GIS5265 - GIS Applications for Archaeology - Module 7

This weeks lab focused on various techniques of surface interpolation. The initial assignment above, shows  the five basic techniques that we performed using ArcToolbox to perform the given study on the Barillis site, while the three examples at the bottom show the three techniques that we had ran on data for the Machalilla site in Ecuador. The latter data came in the AutoCAD format, and required manipulation of a basic text file in WIndows Excel in order to display. The coordinate system remains unknown, but I was able to get the data to comply with a few basic surface interpolation operations.

This extra poster shows yet another surface interpolation operation performed on the Machalilla site data, this time using the IDW tool. The various images show the outcome of running the tool at various power settings. It seemed to me that the tool begins to be distorted once the power is set to well over 100, but the most thorough study seems to be at the powers of 50-100. 

Breakdown of the Methodology

Despite having used five surface interpolation techniques, (all of which performed on the same archaeological sites) it is still difficult to confidently state which is right for any given project. I think it is reasonable to state that each of these surface interpolation tools could have different advantages and disadvantages in estimating a site surface on archaeological sites.

The inverse distance weighted tool (IDW) works by estimating the potential of a general area based on the known value of a given attribute. The "weight" of the color is based off the point containing most densely populated attribute in question. For example; series of positive shovel tests on a site will show as an overall positive area, with the most artifact-rich test most likely ending up around the center. This is a very useful generalization for deciding which area ought to be more tested, but when dealing with a small, isolated feature the probability can become completely distorted on the map.
The result shown after the spline tool is ran expresses a diminished surface curvature and results in a smooth overlay that passes through relevant points. I don't see this being a preferred interpolation method for finding locations of settlements in most cases. In large APEs however, the spline tool may make it easier to see consistency in distance between points.

Both the positive and negative aspects of the Natural Neighbor technique are the fact that the tool tends to generalize trends for the presence of an attribute based of the nearest input features. The positive side of this, being that the tool may overlook possible outlier tests that may have no relevance. Inversely however, it is beneficial to know where highest artifact concentrations are located as they are generally indicative of more human activity.
The Krieging method works by generalizing surface trends based off the z-value of data points in question. The degree of accuracy when dealing with low numbers tends to be quite high. The main disadvantage of the Krieging method however, is that points with high values tend to lose a bit of accuracy despite the fact that they tend to be the main points of interest in archaeology.

The Kernel Density tool takes into account the weight of all specific features (line and/or point) and shows more weight based on higher potency. This method seems to be the most advantageous for archaeologists seeking locations of sites, but the areas do seem more generalized, and weight edges seem to be less sharp than those in tools such as IDW or Natural Neighbor.

Spatial Analyst Potential

The Spatial Analyst toolset provides a great arsenal for statistical analysis within many different fields of study. In archaeology alone these tools can provide us with estimated boundaries of settlements and sites, and/or distributions of specific artifact classes. This information can be extremely useful in estimating a cost of later phases of work in CRM, and obtaining precise spatial information to clients in order to minimize impact from construction.