GIS In Archaeology

Lab Exercise 12

Step 1: Create Study Area Map
Launch ArcCatalog and ArcMap and load the map you saved at the end of the Vector Data Analysis lab.

Evaluating the Impact of Elevation on settlement locations

In the previous lab we examined the impact of agricultural land class and river locations on settlement decisions in the study area. While we used different strategies to examine the relationship both of these were vector based data layers. For this segment of the exercise we will examine the possible influence of elevation which is modeled in a raster based Digital Elevation Model data layer.

Step 2: Examine the digital elevation layer
Looking at map of the digital elevation map (DEM_100M) shows that the entire DEM is much larger than they survey study limit. Right click on the layer in the table of contents select Properties from the popup menu. Click the Source tab of the Layer Properties window and we see that the DEM has elevation that ranges from a low of 699 meters to a high of 5500 meters with a mean at 2151 meters.


These statistics are for the DEM as a whole, not just for the project study area. To obtain the data for just the study area go to the Spatial Analyst toolbar (if the toolbar is not open select View -> Toolbars from the main menu and make sure there is a check next to the Spatial Analyst option). Click the arrow next to Spatial Analyst and select Zonal Statistics from the menu.

Select Survey_Limits as the Zone dataset, Id as the Zone field, DEM_100M as the value raster, uncheck the Chart statistic box and save the Output table as Survey_Area_Elevation_Summary

The output table summarizes all the cells from the DEM dataset that are found within the survey limits. The table indicates there are 36,528 cells inside the survey area and these range from 855 to 1677 meters with a mean elevation of 1042 meters.

Step 3: Reclassify elevation data
It is possible to analyze and compare sites elevation information using the raw elevation values but it can be simpler if the data is collapsed into larger groups. For the purposes of this exercise we will reclassify the elevation data into 100 meter elevation zones.
In some cases the easiest way to reclassify raster data using the Spatial Analyst Reclassify tool. In this case, where the reclassification value can be expressed as a mathematical function of the original data the simplest way to do this is by using the Spatial Analyst Raster Calculator tool. From the Spatial Analyst toolbar select the Raster Calculator tool. Elevations in the 800-899 meter range can be reclassified by an 8 by dividing the elevation by 100 and keeping just the integer portion of this result. In the Raster Calculator window click the Int button, then the DEM_100M layer, the / symbol and enter in the value 100. Then click the Evaluate button to perform this operation.

The result of this operation is the creation of a layer called Calculation in the table of contents. To give the layer a more memorable name right click on the layer entry, then select properties. In the properties window select the General tab and change the layer name to be Elevation Zones

Step 4: Assign Elevation Class Data to Sites
When we use vector data sets it is possible to use spatial overlay or spatial join operations to link sites with the data from other layers. This same strategy cannot be used with a raster based data sets.
As long as there is a field in the attribute table of the sites data set which has unique values for each site we can use the Zonal Statistics function. In the sites table each site has a unique site number so we will use this strategy to assign them their corresponding elevation information.
From the Spatial Analyst toolbar select the Zonal Statistics Tool. Select Sites_training as the Zone dataset, Site_No as the Zone field, Elevation Zones as the value raster, uncheck the Ignore NoData and Chart statistic boxes, check the Join output table to zone layer, and change the output table to be Sites_Training_Elevations. Then click the OK button.

Once the operation is completed right click on Sites_Training in the table of contents and open the attribute table. Scrolling across the table you can see that the sites layer information has been joined to the elevation data.

Click on the Options button and add a new field. Call the field Elevation_Class and define it to be a long integer data type. Once the field has been added to the table right click on the field name and select Calculate Values. Set the value to be equal to the field [Sites_Training_Elevations.MEAN] then click the OK button.

Right click on Sites_training in the table of contents and select Joins and Relates -> Remove Joins -> Remove All Joins. All of the Sites_Training_Elevations fields that were joined are now gone but the Elevation class field remains.

Right click on the Elevation_Class field name and select Summarize from the popup menu. Store the output table as Sites_Training_Elevation_Summary in the Morelos geodatabase. Click the OK button

Open the Sites_Training_Elevation_Summary to see the number of sites by 100 meter elevation zone. Clearly the majority of sites occur in the 800, 900 and 1000 meter elevation zones.

While it appears that there may be a preference for the lower elevation zones it is once again necessary to evaluate this by comparing these data to the overall distribution of elevations within the study area

Step 5: Summarize Survey Area Elevation Data
Right click on the Elevation Zones layer and open the attribute table. The attribute table contains a listing of all the elevation zones in the DEM as well as the number of times each one occurs.

These data cannot be used to compare with the site elevation zones because this table contains many cells which occur outside of the study area.

To obtain a summary of just those raster cells within the study area it is necessary to create a new DEM layer that contains only those cells which are within the study area. One way to do this is by combining a raster layer that takes on the shape of the study area with the DEM layer.

We will use an analysis mask that will allow us to select out only those elevation values which intersect the boundaries of the survey limits layer.

Click on the Spatial Analyst and select Options. On the General set the Analysis mask to be Survey Limits.

Next click on the Extent tab and set the Analysis extent to be “the same as layer dem_100m”.

Finally click the Cell Size tab and set the Analysis Cell size to be the “Same as layer dem_100m”. Click OK to close the options window.

Return to the Spatial Analyst toolbar and select Raster Calculator. Double click on Elevation Zones, then click the Evaluate button.

Notice that a new temporary layer is added to the map that has the same values as the Elevation Zones layer but it is restricted just to the survey limits area.

The new layer will have a name similar to Calculation. Right click on this layer and open the attribute table.

Right click on the Count field and select Statistics. Notice that the layer has a total of 36,458 cells. Close the statistics window.

If we create a new table with elevation class and divide the count values by 36,458 we will obtain the percentage of the survey area covered by each elevation zone. Following this procedure we get these results:
Value Count Area%
8 2988 0.082
9 13652 0.374
10 8023 0.22
11 8143 0.223
12 2632 0.072
13 620 0.017
14 254 0.007
15 101 0.003
16 43 0.001
17 2 0.0001

If we combine this table with the Sites by Elevation zone table we get the following results.
Value Count Area% Sites%
8 2988 0.082 0.24
9 13652 0.374 0.49
10 8023 0.22 0.15
11 8143 0.223 0.08
12 2632 0.072 0.03
13 620 0.017 0.01
14 254 0.007 0
15 101 0.003 0
16 43 0.001 0
17 2 0.0001 0

From inspection sites are more frequent than we would expect by random chance in the lowest two elevation zones (800-899 meters, and 900-999 meters). Once the elevations exceed 1000 meters the percentage of sites is less than the area covered by each of those zones. There is only one site found at an elevation of over 1300 meters.

As with the distance to river data it appears that there is a tendency for sites to be located in the lower elevation zones, once we move into the higher zones the frequency begins to drop off quite dramatically.

Evaluating the Impact of Slope on settlement locations
A final environmental variable that is commonly investigated when looking at settlement patterns is slope. Investigating the impact of slope on settlement regions can be performed in a manner similar to looking at elevation.
Build a slope map from the Digital Elevation Map Data – for the slope map calculate slope using degrees as the unit of measure. Next follow, and modify as needed, the previous steps to investigate any relationships that might exist between slope and site locations. If you round slope measurements off to the nearest integer value you should get the following results:

Value Count Area% Site%
0 3794 0.1 0.05
1 7913 0.22 0.2
2 6100 0.17 0.26
3 4192 0.11 0.2
4 3097 0.08 0.1
5 2325 0.06 0.08
6 1702 0.05 0.03
7 1313 0.04 0.04
8 1107 0.03 0.01
9 813 0.02 0.01
10 700 0.02 0
11 572 0.02 0
12 465 0.01 0.01
13 387 0.01 0.01
14 327 0.01 0
15 283 0.01 0
16 220 0.01 0
17 239 0.01 0
18 182 0.005 0
19 115 0.003 0
20 114 0.003 0
> 20 498 0.01 0

The overall pattern that emerges from this analysis seems to be that area with slopes up to 5 degrees have a slightly higher than random chance of being inhabited, areas with slopes of 6-10 degrees are only very slightly selected against but that land that is over 10 degrees slope has a low probability of being inhabited.