Lab 07 - Overlay & site suitability analysis

This lab will introduce the idea of overlay analysis in terms of a common application, site suitability. Site suitability can take a number of forms but in this lab we’ll approach these concepts in vector and raster form. You will also touch briefly on Terrain analysis, something we’ll see more of in the next lab.

A logging company has been given a license to develop and to cut down trees in the Oakwood area (see the figure to the right). However, there are restrictions on where the company can cut down trees.

The purpose of this exercise will be to select sites where the company can log. We will use the ArcGIS to select these sites according to the selection rules given below.

License Restrictions:

• No trees may be cut down within 10km of the shrine, in order to preserve the aesthetic of the landscape.
• No trees may be cut down within 1km of the sea, the lake, or any river in order to help prevent land erosion.
• The logging sites must be within 5km of existing roads for easy access by heavy logging equipment, since conservation laws will not allow any new roads to be built in this area.

Materials

At this point you should be able to discern how to accomplish what you want to accomplish. We want to show the logging company where they can and cannot log. As is the case with most things in life, there are a few ways we can do this, and the choice is yours (if you got this, you are my new favorite student). Below I walk you through one way to accomplish this, but you will gain far more if you design your own way to arrive at the same end goal.

• Start ArcMap, create a new empty map, and save your map document to your personal folder.
• Add forest.shp, rivers.shp, water.shp, roads.shp and shrine.shp, all of which are located in your Part01 folder.
• In the data frame properties, change the display units to kilometers.
• Arrange the layers correctly
  • Make sure List By Drawing Order (upper left of the Table of Contents) is selected.
  • Arrange the layers in this order (bottom to top): water, forest, roads, rivers, and shrine.
• Change the symbology of each layer:

  • Make the rivers blue.

  • Make the roads black.

  • Make the shrine red.

  • Make the water blue. This requires extra steps:

    • In the Symbology tab, click Categories near the upper left part of the window.

    • Select Unique Values.

    • Select ISWATER as the Value Field.

    • Click Add All Values.

    • Uncheck the box the left of

    • Looking at the Value column, recall that water = 10 and everything else = 1, so double-click the color box next to 10 and change it to a blue color, and then click the color box next to 1 and change it to Hollow.

    • Remember these steps, they will come in handy when making the final map for this analysis

  • Make the forests green. There are two kinds of trees in this layer, so use a different shade of green for each kind. To do so, follow similar steps as were just performed for the water layer.

• In ArcToolbox (the button with the little red box on it) go to Analysis Tools | Proximity | Buffer. Use the following settings:
  • Input Features: roads
  • Output Feature Class: roads_buffer.shp (make sure to save in your Part01 folder) Linear Unit: 5 km
  • Dissolve Type: All
• Click OK and give it a minute to work. When it’s done it won’t look pretty, but we’ll change that later.
• Follow the same steps to create 1 km buffer around the rivers layer and a 10 km buffer around the shrine layer. Make sure you give the output files descriptive names (e.g., rivers_buffer.shp and shrine_buffer.shp).

• We need to select the sea and lake polygons from the water layer. If we were to just do this we would end up buffering both shapes in the layer. Recall that when tools run, they typically operate within the active data frame, and use the current selection (or the whole shapefile if nothing is selected). So to accomplish what we want to do, we want only the actual water from the water.
• To do this, go to Select by Attributes… and use the expression “ISWATER” = 10.
• As you did with the other layers, create a 1 km buffer of the water features. (Note that if features are selected, the buffer tool will only create buffers for those selected features.)

• Now we will find potential logging areas based on access to roads. In ArcToolbox go to Analysis Tools | Overlay | Intersect. Use the following settings:
  • Input: forest and roads_buffer (you will have to select each one individually)
  • Output Feature Class: Solution_1.shp (make sure to save in your Part01 folder)
  • Join Attributes: All
• Click OK.
• Now we will merge the rivers_buffers and Solution_1 layers. In ArcToolbox go to Analysis Tools | Overlay | Union. Use the following settings:
  • Input: Solution_1 and rivers_buffer
  • Output Feature Class: Temp_1.shp (make sure to save this in your Part01 folder)
  • Join Attributes: All
• Click OK.
• To find those areas from Solution_1 that are not within the river exclusion zone (i.e., buffer), go to Select by Attributes…, select the Temp_1 layer, and apply the expression “FID_rivers” = -1
• In the Table of Contents, right-click the Temp_1 layer and select Data | Export Data…. Save the features as a shapefile called Solution_2.shp. Leave the other settings as they are.
• When asked if you want to add the exported data as a layer, click Yes.
• Click the Clear Selected Features button on the main menu.
• In ArcToolbox go to Analysis Tools | Overlay | Union. Use the following settings:
  • Input: Solution_2 and shrine_buffer
  • Output Feature Class: Temp_2.shp (make sure to save this in your Part01 folder)
  • Join Attributes: All
• Click OK.
• To find those areas from Solution_2 that are not within the shrine exclusion zone (i.e., buffer), go to Select by Attributes…, select the Temp_2 layer, and apply the expression “FID_shrine” = -1
• In the Table of Contents, right-click the Temp_2 layer and select Data | Export Data…. Save the features as a shapefile called Final.shp. Leave the other settings as they are.
• In the Table of Contents, remove all layers that have names that start with Temp or Solution or that end with buffer. The following layers should remain: shrine, rivers, roads, Final, forest, and water.

• Using the map below as a rough guide, create a map that indicates which forest stands the company can legally log. Make sure your symbology is clear and that the crucial components of the map are not covered up by your Final layer. (Hint: Use patterns, and/or go to the Display tab of the layer’s Properties and change the transparency.) Use some creativity— don’t just copy the map below! Export a PNG of the map and paste it in the answer sheet.

In this part of the lab, we are interested in assessing how much of Douglas County is covered by tornado sirens. In other words, we want to know how many people can actually hear a tornado siren when it goes off. While the sirens for Douglas County were designed to be heard at distances of up to 5,800 feet, there might be areas in which people can’t hear them.

• In ArcMap, create a new empty map, and save your map document to your personal folder.
• Add countybnd.shp, roads.shp, censusblk.shp, and sirens.shp from your Part02 folder.
• Create buffers around the sirens using the following settings:
  • Input Features: sirens
  • Output Features: sirens_buffer.shp (make sure to save in your Part02 folder)
  • Linear Unit: 5800 feet
  • Dissolve Type: All
• Arrange layers so that you can see the siren points and buffers on top.

• Add kansas2000censusblk.dbf to the data frame.
• Right-click the censusblk layer, select Joins and Relates, and select Join. Join the layer to kansas2000censusblk.dbf using the STFID field.
• Examine the attribute table of censusblk to make sure the join was successful. Do you see all sorts of demographic attributes? If so, the join was successful.
• Right-click the header of the POP2000 attribute and select Statistics. The sum value is the number of people in Douglas County. Record the value somewhere, as you will be entering this into the comments section of your Blackboard submission. Close the window.
• Use Select by Location… to select the census blocks that are within the buffer zone.
  • Hint: Use the are completely within the source layer feature setting. The result should be 1,590 census blocks selected.
• Find the population in these census blocks by opening the censusblk attribute table, right- clicking the POP2000 attribute header, and selecting Statistics. (It will only give you statistics on the selected records.) Record this number.
• Find the population out of the range of sirens by clicking Table Options (upper left corner of the attribute table), selecting Switch Selection, and then using the Statistics tool again. Record this number in yout answer sheet and answer the associated questions.

Total Douglas County population:
Number of people who can hear the tornado sirens:
Number of people who cannot hear the tornado sirens:
Percentage of the population that can hear the tornado sirens:
Percentage of total population that cannot hear the tornado sirens:

Note: Your math should add up. If it doesn’t, reevaluate.

• Switch the selection back (i.e., back to the blocks within range of the sirens) and close the attribute table.
• Take a screenshot of your entire screen and paste it in your answer document.

You have been asked to help a small town in Vermont find a suitable location to build a new school. There are a few considerations to take into account, including the slope of the land surface, the distance to both recreation sites to existing schools, then reclassify these derived datasets to a common scale from 1-10. You will then set the model up to weight them according to a percentage influence and combine them to produce a map displaying suitable locations for the new school.

###
1. Open a new ArcMap and setup * While I would normally say it is ok to create a new data frame and import the data, in this case we will set a few environmental variables, so for reproducibilities sake we will import data into a new map entirely. * Bring in the data > Hint: Getting a permissions error when you try to bring in the rasters? Did you Google it? https://community.esri.com/thread/223412-invalid-raster-dataset-failed-to-create-raster-layer-permission-denied-how-can-i-overcome-this-problem: Think back to the data health page. * Click on Geoprocessing > Environments: * Click the down arrows for the processing extent and set both the processing extent and the Snap Raster to elevation * Click the Raster analysis section and set the cell size to the same as elevation. 

   Framing the work flow

   The town has asked you to identify some suitable areas in which to site a new school. More explicitly, our site suitability analysis will concern itself with 4 factors. We don’t want to spend a lot of money converting a land cover into a developable state. We also don’t want to build on a particularly steep slope. Because of the lack of a robust transit system, we want to place this school further away to maximize the accessibility of that school to the population. We also want to place that school close to existing recreational sites. Finally, we want to prioritize these factors in order of distance to a recreational facility > Distance to a school > Land use type = Slope.

   Given that, we have all the information we need to successfully execute this analysis, but it will be helpful if we frame what we want to accomplish in a visual sense. ArcMap has a wonderful tool, Model builder, that provides an easy means of visualizing a GIS work flow. Because these factors have different “units”, we need to normalize them to some common scale. Lets pick the easiest one, and make them a simple scale from 1-10, with 10 being the best. The tool to do that is the reclassify tool. The next sections will walk you through the analysis.

   Land cover factor

   

   • The first thing we will do is reclassify the land cover. Use the reclassify tool.
     • Set the input to the landuse raster
     • Use the LANDUSE field as the reclass field
     • Change the values to those shown below
     • Click on the folder icon next to the dialog box, create a new geodatabase in your part 3 folder, and save all subsequent rasters in there. Call this Output raster LandC_R

   Slope factor

   

   • If you use the search function, there are a few ways we can calculate slope. For this application we want but we want the slope tool from the Spatial Analyst toolbox.
     • Set the input raster to elevation
     • Save the output as Slope
     • Leave the rest as defaults
   • Your map should look something like this:
   • Finally, we need to reclassify the slope to our common value range. Use the reclassify tool again.
     • Set the input raster to Slope Set the reclass field to value
     • Click on the classify button
     • Set the classification method to Equal Interval, and the number of intervals to 10
     • Call the output SlopeR

   School distance factor

   

   • To calculate distance we will use the Euclidian Distance tool.

     • Set the source features to Schools
     • Call the output distance raster SchDis

   • Then we need to reclassify this raster, in comes the handy reclassify tool.

     • Set the input to SchDis
     • Set the reclass field to Value
     • Perform the same normalization as above (classify button > Equal Interval, intervals to 10)
     • Save the output as SchDisR

   • Your final map after this step should look something like so:

   Recreational site distance factor

   

   • We need to calculate the distance from our features again, re run the Euclidian Distance tool with the following settings

     • Set the source features to rec_sites
     • Call the output distance raster recDis

   • Then we need to reclassify this raster, in comes the handy reclassify tool again.

     • Set the input to recDis
     • Set the reclass field to Value
     • Perform the same normalization as above (classify button > Equal Interval, intervals to 10)
     • Save the output as RecDisR
     • Because we want the area closer to the sites to have a higher value, we need to click the reverse New Values button

   • Your final map after this step should look something like so:

   7 Site suitability and thresholding

   

   • With all of our standardized layers, we are ready to calculate our suitability raster. Search for the Raster Calculator Tool (Spatial Analyst Tools)
     • Create an expression to weight each factor using the table below.
     • Call the output SchSites

   Layer	Weight
   recR	0.5
   schR	0.25
   landuseR	0.125
   slopeR	0.125

   • The results of the site suitability are displayed below. Lets perform one last analysis, and find and highlight the areas with a site suitability score larger than 8. Use the raster calculator tool with the following expression: SchScore > 8, save the output as GoodSites.

   8 What to submit

   • First let’s change the symbology of our map.Open the attribute table of the GoodSites raster and record the number of cells which meet our criteria. Set the GoodSites symbology so that 1 is hollow, and 0 is gray. Set the transparency of the layer to 50, making a poor mans mask of the interesting areas. Turn off all layers other than GoodSites and ScSitesTake a screenshoot of the map, and record the number of cells and area rated as good fro school site development. Submit the answer sheet to blackboard, you’re done!