GIS In Archaeology
Lab Exercise 13
Step 1: Create New Map Document
Start new sessions of ArcCatalog and ArcMap. Add the Mexico_DEM and Countries layers to a new map document.
What is the projection information for each of these layers?
Raster operations in ArcGIS can have problems if the map projection differs from the projection of the raster data you wish to manipulate, so make sure that the map document is using the same map projection as Mexico_DEM
Step 2: Self Test Questions
What is the projection information for the Mexico_DEM layer?
What is the cell size in the Mexico_DEM data set?
What does the acronym DEM stand for?
What are the minimum and maximum elevations in the Mexico_DEM data set?
What is the current display symbology for the Mexico_DEM layer?
Step 3: Apply Hypsometric tinting
Hypsometric tinting is the process of assigning different color values to various elevation ranges. The default symbology of ArcGIS for most elevation layers is a black-to-white color ramp. This color ramp is generally not the most visually appealing, nor the most interpretable.
Right click on Mexico_DEM in the table of contents and bring up the symbology panel of the properties window. The current layer symbology uses a Stretched, Standard Deviations classification scheme with a black-to-white color ramp.
Select and apply the color ramp listed below to which uses a wider range of colors.
By using a hypsometric tinting scheme such as this it is somewhat easier to identify major elevation zones than before.
The new mapping scheme makes it more easy to distinguish the major elevation break between lower and mid-elevations. However, the resulting map shows too much area in the white end of the color scheme which can make it difficult to visualize differences here.
Return to the symbology properties tab. You will notice that the current classification scheme classifies the data based on 2 standard deviations:
Modify the n: value from 2 to 3 to apply a classification based on 3 standard deviations. Using a value of 3 reduces the amount of white and does a better job of showing where the highest elevation peaks are in the data set.
In this map the lowest elevation elevation zones are shown as a light blue color. For this map we will modify the color ramp so that the lowest elevations are displayed as a light green color instead.
From the symobology properties tab right click inside the box containing the color ramp schema and select Properties from the pop-up menu.
We are now presented with the Edit Color Ramp window that shows the detail of how this complex color scheme is organized.
The top entry in this scheme represents the colors used for the lowest elevation values. Highlight this entry and then click the Properties button to view the Edit Color Ramp window.
To modify the lowest elevation color from light blue to light green click the down arrow next to the light blue color to bring up the color picker window. Select the light green color called light Apple.
Notice that the preview: color ramp now reflect this change in color. Click the Apply button on the Edit Color ramp window, then click the OK buttons to close the Edit Color Ramp windows and Apply the new layer symbology to the map. The new map display now uses a light green color for the lowest elevation zones.
Try experimenting with other standard deviation values, classification schemes, and color ramps to view the wide range of control available within ArcGIS.
Step 4: Contour Mapping
Hypsometric tinting is a powerful graphical display to show general elevation trends within a data set. What it lacks, however, is information showing exact elevation values. In contrast, the use of contour mapping provides a means to display definite elevation zones and their associated elevation values.
Click the down arrow next to the Spatial Analyst entry in the Spatial Analyst Toolbox. Select Surface Analysis Contour from the drop down menu. Make sure the Mexico_DEM is the selected Input surface, and set the Contour interval to 500.
Based on this procedure a new layer called ctour is added to the table of contents and this contains a series of elevation contour lines at a 500 meter interval. The combination of both the hypsometric tinting technique and the contour lines provides a greater level of information content than either procedure alone.
Step 5: Generalizing the Contour Map
The contour mapping procedure used in the previous step produced an accurate but somewhat messy set of contour lines. For analytic purposes the most accurate data possible is preferable but often for a visual display a more pleasing, smoothed result is often preferred.
Click the down arrow next to the Spatial Analyst entry in the Spatial Analyst Toolbox. Select Neighborhood Statistics from the drop down menu to bring up the Neighborhood Statistics window. Modify the parameters in the window to use Height and Width values of 5.
This procedure effectively smooths the elevation data by calculating mean elevation values. A new temporary layer called NbrMean of mexico_dem is added to the table of contents. Rerun the previous contour mapping procedure on this layer (make sure that you select the NbrMean layer rather than the original Mexico_Dem layer in the Contour window) and view how this results in a somewhat smoother contour map. The result is somewhat less accurate than the original contour map but the results are probably preferable from a visual standpoint.
Step 6: Modify Hillshade Parameters
In the last exercise we created a hillshade map from the DEM.BIL dataset. At the Hillshade parameters window we accepted the default values (Azimuth: 315; Altitude: 45; Z Factor: 1). The procedure created a somewhat muted looking pseudo 3-D view of the terrain. It is possible to exaggerate the elevation differences in order to create a more dramatic hillshade model.
Rerun a hillshade analysis for the DEM.BIL raster layer but this time at the parameters window specify a Z Factor: of 5. The hillshade layer created by these parameters exaggerates the differences in elevation (the Z-values) and produces a more accentuates the terrain hillshading model.
The Azimuth parameter refers to the angle at which the hypothetical sun is sitting in relationship to the terrain (0 = North, 90 = East, 180 = South, 270 = West). The default setting of 315 places the ‘sun’ source to the northwest of the DEM.
The altitude parameter refers to the height of the sun, but it is expressed in terms of the suns angle in relationship to the ground. An altitude value of 0 places the sun level with the terrain (ie on the horizon), an altitude value of 90 models the sun as though it is directly overhead.
Try playing with various parameter values to see the effects and understand how these parameters interact with one another.
Step 2: Raster Scales Effects on Raster Summaries
In the last exercise we used the Spatial Analyst to create a slope layer from the DEM and a distance to Water layer based on the Rivers layer. The model that we were using was that all other things being equal when select a location for a settlement humans will tend to favor flatter lands over steeper lands and land closer to water over land further from a water source. In order to combine these two variables (slope and distance to water) we used the Raster Calculator to add the two values together.
The idea was to create a summary layer that would combine both the effect of both these variables on human settlement decisions. However, if we examine the resulting map in comparison to the distance to water map we can see that they are extremely similar. This is because the values for distance to water range from 0 to 10,000 whereas slope values only range from 0 to 57.4. This means that the effect of the distance to water layer swamps out the impact of the slope variable. We need to put the slope and distance to water variables onto a similar scale so that neither one dominates the combination
Step 3: Reclassify Slope Values
A brief analysis of sites in the region indicates that in general a slope of 0 to 7.5 degrees has no appreciable negative impact on human settlement decisions (these are basically flat lands). Slopes of 7.5 to 15 degrees have a slight negative impact and slopes over 15 degrees seem to generally have been considered to be too steep for habitation. We will reclassify the lands based on the following table:
Slope Slope Class Slope Class Label
0 – 7.5% 1 Flat (under 7.5%)
7.5 – 15% 2 Moderate (7.5-15%)
15 – 60% 3 Steep (over 15%)
Click the down arrow next to the Spatial Analyst entry in the Spatial Analyst Toolbox. Select Reclassify from the drop down menu. Make sure that the Input Raster is the slope layer. The Reclassify window will convert the current value ranges under the Old values column to be the values in the New values column. Currently there are entries under Old values but we only need 3 and we need to set our own ranges. Highlight the 2nd row in the reclassification table (old value 2.2436 – 5.1605, new value 2) then click the Delete Entries button to the right. That row is removed from the reclassification table.
Repeat the above procedure until there are just 3 entries in the reclassification table (not including the row for NoData NoData). When there are just 3 entries click in first value under old values and change the range from 0 – 2.2436 to 0 – 7.5 (note: it is critical that there is a space between the hyphen and each of the range values). Change the other values to match the slope reclassification table listed above. Then click the OK button.
An alternate method to modify the reclassification table is by clicking on the Classify… button on the Reclassify window. On the Classification window change the Method from Manual to Equal Interval, then the Classes: value from 10 to 3, then type the values 7.5, 15 and 60 as the 3 values in the Break Values table to the right of the window.
Once you have reclassified the slope values a new layer called Reclass of Slope is added t the map. This is a raster data set similar to the slope layer but the values for this layer can only be 1, 2, or 3.
Step 4: Make Slope Reclassification a Permanent Data Set
Follow the procedure used in the last exercise to save the Reclass of Slope layer as a permanent layer (save it in the Rasters folder with the name SlopeClass).
Step 5: Relabel Slope Classification Layer
The previous slope reclassification procedure has produced a useful layer but the default labels are somewhat hard to understand.
We will relabel the laye to make it more legible to the casual reader. First rename the Reclass of Slope layer in the Table of Contents to read Slope Class. Next bring up the layer’s property window and from the symbology tab change the labels from 1, 2, 3 to Flat (under 7.5%), Moderate (7.5-15%), Steep (over 15%). Feel free to change the slope class colors as well if you’d like.
This procedure doesn’t do anything to modify the underlying data but it does make our map document easier for anyone to read and understand.
Step 6: Reclassify Distance to Water Values
Follow the previous steps to modify the distance to water layer to the following distance classes:
Distance Distance Class Distance Label
0 - 750 1 Close (under 750 m.)
750 - 2000 2 Moderate (750 m - 2 km)
2000 - 10000 3 Far (over 2 km.)
Make the layer a permanent data set in the rasters folder called H2ODistClass.
Relabel the layer to be Distance to Water Class, and change the class labels to match the entries listed in the table above.
Step 7: Agricultural Potential
An important factor in the daily life of agriculturalists is the agricultural productivity of the land they farm. Add the layer Ag_Potential (in the V:\Anth497G\Lab_Data\Mexico folder) to the map. This is a polygon vector data set classifying the land into it’s suitability for agriculture. In the attribute table there is a field called LType (Land Type) which classifies the land from 1 to 9 (1 is the most suitable for agriculture, 8 the least suitable, and 9 is for water).
Step 8: Vector to Raster Conversion
In order to summarize this data set along with the slope and distance to water variables it is necessary to convert from a vector to a raster data set.
Click the down arrow next to the Spatial Analyst entry in the Spatial Analyst Toolbox. Select Convert, Features to Raster from the drop down menu. In the Features to Raster window set the following parameters Input features: Ag_Potential, Field: LType, Output cell size: 60, and Output Raster: V:\Anth497G\Lab_Data\Mexico\Rasters\AgLands
Step 9: Agricultural Lands vs. Agricultural Productivity
Currently the agricultural lands are classified into 8 groups (plus lakes) for their agricultural suitability. These classes are arbitrary classes and don’t necessarily reflect the agricultural productivity of the land. If we used the land classes as they stand we would basically be saying that type 1 lands are twice as productive as type 2 lands, three times as productive as type 3 lands, four times as productive as type 4 lands, etc. A study of the area indicates that this is not the case and a more suitable agricultural productivity index would be:
Land Class Productivity Index Description
1 100 Most Productive
2 90 Highly Productive
3 70 Mod-High Productivity
4 60 Low-Mod Productivity
5 40 Low Productivity
6 40 Low Productivity
7 20 Least Productive
8 20 Least Productive
9 0 Lake
Use the previous techniques to create a Productivity Index using the above values. Save this as a permanent data set in the Rasters folder called AgLandIndex.
Step 10: Local Soils vs. Agricultural Productivity
This last procedure created a 60 meter raster data set of agricultural lands indexed for productivity. That’s to say we classified each 60 meter square for the agricultural suitability of the land within that 60 meter square. From an agriculturalists perspective, however, this is not necessarily the most appropriate measure of the agricultural productivity of an area. It is most unlikely that an agriculturalists would limit their selection of agricultural soils to just the 60 meter square surrounding their settlement. Various studies of agriculturalists indicate that they are regularly willing to travel distances of 1 km, 2 km, and even more to farm suitable soils.
Click the down arrow next to the Spatial Analyst entry in the Spatial Analyst Toolbox. Select Neighborhood Statistics from the drop down menu. In the Neighborhood statistics window alter the parameters to match the ones below. Be careful to make sure your options exactly match and Output raster is V:\Anth497G\Lab_Data\Mexico\Rasters\Catchment1Km.
If you set the parameters correctly you will get a data set called Catchment1Km with values ranging from 0 to 79,700.
Step 11: Questions to Ponder
What does this data set represent?
Why did we use Sum as the Statistic type?
Why did we use a Circle Neighborhood
What does a Neighborhood setting of Radius: 1000, Units Map mean?
Step 12: Save Map and Exit
Save map and exit ArcMap.