Home > GIS in Action > ArcView 3.1 Lab >
     
GIS in Action
ArcView 3.1 Lab

Click on PC or MAC to download the Data Set for this lab. PC (532 K) MAC (572 K)

Spatial Analysis II: Distance Measures and Buffers

Note: This laboratory covers material in chapters 8 and 9. It requires the ArcView Spatial Analyst extension. If you do not have Spatial Analyst, skip this lab.

Spatial Decision Support Systems (SDSS). Humans frequently need to consider a variety of factors and information in making spatial decisions. If a county needs to add a new fire station for example, the county planners will require information about suitable locations, relative costs, and potential benefits. While a software system can be designed to support optimal single-site solutions, planners have found that they prefer to be presented with several viable solutions rather than a single optimal solution. This is because there are sometimes intangible reasons for not wishing to employ the single "best" solution. Planners then want to have alternative choices that, while not the best, are still quite good. SDSS are GIS systems that serve data to a heuristic algorithm that produces multiple near-optimal solutions. In this lab, you will be using Spatial Analyst and buffers to produce a number of alternative solutions to a simulated GIS problem.

Buffer analysis, and point-in-polygon overlay. The buffer is a frequently used analysis device. Buffers calculate distances from spatial objects, and produce polygons that reflect the object and the area around it. Buffer zones are frequently used to mitigate environmental hazards. Emergency plans for nuclear power plants, for example, include multiple buffers which describe the level of hazard at increasing distance from the plant.

The overlay operation is one of the generic operations central to GIS. All types of spatial objects can be overlain in order to analyze spatial relationships between sets of objects. In this lab we will perform a point in polygon overlay analysis which will find all of the points from one data set that spatially coincide with the buffer polygons that you will create.

Spatial analysis project design. The success or failure of a GIS project generally lies in the quality of the design. There are frequently several adequate solutions to a problem using GIS, and so selecting an approach usually will devolve to considering costs such as time expended on the project. The concept of parsimony or simplicity is also a good test for the quality of a project design. In this lab, the analytic approach has been designed for you, but it is still worthwhile for you to consider whether or not this is the way you would have designed the project.

Task Set 1

 

 

Task Set 1: GIS Consultant

In this laboratory exercise we will look at an analytic word problem, and develop a GIS solution using ArcView. The problem statement is as follows:

You have been hired in your first job as a GIS consultant. A traveling salesperson, Jorge Rodgers, has presented you with an interesting problem. He works for a marketing firm that sells outdoor spas along with four other salespeople. The management has given Jorge and his colleagues a list of five cities that need to be visited. Each salesperson is expected to set up shop in the downtown area of the city and make business contacts by cell phone. They will be paid a commission on each call regardless of getting a sale as long as the potential customer requests additional information. Jorge estimates that he makes 1000 calls a day. His commission is 25 dollars per customer requesting information. The cell phone service that Jorge uses allows unlimited calling within a 50-mile radius of his registered location. Thus he is interested in making all of his calls to cities within a 50-mile radius of his home base. Jorge has found from experience that the baseline rate for expressing interest in the product is about 4 percent. He also has been given marketing information about the predisposition of people in the 5 regions to purchase his product. Jorge has been asked to rank the cities in terms of his preference. The rankings of the 5 salespeople will be compared, and an attempt will be made to send each person to the city of his or her choice. Jorge is hoping to buy a house and get married within a few years and so wishes to make his ranking choices based on the highest projected commission per call. He also hopes to introduce his company to the new GIS technology and its potential uses. He has asked you to provide him with some information to help him make his rankings.

Basic Information:

Jorge lives in Atlanta, and must pay his own airfare. You may assume that the cost of airfare is 10 cents per mile.

The cities that he must rank and their relative dispositions to buy spas are:

Chicago 4%
Seattle 16%
San Francisco 12%
Denver 8%
Irvine 4%

Note: You will probably need to take notes, and write things down as you think through this problem. Do not start the lab until you have understood the problem. At the end of the lab you will be writing a report.

Note: Before going any further, read through the entire lab, and estimate how much time you expect to spend on the lab including write-up. Write a quote for Jorge, billing your time at 20 dollars per hour. At the end of the project, record your actual time spent, and the true cost of your services. Turn this in with your lab.

Question: Can GIS help Jorge? Just based on the information above, what kind of data do you think you will want to use?

Question: You know that your census data was collected in 1990. How will you deal with the fact that you are using old data? Name three possible solutions that will satisfy your client.

You have determined that you will want to use demographic data for the five US cities listed above plus the location of Atlanta. Start ArcView. Add the Cities.shp data to the new view from your data directory. You will want to change the projection to an equidistant projection with distance units of miles, (The Lab in chapter 2 contains instructions for the projection procedure). Make the theme active, and open its table. Select the six cities that you will need. This can be done most easily by selecting the City_name field, and then selecting the Sort Ascending button.

Next scroll through the city names until you find the six cities you need. You can use the shift key to do a multiple select for all six of them. You can use the Promote button to make sure that you have all six.

With the cities still selected, make the View active. Select Theme->Convert to Shapefile. Give your new theme a name (e.g. mycities.shp), and make sure to save it to your own working directory. Do not delete the Cities.shp theme from your view. Now open the table for your new theme (mycities.shp) and select all of the cities except Atlanta. You will now use Theme->Create Buffers to buffer the five cities you selected. There is a set of buffer dialogue windows that you will need to use properly.

The first is the Create Buffers window. Make sure that the Use only the selected features box is checked. You should have 5 selected features. Press Next when done.

Make sure to save to a new theme. Use the folder icon to navigate to your work directory. Select Finish when done.

Question: Why were you asked to change the projection of your map at the beginning of the lab?

Question: When you look at your 50-mile buffers, how do they compare to your city points? Based on this comparison, what can you conclude about the precision of the point as a spatial object? Do points have areal extent?

Now you have all of your data ready to analyze. This procedure can be programmed, but we will perform it as a computer assisted manual task. This means that you will need to write information down as you go along.

Note: If you want to get screen shots of ArcView windows, you can make the window active, then press Alt and Print Screen. If you open Word and use paste, the window graphic will be pasted into your word document.

Now we will calculate what the total population of the cities in a 50-mile radius of all of our target cities is. Make Cities.shp active. Select Theme->Select by Theme. The Select by Theme window will open. Make sure that the first frame says Are Completely Within, and that the second frame says the name of your buffer data.

Select New Set. Make Cities.shp active. Select Theme->Convert to Shapefile. Name your new theme (e.g. Bufcities.shp). Arrange your view to look something like this.

Question: When you look at this map, what conclusion might you draw about the accuracy of the buffer? Do you think that this is a correct conclusion?

Question: Although the symbols on the map do not look right, what would happen if you portrayed all of the data in this view at a 1:1 scale on your computer screen? Do you think that you need to explain this to Jorge?

We will now work with some simple statistical information. Make the Bufcities.shp theme active. Open the table for the active theme. All of the records should be selected and highlighted in yellow. Use the Select None button to clear the selection set.

Make the Field Type active in your table. Select Field->Summarize. The Summary Table Definition window will pop up. Select Households as the Field and Sum to Summarize by. Click on Add and then click on OK. Your new table, sum1.dbf should look like the table below.

Question: Could sum1.dbf be useful to Jorge? Why?

Notice that Field->Statistics is not available for use. This is because you do not have a numerical field selected and ArcView does not perform statistics on categorical data. Select the numerical field Pop1990. Notice that in the Field menu, both Summarize and Statistics are available. Select Field->Statistics.

Question: Explain the meaning of each of the statistical values in your Statistics result. These do not have to be technical explanations.

We will now create a point data set for each of the 5 urban point clusters. Use the Select Feature tool to select the cluster of points around each city.

For each selection set, select Theme->Convert to Shapefile, and name each new shapefile for the city around which the points are clustered. Summarize the Pop_90 field and the Households for each of the five new themes. Make certain that you save these tables with useful names. Otherwise they will all look alike. Your view should now look something like this:

It might be useful now for cartographic purposes to turn cities.shp off and add the states.shp theme from the same directory in order to provide a spatial context. You have now collected enough demographic information to advise Jorge. Now we will go on to distance measures. Turn off all of the themes except the theme that contains your six original cities.

Make Mycities.shp active and use Theme->Auto-label to label the cities for identification. Now we will calculate the distances from Atlanta to each of the five cities. Use the Measure tool to measure the distances between Atlanta and the other five cities. Record these distances.

Assignment: Write up your solution to the city ranking problem. Remember to trade off the cost of travel against potential higher sales. You will want to write a short concise report to Jorge. Tell him in simple terms how you performed this analysis. If you think that there is any uncertainty in the result, tell him why, and explain whether or not it was unavoidable. If you believe that the data you used had limitations, explain these. Provide him with the rankings and the maximum sales potential based on the 4% take rate, and the travel cost for each city. Turn in this report with one map produced in Layout that you feel focuses the information produced in this project. Include your initial estimate and final cost of your time. Remember that Jorge is a businessman. Be concise.

Go to top



Copyright © 1995-2009, Pearson Education, Inc., publishing as Pearson Prentice Hall Legal and Privacy Terms