GIS

GIS caveats by JP

-GIS incredibly powerful tool but best when it works in the background, shouldn’t fetishize the technology

-criticized as landscape as data approach, arguably just highlights people’s lack of imagination

-most of the rest of academic world uses it as medium for representation rather than a tool for data analysis

-need good knowledge about a landscape to be able to represent it

Formal vs. Web GIS

-formal GIS: desktop packages that let you manipulate data to help answer questions with spatial dimensions

-designed to be modular, combine smaller tools to find what you need

-’Pareto’ GIS/web technologies: have 20% of the functionality they could have but what 80% of users want

-always a cost in terms of amount of data it can handle or functionality compared to desktop, but less of a learning curve

Lines of Sight/Viewshed analysis

-have a digital representation of the landscape, generate a hypothetical line of site

-both observer sight and target height problematic, unless same height cannot assume reciprocity

-tree problem: vegetation on landscape can proclude visibility, at a point in past cannot know exactly how it looks

-viewsheds: 360 degree view around taken from various points on the landscape

What is GIS?

-GIS=entry, editing, storage, query and retrieval, transformation, manipulation, analysis, and display of geospatial data

-all data in a GIS is georeferenced (ie located by means of geographical coordinates)

Layer

-how we decide to slice and dice the world

ex: streets, water features, etc

-can turn layers on and off and overlay them

Types of data in GIS

-Vector data/class/feature: points, lines, and polygons (made by connecting the dotes

-Vector data exs: counties, rivers, census data, habitat boundaries, GPS data

-Raster data/class: cells or pixels, assign each cell a numerical value that might reflect something like elevation or slope

-Raster data exs: satellite imagery, elevation, precipitation, topography

Feature

-actual object or “thing” that a map symbol refers to

-ex: river, in GIS series of connected points that all refer to the same feature

-features in a single layer can have all the same symbology, or different based on attributes

-can combine fields from two tables by joining by attribute (or by spatial location)

Attributes

-the non-spatial properties of feature

-eg name, size, etc

-typically used for symbology, labeling, selecting features

Archaeological studies of the visual landscape (Gillings and Wheatley 2020)

-’visibility’ for archaeologists property of certain locations, manifests in a network of locations, act of perception

-modes: rich description, simple mapping, formal modelling

-methodologically: functionalist approach incl defensibility in like watchtowers, others more experiential

GIS-based study of visibility (Gillings and Wheatley 2020)

-geocomputation within a GIS can calculate Line of site (LOS) between two points, useful for analysing intervisibility

-increased interest in 3D analysis, eg angle-of-view, elevation-of-view, and distance-of-view

-growing availability of large high res DEMs, prioritization of algorithms lthat are more optimised: approximate so less slow

-should be treated as probabilities rather than objective actualities

Fuzzy viewshed (Gillings and Wheatley 2020)

-comes from Fisher 1994

-in view vs. out of view false binary, things can be in the middle

-acuity: being able to see something vs. being able to recognize it, assumption of quality of eyesight

Viewsheds (Gillings and Wheatley 2020)

-delineation and mapping of potential field-of-view=viewshed

-viewshed usually calculated on DEMs and DTMs

-total viewsheds: obtain visibility of entire landscape so each cell in DEM iteratively treated as separate viewpoint

-cumulative viewsheds: how often location can see or be seen, establish if visibility patterns are statistically significant

Limits of visibility studies (Cummings 2008)

-all landscapes experienced differently by different people

-tendency to treat landscapes as visual phenomenon, ignores other senses

-GIS-specific problems: abstracted 2D cartesian views, give no real sense of what particular views look like in reality

-most effective way: to use as many different representations of landscape as are available

Visual prominence in American SW: context (Bernadini et al. 2013)

-arid landscape with vast views, mountains sacred to many Pueblo people, mark directions, homes of deities and ancestors

-many groups have migrated over time, ancestral landscapes have changed

Visual prominence in American SW: methods (Bernadini et al. 2013)

-visual landscapes of SW reconstructed using vector-based analysis of local prominence and population prominence

-skylines generated for database of >1100 Ancestral Puebloan villages in use from 1200-1700

-used Douglas-Peucker line simplification to repeatedly remove points near to trendline to retain most extreme points

-Lprom: measure visual significance from particular human observation location

-Pprom: quantification of social variables affecting viewership, multiply local prominence by number of people living at point

Visual prominence in American SW: Results and implications (Bernadini et al. 2013)

-PProm values: b/c concentrated in areas conducive to farming, many of the high ORS landforms not visible

-doesn’t necessarily mean severing attachment: Ute mountain in memory among contemporary Tewa and Towa speaking populations

-limitations of study: does not include foreground landforms, limits context of viewing to places of residence

Local visibility in the Chaco world: context (Dungan et al. 2018)

-Great houses: massive planned structures, domestic spaces but also likely played a role in local or regional religious practices

-great kivas: formal structures with a single large room serving as a religious venue

-previously GIS applied to tower kivas, large-scale cumulative viewshed analysis of possible communication networks, but not yet addressed views of great houses from landscape

Local visibility in the Chaco world: methodology (Dungan et al. 2018)

-total viewsheds created using 5-cell spacing on 30m resolution DEMs extracted from Aster satellite global DEM dataset

-each total viewshed produced with 5km buffer (max visible distance) around 4km study area (most likely utilized land)

-visibility scores for individual structures calculated as mean decile value for the 3-cell neighborhood

Local visibility in the Chaco world: results (Dungan et al. 2018)

-concluded that great house locations chosen to take advantage of visibility, less concern for kivas

-monumentality of kivas experienced from within the structure, great houses from outside

-landscape use enhanced visibility of great houses, conveyed messages to the wider audience through “Being seen”

MSA and LSA Eswatini: overview (Bader et al. 2025)

-building on legacy collections in the Lobamba Museum in Eswatini

-provide scenarios of raw material provisioning for hunter-gatherers in Eswatini over the past 40,000 years

-used Neutron Activation Analysis (NAA), able to identify primary outcrops for red jasper and green chalcedony

-used least cost path (LCP) analysis with hydrological and geomorphometric estimates of clast transport in rivers to reconstruct routes of raw materials

MSA and LSA Eswatini: GIS (Bader et al. 2025)

-conducted LCP analyses between the two raw material sources and four archaeological sites

-water flow: used Terrain Analysis Toolbox in SAGA GIS to derive topographic indices

-stream power index (SPI): capacity of flowing water to erode and transport sediment, calculated w/ volume of water and slope

MSA and LSA Eswatini: results (Bader et al. 2025)

-minimal LCP distances between two raw material sources and four arch sites range from 10 km to up to 99 km

-river sediments as secondary raw material sources only plausible for one site, confirmed with ground truthing

-distance from next possible source makes direct procurement unlikely, material may have had a high value

-most likely acquired via gift or exchange networks, may have been symbolically loaded