Download the data (~3.4MB) for this project, and unzip the states, counties and lakes (polygons), rivers and roads (lines), and cities (points) shapefiles to a "proj1" subdirectory on your data stick. Although ArcMap treats each of these as single datafiles, each shapefile is comprised of multiple component files in your work directory. Download the 2000 Census Data spreadsheet "sf3.xls" for US counties as well.

Open a new ArcMap session and add the six shapefiles--states, counties, lakes, rivers, roads and cities-- to your map. Check the attribute table for the counties shapefile; if there are negative values for the POP1990 field, you will need to replace the COUNTIES.DBF file in your data directory with an uncorrupted version of this file that you can download here.

Display your map with an appropriate coordinate system:   The initial display will treat the lat-lon coordinates as if they were planar coordinates, so this unprojected map will look pretty distorted. Find the "Projection" tab in the dataframe ("Layers") Properties menu to give your map a decent-looking projection (choose any of the four predefined continental North American contiguous USA projections.)  Zoom your map to just the continental US. (If you include obviously unprojected maps in your project webpages, I will have you dragged out of the lab and shot.)

If you click the "Display" tab at the bottom of the table of contents frame, you can re-sort (drag and drop) the layers in the legend so you have the following display order (bottom to top): counties, states, lakes, rivers, roads, cities.  Now display cities as tiny (1- or 2-pixel) green dots, highways as thin red lines; rivers as blue lines, lakes as blue, states with black outline and clear fill color, and counties with light gray outline color and whatever fill color you like.  This will let you see county boundaries in gray within black state boundaries.  If the map looks too busy, uncheck the display box for the cities and/or roads so you don't see them in the map.

Figure out the following selection problems. For each one, zoom the view to the relevant features and File--Export Map a PNG-format version of your map (reduce the resolution, pixels per inch, to reduce your image to the desired size). Map the continental US only, omitting Alaska and Hawaii. Use any graphic editor (GIMP, Photoshop, MS-Windows Paint or whatever) to crop or edit these if necessary (all rescaling and other edits should be done in RGB mode). Transfer your finished image files to a proj1 subdirectory under your public_html directory on the Copland server for inclusion in your project web page.

1. What is the total area of the five Great Lakes (units are square miles)?
   How many US counties border the Great Lakes? (Note that Arc's Select-by-Location has lots of alternative criteria: share a boundary, intersect, etc., that may yield inconsistent results.)
   If you calculate the population density of each of the 3,140 counties, what is the average?
   If you sum the county populations and divide it by the sum of the county land areas, what is the true population density?
   Export a GIF-format map zoomed to these selected counties with the Great Lakes unselected.
2. How many counties intersect some major river portion of the Mississippi River system?
   What was the total population of these counties in 1990?
   Export a map zoomed to the Mississippi counties, with the rivers unselected.
3. How many counties in the US had negative population growth between 1990 and 1999?
   Export a map of the continental US with these counties highlighted.
4. What percent of the US population lived in counties within 200 miles of New Castle County in 1999? Open Arc Toolbox (little red toolbox icon), select just New Castle County, and use the Analysis Tools-Proximity-Buffer tool to create a 200-mile radius ring feature around New Castle County. It will look egg-shaped, because the Counties shapefile is in lat-lon coordinates and the Buffer tool only works correctly on projected shapefiles. Right-click on Counties and use Data-Export Data to create another version of the Counties shapefile in the coordinate system of the data frame. If you select New Castle County and rerun the Buffer tool on this new shapefile you should get a correct near-circular buffer.
   Export a map of this region, with the counties selected and both buffer rings displayed.
5. How many individual US counties have larger area than the whole state of Delaware?
   Export a map of the continental US with these counties highlighted.
6. How many US counties are intersected by an interstate or equivalent highway in the roads shapefile?
   What was the overall percent population growth between 1990 and 1999 in these counties?
   What was the overall percent population growth in counties not intersected by an interstate highway? (Export a map.) 

Thematic mapping: Using the 2000 Census data, create 3 uniformly-sized choropleth maps of the continental US showing percent white population (white population normalized by total population) in each county using three alternative classification methods--Natural Breaks (the default), Equal Interval and Quantile--using the same color ramp and number of classes (7 to 10) for each map.
Create three equivalent maps showing percent black, one for each classification method.
Create three equivalent maps showing percent Hispanic, one for each classification method.
Export uniformly-sized PNG images (about 400 pixels wide) of each of these 9 maps. (You may want to make the county boundaries transparent so you just see the state boundaries.)
Which method gives the truest-looking representation of the distributions of whites, blacks and Hispanics in the US?

The one cartographically complete map you will make in this course:   Create a map from the 2000 Census data showing percent population 65 or older, by county, for the continental US. Once you get a really swell-looing map in the Data view, switch to the Layout view to compose a proper map with legend, title, scale bar, compass rose, projection information, etc. The page orientation should be landscape, not portrait. Export your finished map as a larger-scale PNG (800-1200 pixels wide) image file.

Editing and summarizing attribute fields:   Add a new "PCTPOPCH" field in the Counties attribute table and use the Field Calculator to populate it with percent population changes between 1990 and 2000. Export this to a PNG image file.
Use the state name field in the Counties attribute table to Summarize 1990 and 2000 county populations to get the Sum county populations for each state. Then open the output table and calculate overall percent changes in population for each State. Join this table to the States shapefile's attribute table. 

The only time you will every use ArcMap's chart utility:   Create a bar graph with a labeled column graph showing percent changes in population in the 12 northeast states (ME, NH, VT, MA, RI, CT, NY, PA, NJ, DE, MD, WV).  Put the 2-letter state abbreviation under each column; include horizontal gridlines and an appropriate title.  Save a PNG image file of this graph to include in your web page. (This little exercise should convince you to use Excel instead.)

Using external data in ArcMap:   Download county-level 2002 Census of Agriculture data from the National Ag. Statistics Service website. Select the "Table 5. Government Payments ..." table (step 1) and "Government Payments - Average per farm (dollars, 2002)" from this table (step 2), for the "United States / All States, Only Counties" (step 3) and click the "Get Data" button. After the 3,000+ county records all display, scroll to the bottom, click "download CSV," and unzip and import the comma-separated values file to Excel.

To split the State\County column into separate columns in Excel, insert a blank column after it, then use the Data-Text to Columns utility, with the "\" character as a delimiter. Create 5-character county FIPS codes by combining the the 2-character state and 3-character county FIPS fields in another empty column. FIPS codes are supposed to be text, not numbers, so that they retain leading zeroes. The Excel expression to combine a 2-character text string in cell B2 with a 3-character text string in cell C2 would be: =TEXT(B2,"00")&TEXT(C2,"000"). Save this edited worksheet in Excel (.XLS) format, exit Excel, and add the file to your ArcMap project. Join this file into the Counties shapefile attribute table based on the matching FIPS codes. Once you have the correct join, create a nice-looking thematic map showing "Who's farming the government!" and save a decent-size PNG image of it for your web page.

Pork and presidential elections:   Download and add the Federal Taxes Paid vs. Federal Expenditures, by State, 1981-2004 table (source: The Tax Foundation) to your project. Access the "Editor" toolbar and "Start Editing." Open the DBF table, select the record for the District of Columbia, and delete it, then "Stop Editing," saving the change. (DC is an extreme outlier because so many federal employees live there.)

Join this table to the States shapefile and create a red-green thematic map of the ratio of federal expenditures to taxes paid in 2004 (field R2004), showing "winner" states (R2004>=1) in green and "loser" states (R2004<1) in red.

Use Select By Attribute to count the tax "winner" and "loser" states that went for Bush and Kerry respectively. Plug these counts into a 2x2 frequency table of states like this:

	"tax winners"	"tax losers"
went for Bush	a	b
went for Kerry	c	d

To estimate the statistical significance of the correlation between Bush/Kerry and winner/loser states, calculate and evaluate the Chi-square statistic: The formula for an (approximate) chi-square value (with one degree of freedom) is

Alternately, you can use Excel's CHITEST function to compare the actual 2x2 frequency table to the 2x2 frequency table you would expect if the 31 Bush states and 19 Kerry states had equal likelihoods of being winners or losers:

	"tax winners"	"tax losers"
went for Bush	20	11
went for Kerry	12	7

When you are done with Arc, put all your GIF-formal graphic files in a "proj1" subfolder under your public_html web-server folder.  Create a web page within the "proj1" subfolder called "index.html" to report your answers and display all your exported maps, your chart, and any other eye-candy that you want to include. In the parent public_html directory edit your home page to include a link to your proj1 subdirectory.

Looking ahead (extra credit)

Load the US states, counties and rivers layers. Set an appropriate coordinate system, e.g., Continental US, Albers Equal Area Conic. Use Data--Export Data to create a copy of the rivers shapefile in this coordinate system. Zoom to just the continental US. Under Tools--Extensions, activate the Spatial Analyst extension; then add the Spatial Analyst toolbar. Set the Spatial Analyst Options specifying the States layer as a mask, specifiying that output maps should be in the dataframe's coordinate system, specifying the current display as the analysis extent, and specifying a cellsize that yields maps smaller than 1000x1000 cells.

Select all the rivers in the Mississippi system from the newly-created rivers shapefile. Use Spatial Analyst's Distance--Straight Line module to create a raster map of distances from the selected features. Switch the selection, so all the rivers that are not part of the Mississippi system are selected; then run the same Distance module on these. Use a Raster Calculator expression such as Distance < Distance2 to define the approximate drainage area of the Mississippi system. The Raster Calculator should generate a binary (0,1) raster. Select raster category identifying the Mississippi drainage area and use Convert--Raster to Features to create a polygon shapefile of this area.

"At Rex Kwan Do we use the buddy system. No more flying solo. You need somebody watching your back at all times. Second off, you're gonna learn to discipline your image. You think I got where I am today because I dress like Peter Pan over here? Take a look at what I'm wearing, people. You think anybody wants a roundhouse kick to the face while I'm wearing these bad boys? Forget about it! Last off, my students will learn about self respect. You think anybody thinks I'm a failure because I go home to Starla at night? Forget about it!"