Lab 2: Map Projections and Coordinate Systems

Note: This exercise is a modification of material originally developed by Kristina Schneider and Dr. David R. Maidment for CE 394K: GIS in Water Resources at The University of Texas at Austin.  These materials may be used for study, research and education.  Please credit the authors.

To download the summarized questions from this lab (MS Word format), click here.

Part 1 - Map Projections and Coordinate Systems

Objectives

  • Gain experience in ArcMap using on-the-fly-projection of various common projections
  • Learn how to define a projection in ArcCatalog and ArcToolbox
  • Learn how to project vector data using ArcToolbox

Introduction

For software to show the geographic location of features, the features must be stored with x and y coordinates tied to a specific origin (e.g. x=0, y=0; a "Coordinate System") and datum that specifies the length of the minor and major semi-axes of an ellipsoidal earth.  Together, the datum and coordinate system comprise the so-called " Spatial Reference " of a data set.

In ESRI lingo, Spatial References are of two types.  The first, called a Projected Coordinate System (" PCS "), is composed of a datum (e.g. WGS84, NAD83, NAD27) and coordinate system parameters (e.g. location of origin and standard parallel(s), type of developable surface, false easting and northing) specific to a map projection.  PCS coordinates are stored in meters or feet relative to the PCS' origin.  The second, called a Geographic Coordinate System (" GCS "), is composed only of a datum and has units of decimal degrees - GCS coordinates are stored in latitude and longitude.  West longitudes and South latitudes are stored as negative values, and the units are decimal degrees, not Degrees, Minutes and Seconds.

One of the principal, traditional strengths of GIS software is the ability to convert data stored in one Spatial Reference to another Spatial Reference.  This map projection process may involve conversion of coordinates from one GCS to another GCS, from a GSC to a PCS, or from one PCS to another PCS.  A computer screen or paper map can only have one Spatial Reference.  Thus, the ability to convert coordinates from one Spatial Reference to another is a key aspect if one wishes to simultaneously view data sets that have different Spatial References.

Within ArcGIS, ArcToolbox has tools to do this conversion, permanently converting (i.e. creating a new data file) GCS data into PCS coordinates, or converting from one PCS to another.  Alternatively, temporary ("on-the-fly") conversion to a different Spatial Reference is done automatically upon adding data in ArcMap.  This on-the-fly conversion permits viewing of data sets that have different Spatial References without first having to convert them in ArcToolbox to a common Spatial Reference.  Of course, for either of these processes to work the software must first know the Spatial Reference of the coordinates being converted.  If this information is missing from a data set, then ArcToolbox and ArcCatalog have tools for creating it, a necessary first step (called "Defining a Spatial Reference " or " Defining a Projection ") before conversion can be successful.  Many older data sets are missing Spatial Reference information and can produce problems if this "defining" step is ignored.

Data

Files for this exercise are contained in the Lab_2_data folder, which can be downloaded as a zip archive here. To use the files, save the archive to your hard drive and extract all of the files to a folder labeled Lab2 (remember the recommended file labeling procedures from Lab 1) in your GISWorkspace folder.

Most of the data are within a geodatabase, called Mapproj.mdb, that contains four feature datasets:

  • World: containing feature classes cntry94 and world30 - countries, and 30º meridians and parallels for the earth
  • USA: containing feature classes States, Counties and Latlong - states and counties of the US and a 5º grid of meridians and parallels
  • Minnesota: containing MNcounties, MNmunicipalities, MNrailroads, MNrivers, MNstate, MNquadindex, and MNwatersheds, coverages of Minnesota showing state, county, and municipality boundaries, railroads and rivers, and 7.5' quadrangle map extents
  • Twin Cities Metro Area: containing TCMAcounties, TCMACTU, TCMAlakes, TCMArivers, TCMAzip5, and TCMAquadindex, coverages of county, city, and township boundaries, lakes and rivers, USPS zipcode areas, and 7.5' quad map extents

         All of these contain data in geographic coordinates relative to the NAD83 datum.

Finally, there is an aerial orthophotograph of the Mound quadrangle contained in four separate files representing quarters of the quadrangle.  This is a 1-meter resolution image obtained from the Minnesota Department of Natural Resources (MnDNR).  The spatial reference for this file is UTM zone 15, NAD83, but this information is not present in a form that can be read by the software.

1. Projections of the World

1.1 The World in Geographic Coordinates

1.11 Loading Data

  • Start ArcMap, choosing “A new empty map”. There will be a single Data Frame called "Layers" in the Table Of Contents (TOC). Remember that it is good practice to set the data source options to record relative path names when beginning a new project in ArcMap (File -> Map Properties, select Data Source Options,and "Store relative path names." Click OK.)
  • Right-click on "Layers", and choose Properties. Select the General tab, and change the Data Frame name to “Geographic Coordinates”. Click OK.
  • Click the Add data button Add button , navigate to your Mapproj.mdb geodatabase and add all of the World feature dataset feature classes.
  • If necessary, drag cntry94 above world30 in the TOC.

1.12 Symbology and Labeling

You want to make the layer world30 rectangles show just their outlines, so that if needed you can show meridians and parallels on top of the countries.  You would also like to label a few of the countries.

  • Right-click on the world30 layer in the TOC and select Properties. Navigate to the Symbology tab and click on the rectangle in the Symbol cell to get the Symbol Selector window. Select the Hollow color and press OK. You should see the world in geographic coordinates.
  • To label just a few features, first right click on the cntry94 layer in the TOC and deselect Label Features (if its on). On the Drawing toolbar (normally at the bottom of the window), there is a Label button label button that allows you to label individual features. It is found underneath the big big A button symbol in the Draw toolbar (click on the arrow next to the "A" to see other options for this tool). If you don’t see the Draw toolbar, navigate to the View menu, and click on Toolbars>Draw to make it visible.
  • Click on the label button label button
  • Keep the selections chosen by the computer in the Labeling Options window. Close the Labeling Options window, and click on the countries on the map that you would like to label.
    Pretty cool, huh?!
    Note: when labeling this way the program assumes you want to label features in the topmost layer of the TOC.  Be sure to have the cntry94 feature class as the uppermost layer in your TOC or you will get numbers from the world30 boxes when you click on the countries to label them.

Where does the label button get the names for the countries?  To answer this question, right-click on the cntry94 in the TOC, select Properties and choose the Labels tab.  The label tab contains a drop-down menu called "Label field" that contains all of the field names in the cntry94 attribute table - it has been set to NAME.  The attribute table for cntry94 contains a field called "NAME" that contains country names (see this for yourself by opening the attribute table for this layer).  This is where the label button looks to find the text for the labels (as do all other automatic labeling tools).  Note that from this tab you can also control the label font, its placement, the scale at which it is displayed, and select from some predefined styles.

1.13 Coordinate Display

  • Move the cursor around on the map and you will see a pair of numbers at the bottom right that change as you move the cursor. These give the location of the cursor, and from the values displayed you can see that these data are latitude and longitude displayed in degrees, minutes and seconds.

If you find that the units displayed are decimal degrees but are described as “Unknown Units”, you can reset “Unknown Units” to “Degrees Minutes Seconds” by right-clicking the Data Frame name in the TOC, clicking the General tab, and changing the Display drop-down menu to “Degrees Minutes Seconds”. 

  • Likewise, you can change from “Degrees Minutes Seconds” to "Decimal degrees" using the same procedure.  Do So.

Now, SAVE YOUR MAP as World.mxd (mxd extension will be added automatically).
 

Question 1:
a) What is the spatial extent of the view shown in degrees of latitude and longitude?

b) Where is the point (0,0) (deg. longitude, deg. latitude) located? Provide a screen capture of your world map to illustrate your answer.
c) Using the New Rectangle tool in the Drawing toolbar, draw a box around Australia. What is the extent of this box?  Give the coordinates of the lower left corner and upper right corner in decimal degrees.
 

1.2 The World in Robinson Projection

So far, we have examined the world as “unprojected” data (i.e. in geographic coordinates). Now we will view the layers cntry94 and world30 as projected data. A common projection for the world is the Robinson projection.

1.21 Creating a new Data Frame and adding data

  • Create a new Data Frame, using Insert>New Data Frame.
  • From the View menu, select Data Frame Properties. Select the General tab, and name the Data Frame Robinson.
  • Right-click on the world30 feature class in the Geographic Coordinates Data Frame, select Copy, then right-click on the Robinson data frame and select Paste Layer.  You’ll see the world30 feature class appear in the Robinson Data Frame.
  • Do the same for the cntry94 feature class. Put world30 on top of cntry94 for visual comparison of meridians and parallels.
  • Right-click on the Robinson Data Frame and select Activate to display the data in this frame.  Without this crucial last step, you will not be able to do anything with this new Data Frame.

To copy the two feature classes as a group, you could have:

  • right clicked on the original Data Frame in the table of contents (with Display tab active) and created a new "Group Layer";
  • Dragged and dropped individual layers onto the new Group Layer;
  • Copied and pasted the entire group layer to the new Data Frame.

This alternative method can be very helpful when dealing with a large number of layers.

1.22 Setting the Coordinate System for a Data Frame

ArcMap has the ability to take data that are either in geographic coordinates or in any number of projected coordinate systems and project them to a new coordinate system within the Data Frame.  To set the coordinate system of the Data Frame (in this case to the Robinson projection):

  • Right click on the new Robinson Data Frame and select Properties.
  • On the Coordinate System tab, under the Select a Coordinate System box, click the Predefined folder. Then select Projected Coordinate Systems>World>Robinson. Click OK.
  • A warning may appear, but we can ignore it for our purposes (just click Yes). Ignore this message every time it appears throughout THIS exercise (in other circumstances you may need to deal with this issue; we'll cover it in lecture).

You will see the world map appear in a Robinson projection. Pretty slick! As you can see, the landmasses appear much less distorted in this projection. It is important here to note that the files themselves have not been projected permanently—they are only displayed in the new projection. This is called “on-the-fly” projection in ArcMap, which is distinct from create new data files in different projections, sometimes referred to as "hard projection", using ArcToolbox. 

  • To better distinguish the two views of the world, change the color for cntry94 in the Robinson Data Frame.

The Robinson projection is a relatively new map projection for the Earth designed to present the whole Earth with a minimum of distortion at any location. If you move the cursor over this space, you'll see that the coordinates are now projected coordinates (i.e. eastings and northings) reported in meters.

1.23 Create your own Map and layout of several Data Frames

  • Create another Data Frame, and copy/paste the layers from the World Data Frame into the new data frame. Be sure to put world30 on top of cntry94 for visual comparison of meridians and parallels.
  • In the new Data Frame, play with the different projections available in the Coordinate System tab of the Data Frame Properties, and explore the different shapes the world can take.
  • Note that if you want to view a different Data Frame, right-click on the Data Frame name, and choose Activate.
  • After you’ve experimented with different projections, select one for a layout you will turn in.
  • You can create a layout that contains several different data frames. Switch to the Layout view, and you’ll see all three Data Frames displayed. By default, the Data Frames are on top of each other.
  • Resize and reposition these frames to create a more attractive map layout.

Be sure to add all the necessary elements to your layout (title, name, date, etc.). Consult the class layout guidelines for help. Also, be sure it is clear which Data Frame corresponds to which projection - give each a caption or title.

  • Save the map file.

 

Map to be turned in: a color layout showing the world in geographic coordinates, in Robinson projection, and a projection of your choosing.  Before finalzing your layout, drag the latlong grids below the countries to enhane map readability.  Title the layout "Map 1: Projected And Unprojected Maps Of The World".

 

2. Projections of the United States

2.1 United States in Geographic Coordinates

We will now examine map projections used for the continental United States. We could continue adding Data Frames to the previous map file, but to simplify things, let’s create a new map file.

  • Use File>New>Blank document to create a new map file.
  • Rename the Data Frame “Geographic Coordinates” and set Display units to decimal degrees
  • SAVE the map file as USA.mxd.
  • Add the states and latlong feature classes from the USA feature dataset of mapproj.mdb.
  • Move states below latlong in the TOC, if necessary.
  • Use the Zoom In tool and zoom to a view of just the continental US (exclude Alaska and Hawaii). Use the Pan tool to move the US into the center of the view window if necessary.

Question 2 :
a) What is the geographic extent of the United States? Give the eastern and western limits of longitude and the northern and southern limits of latitude of the continental US (not including Alaska or Hawaii) to the nearest degree.

b) Which parallel defines much of the border between the United States and Canada?

c) If we removed a wedge out of the earth cut along the meridians defining the most eastern and western points in the continental United States, how much of the globe would we have cut out? Give your answer as a percent of the total volume of the earth (assume the earth is a sphere for this problem).

 

2.2 United States in Albers Equal Area Projection

The Albers Equal Area projection has the property that the area bounded by any pair of parallels and meridians is exactly reproduced between those parallels and meridians in the projected domain. That is, the projection preserves the correct area of the Earth. For example, an island with an area of 100 km2 will exhibit an area of 100 km2 in the Albers Equal Area projection. The drawback to this projection is that it distorts direction, distance and shape somewhat.

  • Create a new Data Frame and copy and paste in the layers latlong and states from the previous frame.
  • Make sure states is below latlong to enhance visual comparison of meridians and parallels.
  • Rename the data frame Albers Equal Area.
  • Bring up the Data Frame Coordinate System tab. Select the coordinate system Predefined>Projected Coordinate System>Continental>North America>USA Contiguous Albers Equal Area Conic. Zoom into the continental US.

  • Compare the United States in geographic coordinates and in the Albers projection. You should change the color of one of the layers to further distinguish them.

You will see that in geographic coordinates, the United States appears to be wider and flatter than it does in Albers Equal-Area Projection. This does not occur because Canada is sitting on the USA and squishing us! This effect occurs because as you go northward, the meridians converge toward one another while the successive parallels remain parallel to one another. When you reach the North Pole, the meridians converge at a point. In an unprojected view, the meridians are drawn as parallel lines instead of converging lines. Drawing the meridians in this manner distorts the regions between them. As you approach the poles, the meridians have to be drawn farther and farther apart in order to make them parallel. For this reason, distortion of the regions between parallels increases as you move toward the poles. A more precise geometric explanation is provided below.

If you take a 5 degree box of latitude and longitude, such as one of those shown in your Arcmap file, the ratio of the East-West distance between meridians to the North-South distance between parallels is Cos (latitude): 1.  For example, at 30°N, Cos(30°) = 0.866, so the ratio is 0.866 : 1, at 45°N, Cos(45°) = 0.707, so the ratio is 0.707 : 1. In the projected Albers Equal Area frame the result is that square boxes of latitude and longitude appear as elongated quadrilaterals with a bottom edge longer than their top edge. In geographic coordinates, the effect of the real convergence of the meridians is lost because the latitude and longitude grid form a set of perpendicular lines, which is what makes the United States seem wider and flatter in geographic coordinates.

  • SAVE your map document as USA.mxd.

Map to be turned in: A layout showing the United States in geographic coordinates and in the Albers Equal Area projection.  Before finalzing your layout, drag the lat/long grids below the countries to enhance map legibility.  Title the map "Map 2: Projected And Unprojected Maps of the Conterminous United States".

 

3. Projections of Minnesota

3.1 Minnesota in geographic coordinates - displaying a subset

  • Create a new map document, Minnesota.mxd. Name the Data Frame "Geographic Coordinates".
  • Add in the feature classes counties and latlong (latlong on top) from the USA feature dataset of the mapproj.mdb geodatabase. The feature class counties contains counties of the United States, including Alaska and Hawaii. Make sure latlong is on top of counties in the table of contents.
  • To make it easier to determine the counties in Minnesota, right-click counties, select Properties, and go to the Symbology tab. Choose to display the counties by Categories>Unique values, and select STATE_NAME from the Value Field drop-down menu. Press the Add Values… button (don't choose Add All Values!), select Minnesota, and press OK (you may first have to click the button "Complete List" if Minnesota is not displayed in the list). Deselect the check mark for <all other values>, which will leave only counties with the STATE_NAME as Minnesota to be shown. Click OK.

  • Only the counties in Minnesota remain on the map.  Zoom in to see a larger view of Minnesota counties.

The latitude/longitude grid displayed is in intervals of  5-degrees. You can determine what latitude or longitude a particular line represents by moving the cursor to any line and reading the numbers displayed at the bottom right below the map view window.

Save your map document as Minnesota.mxd.

 

Question 3:
What is the geographic extent of Minnesota to the nearest degree in North, South, East and West?

 

3.2 Minnesota in Lambert Conformal Conic Projection

The Lambert Conformal Conic projection is a standard projection for maps of areas whose East-West extent is large compared with their North-South extent. This projection is "conformal" in the sense that lines of latitude and longitude, which are perpendicular to one another on the earth's surface, are also perpendicular to one another in the projected domain.  Angles remain undistorted in this and all conformal projections.

  • Create a new Data Frame, copy and paste Latlong and Counties to it from the previous Data Frame. In the Properties for the new Data Frame, rename the Frame Lambert Conformal Conic, move to the Coordinate System tab and select Predefined>Projected Coordinate System>Continental>North America>USA Contiguous Lambert Conformal Conic projection.
  • Click OK.

Notice how the meridians now fan out from the north pole (a consequence of using a conic projection centered on the axis of rotation of the earth). The display shown is that produced by cutting the cone and unfolding it so that it lays flat.

  • Zoom to Minnesota in the Lambert Conformal Conic projection.

Notice that Minnesota appears to be slightly tilted to the left. This occurs because the Central Meridian of the projection is 96ºW, which would appear as a vertical line in the display if it were shown. Regions to the West of this meridian appear tilted to the right while those to the East (much of Minnesota) appear tilted to the left. Morris sits just to the east of this meridian (95º55'W).

3.3 Minnesota in Universal Transverse Mercator (UTM) Projection

The Universal Transverse Mercator projection is actually a family of projections, each having in common the fact that they are Transverse Mercator projections produced by wrapping a horizontal cylinder around the earth. The term transverse arises from the axis of the cylinder being perpendicular or transverse to earth's rotation axis. In the Universal Transverse Mercator coordinate system, the earth is divided into 60 zones, each 6° of longitude in width, and the Transverse Mercator projection is applied to each zone. Most (if not all) of the geospatial data available from Minnesota state agencies are provided in the UTM projection (zone 15).

  • As before, create a new Data Frame and copy/paste Latlong and Counties to it from any of the  previous Frames. 
  • Double-click on the new Data Frame, and follow previous steps to rename it UTM Zone 15N. Click on the Coordinate System tab and select Predefined>Projected Coordinate System>Utm>NAD 1983>NAD 1983 UTM Zone 15N projection.

The parameters in the "Current coordinate system" box mean that the Central Meridian of Zone 15 is at 93°W so that it covers from 90°W to 96 °W; the Reference Latitude is 0.0000 (the equator, which is 0°N); the origin of the coordinate system is at the intersection of the Central Meridian with the Reference Latitude and thus is at (0°N, 93°W), where the coordinates are (x, y) = (500000, 0) m. The False easting of 500,000m ensures that all points in the zone have positive x coordinates. The y-coordinates are always positive in the northern hemisphere because 0 is at the equator. In the southern hemisphere, a false northing of 10,000,000m is applied to the equator to ensure that the y-coordinate is always positive.

The Scale Factor of 0.9996 means that along the Central Meridian, the scale of the map is slightly reduced (distorted).  True (undistorted) scale is only achieved at two lines of secancy, which are 1.5 degrees to either side of the central meridian.  The scale factor 0.9996 describes the maximum distortion within the zone; scale distortion away from the central meridian is less than this (more closely approximating a scale factor of 1, which exists only along lines 1.5 degrees away from the central meridian).

  • Click OK to see the projection applied. The pattern of meridians and parallels looks very different from those of the other projections we’ve looked at. Note how the meridians converge at both the North and South Poles.

  • Zoom in on Minnesota.

  • As is stated above, most (if not all) geospatial data that is available from various state agencies (DNR, DOT, etc.) come in the UTM projection. Add any (or all) of the data contained within the Minnesota feature dataset to make sure that they line up properly with the projected Minnesota counties. Experiment with the symbology and order of the different layers you added.

  • SAVE the Minnesota.mxd map document.

Question4: 
a) In the lab procedure, you referenced the UTM coordinates for Minnesota to Zone 15.  Why was this zone chosen instead of the others?
b) How many 7.5' quadrangles are needed to cover the state of Minnesota (hint: you might want to add the MNquadindex and use the attribute table).
c) How many 7.5' quadrangles are needed to cover Stevens County?

 
To be turned in: A color layout showing Minnesota in Geographic, Lambert Conformal Conic, and UTM projections and a color layout showing Minnesota (UTM projection) with one or more of the additional coverages you added to the map. Before finalizing your layout, drag the latlong grids below the counties to enhance map legibility.  Be sure to save your map document after you complete the layout.  Title the layouts "Map 3A: Projected And Unprojected Maps Of Minnesota " and "Map 3B: Minnesota Geography."
 

4. Projections of the Twin Cities Metro Area (TCMA)

4.1 Twin Cities in Geographic Coordinates - Zoom to Layer, Select by Location, Export to Shapefile

You have viewed the effect of different projections on different scales. from the world to country and state levels. In the next few steps, you will take a look at the Twin Cities Metro Area and the effect that two map projections have on a map of the city.

  • Create a new map document, TwinCities.mxd, and add in the layers TCMAcounties, TCMACTU, TCMAlakes, TCMAquadindex, TCMArivers, TCMAzip5 from the Twin Cities Metro Area feature dataset.

The feature class TCMAcounties and TCMCTU are coverages of the counties, cities, and township units of the Twin Cities Metro Area. The classes TCMAlakes and TCMArivers show the lakes and rivers. The feature class TCMAquadindex is a mesh of 7.5 minute quadrangles for the TCMA with map sheet names for each quad, and TCMAzip is a coverage of different 5-digit USPS zip codes covering the metro area. All of these data were downloaded from the MetroGIS DataFinder and are in the UTM projection (zone 15).

  • Double-click on the Data Frame name and rename it TwinCities UTM.
  • Right-Click on the TCMAcounties layer and select Zoom to Layer.
  • Rearrange the layers in the TOC to make the clearest map and symbolize the different layers to distinguish the various features.
  • You may also hide different layers by unchecking the box next to the coverage in the TOC window.

If we didn't have a feature class containing 7.5' quadrangle coverage of the TCMA, we could easily extract this information from the layer containing 7.5' quadrangle coverage for the entire state (MNquadindex). At this point, we will produce a new file that contains a subset of the 7.5 quad outlines and names for quad maps that cover the TCMA.  So far, in order to focus on particular features in a layer, we have simply hidden them from display by not including them when we've symbolized.  Now, let's extract these quads from the MNquadindex layer and save them as a new file.  The steps are to first select (highlight) the ones we want, then "export" them to a new file, a shapefile in this case. To select the 7.5 minute map sheets that encompass the TCMA, you will use one of the many selection tools available in ArcMap. 

  • Add the MNquadindex layer (from the Minnesota feature dataset) to your TCMA data frame and zoom to the MNquadindex layer. If you drag it to the bottom of the TOC, then you can see where the Twin Cities counties fall within the state of Minnesota (if you didn't already know).

  • From the Selection menu at the top of the ArcMap window choose Selection>Select by Location and fill in the resulting window as below. This will select features from MNquadindex that intersect (see the Preview graphic at the bottom of the dialog box that shows polygons that intersect polygons) TCMAcounties.
    • Select feature from MNquadindex that intersect TCMAcounties.

You'll see a subset of the quadrangles selected on the map after you click Apply. Close the Select by Location window.

To export the selection to a new file:

  • Right click on MNquadindex in the TOC, and select Data>Export Data... ; ensure that "Selected Features" appears in the "Export:" drop-down menu and that the "Save as type:" reads "shapefile".
  • Name the resulting file TCMAquads.shp, Browse to the location where you want to save it (Lab_2_data folder on your hard drive, perhaps in a new folder), then click OK and answer yes to add it to the display.  Go to Selection>Clear Selected Features to deselect the quads you previously selected.

  • Delete MNquadindex from the TOC.
  • Right click on the new TCMAquads layer in the TOC and select Label features to show the map names. You can open the attribute table of this file, called TCMAquads.dbf, in Excel and copy the list of map names to paste into a Word document.  If you do so, be careful not to overwrite the original dbf file.
  • SAVE your TwinCities.mxd file.

Question 5:
How many 7.5' quadrangle sheets are needed to cover the Twin Cities Metro Area?  Hint: Open the Attribute table for the TCMAquads layer and examine the bottom of the window.  The number of records (there is one record for each quad) in this file is given.

 

4.2 Twin Cities Metro Area in Geographic coordinates

The Twin Cities coverages you were provided with for this lab were already projected, but we can actually undo the UTM projection and display the data in geographic coordinates.

  • Create a new Data Frame; name it Twin Cities Geographic. Add the various Twin Cities layers as in the previous Frame.
  • Click the Coordinate System tab in the frame's Properties window and select Predefined>Geographic Coordinate System>North America >North American Datum 1983. Under the General tab in the frame's Properties window set the display units to decimal degrees (rather than meters as with UTM). Click OK.
  • Right-click on TCMAcounties and select "Zoom to Layer". Play with the Symbology until it looks similar to that in the UTM Data Frame.

There are, in fact, better ways to make the layer symbology match that of the earlier Data Frame.  You could either copy and paste the layers in from the the previous Data Frame or, to preserve this symbology for future use, you could create a "Layer File".  A Layer file contains no actual spatial data, only a description of a layer's symbology.  Once created, it can be applied to the same spatial data in other data frames or in other map documents.  This can be very useful when sharing previously symbolized data with others. Symbology is only preserved within a map document and is otherwise not attached to a data file unless it is saved separately as a layer file.  Without a layer file, a complicated color scheme for a geologic map with many colors and patterns representing rocks types, for example, would have to be recreated each time the file was added to a map document. 

To create a layer file:

  • Activate the UTM Data Frame (right click, Activate)
  • Highlight the layer with symbology you with to preserve in the TOC
  • Right-click on the layer title in the TOC, choose "Save as Layer file...", and save the layer with the default name to a location on your hard drive.

To apply a layer file's symbology:

  • Activate the Geographic Data Frame
  • Bring up the Symbology tab for the same layer Properties, click the Import button in the upper right, select "import symbology definition from another layer in the map or from a layer file" and browse to your saved layer file before clicking OK.

 

Map to be turned in: A Layout showing the Twin Cities Metro Area in the UTM projection and in geographic coordinates. Make sure to organize the layers (including hiding or deleting unnecessary layers) and standardize (and clarify) the symbology.  Title your map "Map 4: Projected And Unprojected Maps Of the Twin Cities Metro Area, Minnesota ".
 

5. Projection in ArcToolbox

5.1 Project Wizard

Up until this point, we have changed the projection of the Data Frame ("on-the-fly projection") but not the actual projections of the feature classes.  Within the ArcToolbox module, ArcGIS offers a set of tools to project data files and save them in a new coordinate system.  There are also a set of tools for assigning projection information to data that are already projected, but for which such information is lacking.  This latter process is referred to as defining a projection or defining a spatial reference, and should not be confused with the process of actually producing and saving new coordinates for a data file through projection.  We begin first with the projection tool,  "Project Wizard", to project a feature class.

If on-the-fly projection works so well, why bother projecting data to new coordinate systems?  There are two good reasons: 1) On-the-fly projections is not as rigorous as actual "hard" projection of data to new coordinates, and slight mismatches that result from on-the-fly projection of data in different coordinate systems are often unacceptable for some applications; 2) On-the-fly projection works best for data that are in geographic coordinates but not as well for data that are in projected coordinates with different datums and projections - e.g. on-the-fly projection of data that are in an Albers projection and use a NAD27 datum into a Data Frame that is displaying in a UTM projection with a NAD83 datum. The best possible co-registration among data layers is only achieved by having all data in the same coordinate system, which often requires "hard" projection of data to a new coordinate system. 

  • If you have not already done so, SAVE your map document and close ArcMap. We do this before projecting any data because we cannot work on files in ArcToolbox while they are open in ArcMap.  (One might think it would be sufficient to simply remove the data from ArcMap and leave the program open so that one might later add the newly projected data.   Not so; ArcMap sometimes gets confused when you do this, often not displaying the new file properly.) 
  • Open ArcCatalog from the Windows Start button (Start>All Programs>ArcGIS>ArcCatalog).
  • Open ArcToolbox by clicking the ArcToolbox icon
  • Navigate to "Projections and Transformations", which is located in Data Management Tools. The picture below shows the tools available for projection in ArcGIS v. 9.0/ArcInfo (some of these tools are not available with an ArcView or ArcEditor license) that permit projection and projection definition of "Features" and "Rasters" (more on these data types later in class).

We are going to project the contents of the Twin Cities feature dataset from geographic coordinates to Minnesota State Plane coordinates. The Minnesota State Plane projection is based on the Lambert Conic Conformal Projection and is divided into three zones (north, central, and south) that each have slightly different projection parameters. The zone boundaries coincide with specific county boundaries and can be found on a projection summary map published by the MN Department of Transportation (click here). The Twin Cities Metro Area is in the southern state plane zone.

  • Expand the "Feature" toolbox by clicking on the + sign next to it, then double click on "Project" tool.
  • Click on the yellow folder button adjacent to the "Input Dataset or Feature Class" form field, and navigate to the Twin Cities Metro Area feature dataset (notice that we are projecting all the feature classes within the dataset in one operation) in the Mapproj.mdb geodatabase on your Y: drive. The "Input Dataset..." and "Output Dataset..." fields should now appear similar to that shown in the picture below. Be sure to check the output dataset so that the projected data are created within the geodatabase (Mapproj.mdb) rather than the Twin Cities Metro Area feature dataset that you are trying to project. Otherwise, you will get an error message when you choose the output coordinate system.

  • Press the Select Coordinate System button , press the Select… button. This button will allow you to select a predefined coordinate system. Navigate through to: Projected Coordinate System>State Plane >NAD 1983>NAD 1983 StatePlane Minnesota South FIPS 2203.prj and click Add, then click Apply.
  • Press OK. The window now displays the Output Coordinate System.
  • The optional "Geographic Transformation" field allows us to convert the data to another datum.  Because our new file will use the same datum (NAD83) as the old file, we do not need to enter anything here.
  • Press OK.

With just a few clicks, you've projected all of the feature classes in the Twin Cities feature dataset to MN state plane (south zone) coordinates!! This process would have taken many more steps and a lot more time in earlier GIS software packages. You have the power! And you’re not afraid to wield it!!

  • Within the ArcCatalog tree, browse to the Maproj.mdb geodatatase and  you'll see that you've got a new feature dataset with copies of the all of the feature classes in the Twin Cities Metro Area feature dataset with a "_1" added to the name of each:

  • Open a new ArcMap document. 
  • Add the contents of the TCMAProject feature dataset. 
  • Open the Properties of the Data Frame and change the name to Projected. Then go to the Coordinate System tab and notice the projection is NAD_83_StatePlane_Minnesota_South_FIPS_2203, as required. Pretty cool!!

  • Notice the numbers in the lower right hand corner are not latitude and longitude any more. They are in Minnesota southern zone State Plane coordinates of meters (we could have specified the state plane projection in feet too). Right click TCMAcounties_1, and select Open Attribute Table. Navigate to the right-hand end of the table. You’ll see that two new fields have been created, Shape_Length and Shape_Area, which refer to the length (m) of the perimeter and the area (m2) of the polygon, for each feature in the TCMScounties_1 layer.

 

Question 6 : 
a) Which zip code serves the largest area in the Twin Cities Metro Area? Which serves the smallest? What are the respective zip code areas in square kilometers?

b) Name the five largest lakes in the Twin Cities Metro Area and give the area of each in square kilometers.

 

6. Defining a Spatial Reference

To project data on-the-fly, ArcMap must know the coordinate system (also called the Spatial Reference) of the data.  Depending on the data type (e.g. shapefile, feature class, coverage, images; more on this later), this spatial reference information is either stored internally (geodatabase, coverage, some shapefiles) or within a separate file.   When spatial reference information is lacking,  ArcMap cannot successfully project data with different coordinate systems.  This is a common problem for lots of GIS data, such as shapefiles created with, or for older versions of, ArcView, many aerial photographs, and maps you might yourself scan for use in a GIS.  

Tools are available in ArcToolbox to create spatial reference information files.  Spatial references can also be defined in ArcCatalog.   ArcToolbox contains "Create Spatial Reference" and  "Define Projection" tools for this procedure; in ArcCatalog we can define or alter a file's spatial Properties.  Both techniques are examined below.

6.1 Defining a spatial reference for an aerial photograph or image in ArcCatalog

  • Open a new map document in ArcMap.
  • Add the TCMAlakes.shp shapefile from the geodatabase.  This file shows lakes in the Twin Cities area.  As before, the Data Frame adopts the coordinate system of the first file added, in this case, the UTM zone 15 projection.
  • Click the Add data button and add the layer labeled MoundNE, a digital orthophoto of the NE quarter of the Mound, Minnesota, 7.5' quadrangle. Note the dialog box that pops up.

  • The orthophoto is spatially referenced because there are x/y coordinates connected to different areas (pixels) in the raster grid (you could load the image by itself and read the coordinates in the lower panel of the ArcMap window), but the image is not referenced so that ArcMap can recognize the coordinate system. ArcMap will display the layer using the x/y coordinates and values stored for each point, but the coordinates cannot be reprojected without datum and projection information.
  • When this occurs, ArcMap uses the coordinate system defined by the first object placed into the data frame. In this case, we are lucky because the TCMAlakes layer and the MoundNE_doq.jpg image are actually projected the same (UTM zone 15). Thus, if you click OK, then the image shows up in the appropriate location. If you started with a layer that was in geographic coordinates (decimal degrees), the image might end up far out of the data frame because UTM coordinates are usually in the millions of meters whereas lat/long coordinates are many orders of magnitude smaller. The only way to solve this problem would be to specify the spatial reference of the orthophoto and then reproject it and/or the other layer(s) such that they are the same.
  • To solve this problem, we must spatially reference the orthophotos before we import them into ArcMap. Close ArcMap, do not save this map document.

  • Within Arc Catalog, browse to the MoundNE_doq.jpg file and click the metadata tab in the right half of the window. You will notice that there isn't much information there because the metadata are actually contained in a html file that accompanied the downloaded image. Look in the Lab 2 data folder and open the file labeled MoundNE_doq _metadata.html. You will see that the image is projected to UTM zone 15, NAD83.
  • Right-click on the MoundNE_doq.jpg file name in the Catalog tree, choose Properties and bring up the Raster Dataset Properties dialog box and scroll down the box to the see the view shown below:

As shown above, the Spatial Reference for this file is <Undefined> - ArcMap thus had to assume a default.  To fix the problem we need to explicitly Define the spatial reference.

To define the spatial reference:

  • In the Raster Dataset Properties dialog box, Click the Edit... button, then Select.., then navigate to Projected Coordinate System>Utm>NAD 1983>NAD 1983 UTM Zone 15N.prj and click Add and OK twice.
  • To check your result, examine the dialog box.  The spatial reference for the photo is now defined, as shown below.

  • Open a new map document in ArcMap, add the TCMAlakes shapefile first and then add the Mound NE photo. 

  • Zoom into one of the lakes within the Mound quadrangle and note the correspondence of the outlines with the lakes on the photo.

  • Because the lakes in the shapefiles are polygons, they are impossible to see through when filled. To get around this problem, you can either make the polygons hollow and make the outline a lighter color. You can also make the polygon fill transparent by selecting the Display tab under the Properties menu of the TCMAlakes layer. Change the transparency value from 0% (opaque) to 100% (completely transparent)

  • SAVE the map document but do not close ArcMap.

 

Question 7:
Give two plausible reasons why the shores of the lakes on the photo do not exactly coincide with the lake outlines of the shapefile (see the image above for example).

 

6.2 Defining a spatial reference for a shapefile in ArcToolbox

A Spatial Reference can also be defined in ArcToolbox.  You might use ArcToolbox for this procedure instead of ArcCatalog if you had a lot of files to define.  ArcToolbox will allow you to do this in "Batch" mode, permitting definition of a spatial reference for many files in a single pass.  We will do it for the three remaining orthophotos, but it could be done just as easily within ArcCatalog using the procedure you just completed (by simply repeating the same procedure three times).

  • The other three orthophotos in the Lab_2_data folder represent the other three parts (quarters) of the complete Mound 7.5' quadrangle.  Like the Mound NE photos, they also lack a prj file and thus have an assumed spatial reference.  Using the above procedure verify this in ArcCatalog, but DO NOT correct it in ArcCatalog.
  • Open ArcToolbox and find the "Define Projection" tool by navigating the path: Data Management Tools>Projections and Transformations.  It's at the bottom of the "Projections and Transformation" toolbox.
  • Open the Define Projection tool, click the yellow folder button, browse to your copy of the first of the three additional orthophotos and double-click on the file name to complete the first step of the wizard, opening a dialog box like that below.

  • Press the Select Coordinate System button , then press the Select… button. This button will allow you to select a predefined coordinate system.

  • Navigate through to: Projected Coordinate Systems >Utm>North American Datum 1983>NAD 1983 UTM Zone 15N.prj, then click "Add", "OK", and "OK". Because you know that the MoundNE orthophoto is already properly projected. You could also import the projection information from it to apply to the other orthophotos. To do this, choose Import... (not Select...), and choose the MoundNE_doq.jpg file. You should see the projection information show up in the dialog box. Click "OK". Repeat this procedure (whichever variation you prefer) for each of the three orthophotos (MoundNW, MoundSE, and MoundSW).

As with the similar procedure in ArcCatalog, these steps make a .prj file that ArcMap can use to properly align data during on-the-fly projection.

  • Add the three additional orthophotos to the ArcMap document you still have open.  You should now have a map that, when magnified, resembles the one below.
Question 8:
You have downloaded digital orthophotographs and several shapefiles from the web.  List the general steps (not the detailed procedures) that you will do to assure that the data will display properly (i.e. with the proper coordinates) in ArcMap.

 

That's it....you're done.

Last updated Thursday, February 14, 2008 2:25 PM
Comments and questions to jonesjv@morris.umn.edu
Geology Discipline, University of Minnesota Morris