Lab 2

Geospatial Archaeology: Lab Assignment 2

Archaeologists often manage their data in table format. Whether it’s tables of coordinates from excavations, inventories of artifacts in museums, or results from lab analysis. We are endlessly linking and rearranging tables. As we saw in In-Class Ex 2 and 3 there are various strategies for linking tables with spatial entities. The most convenient spatial entity is the geographical Point feature because it’s discrete and can link directly to records in other tables.

Communities living in the Callalli valley about a half-day’s walk from the Chivay obsidian source, turned to an economy emphasizing herding and high-altitude farming (tubers and quinoa) by around 500BC and increasingly depended on this subsistence economy in the millennia that followed. Obsidian was used by the earliest foragers but did it continue to be used by later agro-pastoralists?

Before beginning, download a fresh copy of the Callalli geodatabase Download a fresh copy of the Callalli geodatabase and make sure your catalog is connected to the Callalli folder.

In a previous In-Class exercise you learned to use the Pivot Tables tool to reduce the complexity of tables and join them back to spatial references. While we Pivoted on SiteID previously to link these artifacts to space at the coarser “SiteID” scale let’s Pivot on ArchID this time so that we maintain our spatial resolution. Recall that all collections get an ArchID (the primary key) and some of these are inside of sites which are indicated by the SiteID.

Redo the Pivot Table on Lithics_Lab1 as described in In-Class Ex 2 pivoting on ArchID and use Material and Peso (weight) as the quantitative value. Save the output to the Geodatabase as “Lith1pivot_ArchID_mat_wt”.
Open the Attribute table. For symbolizing it would be useful to have the total weight as well. Attribute Table Menu > Add Field… create a field called “Total_Wt” and make it Float type.
Before proceeding to the next step, start an editing session or your changes will be permanent!
add two columns in Field Calculator. Now right click the new field header and choose “Field Calculator…” and add the Not_Obsidian and Obsidian columns together into the Total_Wt field. Spot check the results to make sure it worked as expected. Close the attribute table.
Symbolizing the lithics: first, stop the editing session! We’d like to map these results Right click All_ArchID and Copy the layer. Now choose Edit menu > Paste.
Click the title of the new Duplicated layer and rename it Lith1_Wt. Check it on so you can see it.
Join the non-spatial lithics pivot table to the new spatial point layer. Join “Lith1pivot_ArchID_mat_wt” to the Point geometry feature “Lith1_Wt” using ArchID as the Key Field in both tables and only keeping Matching Records. Remember to start with the spatial side.

Open the attribute table on the result. Scroll over and note, does the field show Weight of obsidian in each ArchID? Note that some have 0 obsidian – these must have “Not Obsidian” lithic weight because recall that this table was previously subset to only show the records that have ArchIDs that appear in the Lithics_Lab.

Question: How far is obsidian from the River?

Construct the Euclidean Distance measure on Hydrology and use Zonal Query with Table (Statistical Type: Mean) to get the mean distance of every ArchID location in the survey.
Repeat the steps from the previous In-Class Exercise where a Euclidean distance raster is produced from Hydro_lines100k using a 30m cell resolution.
Next we'll query against this cost raster using the "Zonal Statistics as a Table" tool in Spatial Analyst Toolbar > Zonal.

this is that query tool that, in a sense, looks at the ground beneath each Point location and asks "what cell of the raster am I standing on?" and attaches that value to the attribute table for that point.
Query for the MEAN statistic (the others useful when the input vector is a Polygon or Line feature).

Join the Zonal Query output to the Lith1_Wt by ArchID.
Open the Obsidian table. With the “Keep matching records” join we did earlier, these joins can get messy (the cardinal information has errors) and you can get strange results. On top of that, we have an additional join to the Distance from Hydrology table. At this point it’s a good idea to export this into it’s own Feature so that it is simpler.
From the Table of Contents choose Export Data… > Export > Save as Feature in the geodatabase. Call it Lith1_pivot_disth20.
- Add the new layer (the pivot table you just made) and rename it Obs H20
- Copy this layer and Edit menu > paste and rename the new one Non-Obs H20
Right click Obsidian H20 the layer > Properties > Definition Query. Query for only those records whose values match Obsidian > 0.
Do the same for the other layer but where Non-Obsidian > 0. Call it "Not_Obs_H20" and note the total number of objects is 106.

Definition Queries provide a powerful way to organize and present your data because you're creating different views on the same collection of Features. You haven't had to spin off a lot of little files (a pain to manage) for each view. You still have one Lithics - Weight - Distance to Water but you're displaying different subsets of the contents.

Derive a Ceramic Period pivot Table: Pivot the Ceram2_Lab table using ArchID as the Input Field, Period_gen as the Pivot field and ItemCnt as the Value field.

Compare with Time Periods

Select all the rows from the Formative and F-Middle Horizon and then scroll down and hold down Control Key and also select the 4 rows that have a "1" in the MH column (Middle Horizon). You should have 34 rows selected.
Choose Table Menu > Export…. And save the table to your geodatabase call it “EarlierSherds” Add to map.
Table Menu > Switch Selection so you have the other 55 rows selected. Export… save “LaterSherds”
Open the EarlierSherds table. Right Click the top of the ArchID menu and choose Summarize…. And scroll down and select Sum from F-MH, MH and MF. Specify the output table: choose your Callalli GDB and call it “EarlierSherds_ArchID_sum”
Open LaterSherds. Right Click the top of the ArchID menu and choose Summarize…. And scroll down to LIP-LH, LIP, and LH and choose “Sum”. Specify the output table: choose your Callalli GDB and call it “LaterSherds_ArchID_sum”

Data Visualization

When we are doing fieldwork we try to record the richness and diversity of the phenomena we're documenting. We don't want to simplify too much in the field because we can never gain back that detail.

However, when we're doing analysis back in the lab we're often forced to make aggregations to detect general trends and patterns. We're now in a situation to look at two classes of data (Lithics and Ceramics) using two trends each: obsidian and non-obsidian, and early sherd and later sherd styles. Separately we have distance to water raster that we can query (Zonal Statistics) against the other datasets.

Turn on and off layers to investigate patterning. What is the relationship between material density and proximity to streams? What if you knew that chert cobbles are found in the stream beds? What is the relationship between time period and obsidian use?

There is a lot of overlap and we are going to look at a very granular view (we linked to ArchID not SiteID so every collection point gets its own dot). There are also many areas with no-data. This is typical for archaeological studies.

Further exploration: Consider using the View > Graphs > Create Graph function in Arcmap to plot these relationship.

When you've got a Graph you're satisfied with click on the title bar and choose "Add to Layout" to include it on the periphery of your final map.

Nevertheless do you see whole areas that show no obsidian and yet lots of data from some time period identified by sherd styles (Early or Late)? Farmers reoccupy rich farmlands with water. Herders reoccupy grazing areas. Despite these overlaps archaeologists use layers of maps to investigate the relationships between land and resources through time.

Make a nice map illustrating some aspect of these patterns. You can use “dumb” Graphics labels to put text on your map indicating the areas you think are perhaps revealing some pattern. Feel free to use other symbology (pie charts, bar graphs) but make sure it’s not cluttered with giant pie charts by experimenting with the settings.

True maps contain both a scale bar and a north arrow (otherwise they’re just diagrams). Make sure to include these in your final map together with a label.

Open the properties on the Scale bar and look through the settings to make sure the increments are round numbers and reasonable for the map.

Conclusion: Please include a one – two paragraph discussion with your map submission. Razor-sharp obsidian was clearly useful for hunters but what role do you think it had for herders? Obsidian also has symbolic value but the relative valuation can change through time.