented by total viewsheds to test the theory.

The below graphical abstract produced by Mereke van Garderen explains the method a bit better: a total viewshed is created, a visual neighbourhood configuration is formulated representing a visibility hypothesis, the fit between this configuration and each cell in the total viewshed is evaluated and represented as a new raster.

Want to know more? Read the full open access paper here.

Screen Shot 2018-05-29 at 12.51.52
Graphical representation of the VNC method. Figure by Mereke van Garderen.

Highlights of the paper

    A new approach is presented for the formal representation and evaluation of complex visibility theories.
    Visual Neighbourhood Configurations (VNC) represent a theorized distribution of visual properties in a small area.
    Total viewsheds are input to the approach and are formally compared against the VNC representing the archaeological theory.
    A software tool has been developed to implement VNCs with a wide range of analytical techniques.
    VNCs represent a step towards more complex theoretical formal visibility studies.
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at WordPress.com.

Up ↑

%d bloggers like this: