Lesson 3

Choose the data

CHOOSING THE DATASETS TO USE in the analysis is your first goal in this lesson. You came across a few situations in lesson 2 (with rivers, parks, and census units) in which different datasets stored similar features and attributes. How do you choose one dataset over another, which you’ll be doing in exercise 3a?

Here are some considerations:

•One dataset may represent features with a more suitable geometry type for your needs (think of cities as polygons versus cities as points).

•Among datasets of the same geometry type, one may represent features with a more appropriate degree of detail.

•One dataset may be of more recent vintage than another. Or it may be more complete, including features that the other lacks.

•One dataset may be more accurate than another when viewed against imagery or other basemaps. Or it may have been created by a more authoritative source. Or it may conform better to another dataset you’ve already decided to use.

•One dataset may need less preliminary processing of spatial or attribute data to make it ready for analysis.

Your second goal is to choose a coordinate system for the analysis. In reality, that decision is often determined by prevailing standards in your organization, or by more or less default best choices for a given study area. For example, it makes sense for you to use the state plane coordinate system of 1983 because it is specifically designed to minimize spatial distortion for more than a hundred local areas (such as yours) within the United States. Although the choice of a coordinate system bears directly on your project, exercise 3b also contains background material about coordinate systems (what they are, why they matter, how they’re managed) that is important to know for any GIS work.

Exercise 3a: Choose the datasets

In this exercise, you’ll add layers to ArcGIS Pro, compare them, and decide which datasets to use in the analysis. You’ll add this information to the data requirements table you used in lesson 2.

Open the data requirements table

1)Using Windows Explorer, open the data requirements table you’ve been working on from \EsriPress\UGIS4\ParkSite\MapsAndMore.

Choose the parcel data

Parcels are the individual property boundaries from which your candidate park sites will eventually be selected.

In lesson 2, you previewed the Parcels shapefile and the Vacant Parcels table in the City of LA folder. Since you don’t have any other parcel data, you will use these two datasets.

1)In row 1 of the table, in the Dataset column, enter Parcels.

2)In row 2, under Dataset, enter Vacant Parcels.

Recall that the Vacant Parcels dataset is a “stand-alone table.” It does not include any polygons. You must join the Vacant Parcels table to the Parcels layer using the attribute table to see vacant parcels on the map. You’ll do this table join—along with all your other data preparation—in lesson 4, but you’ll make a note of it here.

3)In row 2 of the data requirements table, in the Preparation column, enter join table to Parcels.

You must also calculate parcel acreage to satisfy your third requirement.

4)In row 3, under Dataset, enter Parcels. Under Preparation, enter calculate area.

Get started

1)Start ArcGIS Pro and open your LARiverParkSite project. You’ll work in ArcGIS Pro in this lesson and access data from the Catalog pane.

2)Make sure the Catalog pane is either open or visible as a tab on the right side of the ArcGIS Pro window. If necessary, open the Catalog pane, on the View tab, and click Catalog Pane.

3)Insert a new map and name it Lesson3a in the Catalog pane.

Choose the city limits data

The requirement in row 4 is that the new park be within the city limits. You assumed this requirement meant that you’d need spatial data representing the boundary of Los Angeles (which you have), but actually, there’s an even simpler solution.

1)In the Catalog pane, expand Folders > ParkSite > SourceData > City of LA folders and add VacantParcels.dbf to the Lesson3a map.

2)Open the table.

3)Find the CityCode field on the right side of the table. This field tells you which city each vacant parcel is recorded in.

4)Make sure you’re scrolled to the top of the table. Right-click CityCode and click Sort Ascending. The first record is Los Angeles.

5)Right-click CityCode again and click Sort Descending.

The value for CityCode doesn’t change, which tells you that every record in the field has the same value. This means that all your vacant parcels—and consequently, all your potential park sites—are prequalified as belonging to Los Angeles. It turns out that you don’t need a spatial boundary after all to guarantee that a parcel lies within the city limits.

If the Sort Ascending/Sort Descending command is dimmed, try double-clicking the CityCode field heading to sort the field. The arrow to the right of the field name will indicate whether the field is sorted ascending or descending .

6)Double-click the CityCode field heading again to switch the sorting back to ascending.

7)Close the table.

8)In row 4 of the data requirements table, under Dataset, enter Vacant Parcels.

9)Also in row 4, under Spatial Data, delete cities. Under Attribute Data, enter city name.

Add and symbolize the river data

To analyze the distances of parcels to the river, you need spatial data representing the Los Angeles River. As you saw in lesson 2, you have two datasets that might work—the River feature class in the geodatabase or the LARiver shapefile. You’ll compare these two datasets against an imagery basemap to see which dataset might be more appropriate.

1)In the Lesson3a map, change the basemap to Imagery.

2)In the Catalog pane, expand ESRI.gdb and then Hydro.

3)Drag River to the map display. The display should zoom to the extent of the River feature class, and you should see all the rivers in the LA area.

The only river you want to see on the map is the LA River. In lesson 1, you solved this problem with a definition query. Another approach is to symbolize the Los Angeles River and no other features.

4)On the Appearance ribbon, click Symbology.

One option is to symbolize data by category. This works best when the attribute of interest is a name, a description, or a number that isn’t a quantity, such as a code or a ranking. The Unique Values method assigns a different symbol to each unique value in the specified field.

5)Click Unique Values.

6)Make sure the Value Field drop-down menu is set to NAME.

7)Click the Add Values button.

8)When prompted to generate the full list, click Yes.

9)Scroll down to Los Angeles River. (The list is alphabetical.) Click on it to highlight the row, and click OK.

In both the Contents and Symbology panes, Los Angeles River is added and assigned a color from the current color ramp and a line width. By default, other features in the layer will get the symbol next to < all other values>. You don’t want to see those features, however.

10)Click the More drop-down arrow, and then click to clear the Show all other values check box. Clearing this check box will turn off the other values in the layer.

11)Click OK.

12)Click the symbol patch next to Los Angeles River to open the Formal Line Symbol pane. Then go to the Properties tab.

13)Change the color to Apatite Blue. Change the line width to 2 pt. Compare your settings to the figures, and click Apply.

On the map, only the Los Angeles River is displayed using the symbology you defined.

14)Close the Symbology pane.

Symbolize the other river data

1)In the Catalog pane, under City of LA, drag LARiver.shp to add the layer to the map.

2)Under Contents, click the LARiver symbol patch to open Symbology.

3)Change the color to Cretan Blue and the line width to 2 pt. Click Apply.

4)Close the Symbology pane.

Choose the river data

Even at this relatively small (zoomed-out) scale, it’s obvious that the two representations of the river have different spatial extents. One goes to the sea, while the other stops abruptly halfway there. Now find out why.

1)In the Catalog pane, expand the MapsAndMore folder.

2)Drag LosAngeles.lyrx to the map.

3)In the Contents pane, right-click the Los Angeles layer and click Zoom To Layer.

This is the layer file (the file of saved layer properties) that you created in lesson 1.

4)Open the layer properties of the Los Angeles layer.

5)On the Layer Properties dialog box, click Definition Query.

As reflected on the map, the symbol is a hollow yellow outline, and there is a definition query on Los Angeles.

6)Click the Source tab to see the Data Source window.

A layer file isn’t a feature class—it doesn’t store feature coordinates and attributes. Every layer in a map must point to a feature class on disk, and the same goes for a layer file. This layer file points to the City_ply feature class.

7)Close the Layer Properties dialog box by clicking OK.

You can see the likely explanation for the different extents of the two river layers: in the LARiver layer, the river stops at the city boundary. The City of Los Angeles, which provided the source data for this layer, doesn’t need to maintain features beyond its own jurisdiction. (It’s interesting that even a natural feature such as a river might be defined in part by administrative or political considerations.)

8)Zoom in on the source (the northwestern end) of the river. Set the map scale to 1:24,000.

9)Pan along the river and visually compare the two river representations. Go all the way to where the LARiver layer stops at the city boundary in the southeast.

The two representations of the river aren’t identical, but they’re close. At large scales, such as 1:2,500, the river in LARiver is noticeably more sinuous and conforms more closely to the imagery, but for your analysis, either dataset is perfectly adequate. You’ll use the LARiver dataset, which is already clipped to your area of interest.

10)In row 5 of the data requirements table, under Dataset, enter LARiver.

When you previewed this dataset in lesson 2, you saw that it was composed of 265 features. You can combine them into a single feature with a data processing operation called Dissolve. It will help get the data ready for lesson 6, in which you must create a half-mile buffer (proximity zone) around the river.

11)In row 5, under Preparation, enter dissolve.

12)Save your project.

Add and symbolize the park data

To analyze the location of candidate sites in respect to existing parks, you need spatial data representing parks. You’ll compare the Parkland dataset you used in lesson 1 to the Parks dataset you previewed in lesson 2.

1)In the Contents pane, turn off the River layer and click its side arrow to collapse it.

Leave the LARiver layer turned on.

2)In the Catalog pane, under ESRI.gdb, expand Landmark.

3)Drag Parkland to the map.

4)In the City of LA folder, drag Parks.shp to the map.

The order of layers in Contents should look like the figure. Your symbology for the parks layers may be different.

5)Under Contents, right-click the symbol patch for the Parks layer to open the color palette. Change the fill color to Chrysoprase.

6)Right-click the Parkland symbol patch, and change its fill color to Tzavorite Green.

7)Select the Parks layer, and click the Appearance tab. In the Effects group, change Layer Transparency to 50%.

The symbology makes it clear where the two layers agree and disagree on the representation of features.

Examine a park in two layers

You can start with a close look at an example park. Your map should still be zoomed in to where the river crosses the city limits in the southeast. The scale should still be 1:24,000.

1)On the Map tab, in the Inquiry group, click the Measure tool and click Measure Distance.

2)Click the Choose Options button , click Distance Units and Miles, and if necessary, click to clear Feet.

You’re going to follow the river north about 2.25 miles.

3)Click at the end of the river to start the measurement. Move your pointer north along the river. (You don’t have to drag it.)

4)As you move the pointer, the Segment length value changes in the Measure dialog box.

5)When you reach the top of the map display—assuming your measurement is still less than 2.25 miles—press and hold the C key on your keyboard.

The pointer changes to the hand icon .

6)Pan north and release the C key to continue the measurement.

7)When the length of the measurement reaches about 2.25 miles, double-click to end the measurement.

Half a mile due east of the end of your measurement is a park.

8)Switch back to the Explore tool on the Map tab, and click on the center of the park where the two park layers overlap to identify the park.

Its name is Pecan Playground. The attributes in the pop-up window come from the Parks layer, not from Parkland, because of the layer order in the Contents pane.

By default, the topmost layer is identified. You can choose other layers from the drop-down menu under Explore.

9)In the lower-right corner of the Identify window, click the Zoom tool , which will zoom to the feature Pecan Playground.

The map zooms in on the park.

10)Create a new bookmark named Pecan Playground.

11)Close the Identify window and the Measure dialog box.

12)Turn the Parkland layer off and on a few times.

Neither feature conforms to the image: both encroach, for example, on the surrounding streets. The Parks feature seems better, however, because it includes the swimming pool and the play area at the north end of the block. You’ll come back to this place in lesson 5 to do some spatial editing.

13)Under Bookmarks, click Dodger Stadium to zoom to it. Your view is now centered on Dodger Stadium. Compare the Parkland and Parks layers.

Choose the park data

You’ll follow the river upstream and look at some more parks to help you reach a decision.

1)In Contents, drag the Los Angeles layer directly underneath the LARiver layer.

2)Set your map scale to 1:36,000. Start panning northwest along the river’s course.

You’ll soon come to Griffith Park, the vast park at the river’s elbow. Both the river and the city limits run along the park’s edge. Notice that to the east (Glendale) and north (Burbank) of the city limits, there are features from the Parkland layer, but no corresponding Parks features. This is another case of jurisdiction. The City of Los Angeles, which supplied the Parks dataset, doesn’t maintain data for parks in other cities.

This difference in managing data raises a question. The new park must be a quarter mile away from existing parks: Does that mean any park in any city or just any park in Los Angeles? You’ll interpret it to mean any park in any city, because, by and large, there are no residency requirements for park use. On that basis, you should choose the Parkland feature class for your analysis. You don’t know if it’s more or less accurate on the whole (you only examined one park), but it covers a part of your area of interest that the Parks feature class doesn’t.

3)In row 6 of the data requirements table, under Dataset, enter Parkland.

Add more park data

In lesson 2, you previewed the NewParks feature class and Vista Hermosa Park, which has been completed but is not in NewParks.shp. You should find out if these features are included in the Parkland layer.

1)In the Catalog pane, expand the ParkData folder and add NewParks.shp to the map.

2)In Contents, drag the NewParks layer directly above the Parks layer.

3)Right-click the symbol patch for the NewParks layer, and click Macaw Green.

4)Zoom to the NewParks layer.

How? In the Contents pane, right-click the layer, and click Zoom To Layer.

5)Turn the NewParks layer off and on a couple of times.

You can see parks on the basemap imagery, but they’re not represented in the Parkland layer. Another data preparation task in lesson 4 will be to add these missing parks into the Parkland feature class.

6)Zoom to Vista Hermosa Park using the bookmark you created in lesson 2.

There’s no Parkland feature here, either, or you’d be looking at it. In lesson 5, you’ll create a park feature in this location.

7)In row 6 of the data requirements table, under Preparation, enter edit features.

Choose the census data

In lesson 2, you decided that you would represent neighborhoods with census units. Your two choices are block groups or tracts. Block groups meet all your attribute needs, either directly or indirectly. Because they are smaller than tracts, block groups also portray demographic patterns in more detail.

1)In rows 7 through 10 of the data requirements table, under Dataset, enter block groups.

Population density and age under 18 attributes don’t exist as such in the block_groups table, but you can derive them from other attributes.

2)In rows 8 and 9, under Preparation, enter calculate.

The last analytical requirement is to count how many people live within a quarter mile of each proposed park site. In lesson 6, you’ll do this by creating quarter-mile rings (buffers) around the proposed sites, counting the census block points in each ring, and summing their populations.

3)In row 11, under Dataset, enter block centroids.

Choose the final map data

The last rows of the table list data that you may need for cartographic reasons in the final map.

Row 12 lists political boundaries. Your source data has feature classes of cities, counties, and states. Until you design your final map, you won’t know for sure, but you can foresee a likely need for city and county boundaries, but not for state boundaries.

1)In the Catalog pane, under ESRI.gdb, expand Boundary.

In lesson 2, you previewed City_ply and saw that its extent covered the entire United States. That’s more data than you’ll need. You never previewed the County feature class, so do that now.

2)Under Boundary, drag the County feature class to your map.

3)Zoom to the layer and look at the attribute table.

Note that the attributes include both the county and state names along with census attributes similar to those you’ve seen in other layers.

4)Turn off the County layer and close its attribute table.

5)In row 12 of the data requirements table, under Dataset, enter City_ply. Press Enter to start a new line and enter County.

6)In row 12, under Preparation, enter reduce extent next to both City_ply and County.

In row 13, in the Spatial Data column, you have a need for roads listed.

7)In the Catalog pane, under ESRI.gdb, expand Transport and drag Mjr_rd to your map.

The data covers the Greater Los Angeles Area, which is appropriate for your map, but you also know that roads are available on the Topographic basemap.

8)Change the basemap to Topographic using the Basemap button on the Map tab, in the Layer group.

9)Zoom in so you can see the difference between the Mjr_rd layer and the basemap. Optionally, switch off between the Imagery basemap and the Topographic basemap to compare accuracy.

The Topographic basemap is more detailed and more accurate. The layer is also updated automatically for you by Esri, so your maps will stay current into the future. The roads on the basemap are also already cartographically styled nicely with labels, so it will be a lot easier to work with. One consideration with basemaps is that they are essentially an image—you can’t query or select from them, and you can’t use them as the input for tools. Your project, however, doesn’t include roads in the criteria; you simply want roads for reference and cartographic purposes. For these reasons, you will use the Topographic basemap to display roads rather than using the Mjr_rd dataset.

10)In row 13 of the data requirements table, under Dataset, enter basemap.

Relief and imagery are also available as online basemap layers.

11)In rows 14 and 15 of the data requirements table, under Dataset, enter basemap.

Save your work

All the datasets you need for the analysis have been specified, so you can save and close your work.

1)Save and close the data requirements table.

2)Zoom to the Los Angeles layer.

3)Close the Lesson3a map.

4)Save your project.

5)If you are continuing to the next exercise now, leave ArcGIS Pro open; otherwise, close ArcGIS Pro.

In the next exercise, you’ll take a step back from the project. Analysis operations take place within a particular coordinate system, and choosing that system is an important part of setting up the project. To make a good choice, you should have some background knowledge of what coordinate systems are and how they’re managed in ArcGIS Pro.

Exercise 3b: Choose a coordinate system

A coordinate system is a mesh of perpendicular intersecting lines superimposed on a surface. The point of intersection of any two lines is a unique location, which can be specified with two values (a coordinate pair). The values are measurements from a given reference point in a given unit of measure. The unit of measure may be an angular unit, such as degrees, or a length unit, such as feet or meters.

Coordinate systems can be applied to any surface, but you’re interested in the surface of the earth. If you represent the earth with a spherical model, such as a globe, you have a curved surface to deal with. If you represent it with a flat model, such as a map, the surface is flat.

A geographic coordinate system is applied to round or roundish earth models. Geographic coordinate systems use an angular unit of measure because angles are better for measuring locations on a curved surface. A projected coordinate system is applied to flat earth models. Projected coordinate systems use a length unit of measure.

Every usable spatial dataset has a coordinate system. The locations of its features are specified by coordinates that are correct within its own framework but that would be wrong or absurd in another system. Fifty degrees of arc isn’t the same as 50 meters, which isn’t the same as 50 feet. Knowing the coordinates of a point doesn’t tell you where the point is on the earth unless you also know the coordinate system.

There are many different geographic and projected coordinate systems. As long as two datasets use the same system, their features will align correctly on a map. Datasets with different systems must be reconciled. This basically means that one dataset’s coordinate system must be mathematically converted, or projected, to the other. This can be done with data processing tools. ArcGIS Pro does it automatically when you add layers to a map. What ArcGIS Pro reconciles, however, are the coordinate systems of the layers in the map, not the coordinate systems of the datasets that the layers point to. This reconciliation of coordinate systems in a map is called “on-the-fly” projection.

In this exercise, you’ll see how ArcGIS Pro manages on-the-fly projection of coordinate systems. You’ll also look at how map display and map measurements change with different geographic and projected coordinate systems. Finally, you’ll select a coordinate system for your analysis project.

Add data and check the coordinate system of a dataset

Every spatial dataset has a coordinate system. Now you’ll see how to find out what it is.

1)If necessary, start ArcGIS Pro and open your LARiverParkSite project. Create a new map, and rename it Lesson3b.

2)In the Catalog pane, under ESRI.gdb, expand Boundary and drag the State feature class to the Lesson3b map.

3)In the Layer Properties dialog box of the State layer, click Source. Scroll down to Spatial Reference and expand the group.

The window displays the current coordinate system: GCS WGS 1984.

“GCS” stands for geographic coordinate system. “WGS 1984” is the World Geodetic System of 1984, also called WGS84.

Under the name of the coordinate system are details giving the essential properties of the system:

•Its angular unit of measure (degrees of angle)

•Its prime meridian (Greenwich)

•Its datum (D WGS 1984)

WGS 1984 is a geographic coordinate system. Geographic coordinate systems, and datasets that use them, are said to be “unprojected.” If you checked the other feature classes in ESRI.gdb, you’d find that they all have this same geographic coordinate system. The datum includes the name and dimension of a particular spheroid, or rounding of the earth model. For more information on datums, see the sidebar “Datums.”

4)Close the Layer Properties dialog box.

5)Hover over the map.

At the bottom of the application window, the coordinates of the pointer are reported in decimal degrees. The first coordinate (longitude) tells you the angular position east or west of the prime meridian. The second coordinate (latitude) tells you the angular position north or south of the equator.

6)In the Contents pane, double-click Lesson3b to bring up Map Properties (or right-click Lesson3b and click Properties).

7)In the Map Properties dialog box, click Coordinate Systems. Confirm that the coordinate system is GCS WGS 1984, the same as the State layer.

A map is a container for layers in a project. You can insert more maps to manage separate sets of data within your project. Why would you want to do that? It’s sometimes a matter of layout: you may want to print a single map sheet that contains a main map, an overview map, an inset map, or some other combination of views. (You’ll see how maps work in layouts in lesson 8.)

One crucial function of a map is to enforce the spatial alignment of the layers it contains. When you first add a layer to the map, it adopts that layer’s coordinate system as a standard. Any layers added after that are converted (projected) by ArcGIS Pro into that same system. This process of on-the-fly projection happens automatically as new layers are added.

8)Close the Map Properties dialog box.

Project data on the fly

To see how on-the-fly projection works, you’ll insert a second map, check the coordinate system, and then copy the State layer to the new map. To see the contents of both maps at the same time, you must open another map.

1)On the Insert tab, add a second map and name it Lesson3b Web Mercator.

2)Reposition Lesson3b Web Mercator on the right side of the Lesson3b map by dragging the tab of Lesson3b Web Mercator to the center of the map.

When you release the mouse button, you’ll have the two maps side by side. A second map is usually added in order to add a new set of data layers in a new area of interest. In this case, the second map will allow you to compare two different map projections side by side.

3)On the View tab, click Link Views, and then on the drop-down menu, click Center And Scale.

4)Open Map Properties for the Lesson3b Web Mercator map.

5)Click Coordinate System. Note that it is WGS 1984 Web Mercator Auxiliary Sphere. This is the projection of the Topographic basemap. Close the Map Properties window.

6)In the Contents pane of the Lesson3b map, right-click the State layer and click Copy.

7)Click the Lesson3b Web Mercator tab to make it active, and then right-click Lesson3b Web Mercator in the Contents pane to paste the State layer under Contents.

8)Examine the coordinate system of the Lesson3b Web Mercator map. Notice it changed to the projection of State: WGS 84. The two maps are identical and have the same projection.

With this new projection in the Lesson3b Web Mercator map, the basemap, which is in WGS 1984 Web Mercator Auxiliary Sphere, is now being projected on the fly to WGS 84.

9)Change the coordinate system of the Lesson3b Web Mercator map back to WGS 1984 Web Mercator Auxiliary Sphere in Map Properties. Click OK.

10)Compare the map appearance, and notice especially the shape of Alaska.

Make an area measurement

Coordinate systems change measurements as well as appearances. The Mercator projection you’re using now is a conformal projection, which means that it shows shapes correctly. (Look at Alaska on a globe, and you’ll see this is true.) But, it distorts area measurements—quite significantly in extreme latitudes.

1)Zoom in on Alaska—it doesn’t matter how far exactly. Because the two maps are linked, both will zoom together.

2)On the Map tab, click the Measure tool and click Measure Features.

3)Under Options for the Lesson3b Web Mercator map, change the area Units and Mode. Click Area Units > Square Miles. Click Square Feet to turn it off. Then click Mode > Planar.

4)Under Options for the Lesson3b map, change the area Units and Mode. Click Area Units > Square Miles. Click Square Feet to turn it off. Then click Mode > Geodesic.

5)Click anywhere inside one of the two maps. Leave the Measure dialog box open.

6)Open the State layer attribute table.

7)Select the record for Alaska. (It should be the second one.) Scroll all the way to the right to see the SQMI attribute.

The value in the table, which is correct, is 581,369 square miles. (You can find other values cited in reference sources, but they’re similar to this one.) That means that the map measurement is an exaggeration by a factor of more than five.

8)At the top of the table window, click the Clear button . Close the table.

The Measure tool is doing its job correctly: it has measured the area of Alaska as distorted by the Mercator projection. (The Mercator projection has its virtues, but making area measurements in high latitudes is not one of them.)

Change the map’s coordinate system

You can change the map’s coordinate system whenever you want—before or after adding layers.

1)Open the map properties of the Lesson3b Web Mercator map.

2)Click Coordinate System. Change the projection from WGS 1984 Web Mercator Auxiliary Sphere to Alaska Albers Equal Area Conic, which is found under the Projected coordinate system > Continental folder > North America folder.

This system, unlike the Mercator, represents area (the size of features) correctly.

The new coordinate system is applied to the map. Both the basemap layer and the State layer are projected (reprojected, if you like) on the fly.

Coordinate systems are organized in folders. The two main folders are Geographic Coordinate Systems and Projected Coordinate Systems. Both folders contain several subfolders. Both the Web Mercator Auxiliary Sphere and Alaska Albers Equal Area Conic coordinate systems are in the Projected Coordinate Systems folder.

The more you learn about coordinate systems, the more useful this folder structure becomes. At the outset, it can be a bit overwhelming.

Choosing a coordinate system

A suitable coordinate system for your project depends largely on your area of interest and may also depend on whether you want to preserve a particular spatial property, such as shape or area, without distortion. A few systems are common standards. For example, web-based maps typically use the Web Mercator coordinate system. Local-area maps within the United States are frequently based on the state plane coordinate system, while local-area maps around the world (including the United States) often use the Universal Transverse Mercator (UTM) system.

On the Map Properties Coordinate System tab, you can search for coordinate systems covering only your area of interest. (This doesn’t mean all the results are appropriate for your map. A system designed for a world map is displayed with every spatial filter but may not be a good choice for a local map.) You can also enter text to find coordinate systems that include the text as part of their name. This is often a good way to find systems specifically suited to your area of interest.

For more information, see the sidebars on “Geographic coordinate systems” and “Projected coordinate systems.”

Geographic coordinate systems

A geographic coordinate system (GCS) is based on a spheroidal model of the earth. Sometimes a perfect sphere is used, but usually a slightly squashed “oblate spheroid” is used to reflect the fact that the earth bulges at the equator and is flattened at the poles.

The reference lines of a geographic coordinate system are parallels and meridians. Parallels are lines that circle the globe parallel to the equator. Meridians are lines perpendicular to the equator that converge at the poles. By convention, the origin of the system (its 0,0 coordinate) is the intersection of the equator and the prime meridian, the meridian passing through Greenwich, England.

Geographic coordinates, commonly called latitude-longitude values, are measurements of angle, not distance. Angles are a constant unit of measurement on a sphere, whereas distances are not (because meridians converge).

Angle measurements are usually expressed in degrees, minutes, and seconds. A degree has 60 minutes; a minute has 60 seconds.

Latitude is angular position north or south of the equator. The equator is 0° latitude, the North Pole is 90° north, and the South Pole is 90° south.

Longitude is angular position east or west of the prime meridian. The prime meridian is 0° longitude. Its anti-meridian (on the other side of the world) is both 180° east and 180° west.

A latitude-longitude pair defines a unique position on the earth’s surface.

The unique location of Dodger Stadium, for example, would be written like this: 34°4′26″N, 118°14′27″W and spoken like this:

“34 degrees, 4 minutes, 26 seconds north latitude; 118 degrees, 14 minutes, 27 seconds west longitude.”

For computer calculations, these values are converted to decimals. The location of Dodger Stadium in “decimal degrees” is 34.073, –118.24. The minus sign is used for west longitude and south latitude.

The fact that there are many different geographic coordinate systems is a source of trouble for GIS users. What makes two systems different is disagreement about the exact latitude-longitude values of particular locations. Why is there disagreement about that? In simple terms, it’s because different spheroid models of the earth have been developed over time by different earth scientists using different technologies. Changing the shape or size of the model ends up changing the coordinates of points on the surface—usually not by very much, but sometimes, in sensitive applications, enough to be of concern. This issue is taken up in more detail in the “Datums” sidebar later in this lesson.

Projected coordinate systems

The earth’s surface can be modeled very well on a spheroid, but not as well (except over small areas) on a plane. To make a map, you more or less have to flatten a sphere, which is like squaring a circle, only harder. It can’t be done without radically adjusting the spatial properties and relationships of features on the surface: their shapes, sizes, and relative distances and directions.

The name for any such radical adjustment is a map projection. A projection is a mathematical formula (there are lots of different ones) for translating the world into flat space. All map projections introduce spatial distortion. They are variously designed to minimize certain kinds of distortion or to distribute it in certain ways over the map surface. Some projections correctly preserve feature shapes but distort their areas. Some preserve areas but distort shapes. Some compromise. Some have special properties, such as keeping true distance measurements from a single point to all others, or ensuring that courses of constant compass bearing are plotted as straight lines. The smaller the area being mapped, the less distortion there is of any spatial properties. Areas up to medium-size countries (say about the size of Nigeria or Bolivia) can be mapped with distortion low enough to be insignificant for most purposes.

A projected coordinate system consists of a map projection, a length-based unit of measure, an origin point for measurements, and other parameters, such as standard lines that define the distortion pattern on the map. Manipulating these parameters is what allows you to customize a coordinate system for a specific area of interest. Because a projection is applied to a particular spheroid and its definition of latitude-longitude values, a projected coordinate system also includes, or presupposes, a geographic coordinate system.

The idea of projection includes both going from a geographic (unprojected) system to a projected system, and going from one projected system to another (sometimes called reprojection). To go from one projected system to another, ArcGIS Pro undoes the map projection, goes back to the underlying geographic coordinate system, and applies a new projection to it. ArcGIS Pro stores thousands of map projection formulas and can run these calculations quickly.

When you add an unprojected dataset to a map (that is, a dataset that stores feature coordinates as latitude-longitude values), the data still must be projected in some sense to be viewed as a flat map on your monitor. In ArcGIS Pro, this default “pseudoprojection” has the display properties of a map projection (specifically, the Plate Carrée), but none of the other properties or parameters of a projected coordinate system.

Measure Alaska again

Equal-area projections (of which the Albers Equal Area Conic is one) preserve the spatial property of area, or size. The trade-off is that they don’t represent shapes correctly. Because this projection is customized for Alaska, however, most distortion of all types is pushed outside the area of interest.

1)On the Map tab, click the Measure tool, and then click Measure Features to measure Alaska again using this new projection.

The area is given accurately as 581,333 square miles.

2)Switch from the Measure tool back to Explore .

3)Zoom to the full extent .

This coordinate system is obviously unsuitable for a world map. It doesn’t show the whole world, for one thing. The price for mapping Alaska with very low distortion is that places such as Australia and South America are severely distorted—which is fine, as long as you keep your map zoomed in to Alaska.

Now you can summarize how things stand with the layers in this map:

•The geographic coordinate system of the State layer is GCS WGS 1984. It has no projected coordinate system.

•The projected system of the basemap layer is Web_Mercator. Its underlying geographic system is GCS WGS 1984.

•The map has been set to the Alaska Albers Equal Area Conic projected system. Its underlying geographic system is GCS North American 1983.

•The State and basemap layers have been projected on the fly into the Alaska Albers system.

No matter how often you change coordinate systems, ArcGIS Pro keeps the data in alignment.

4)Change the projection back in Lesson3b Web Mercator to WGS 1984 Web Mercator Auxiliary Sphere (under Projected > World) from Alaska Albers Equal Area Conic.

5)On the View tab, click Link Views to turn off link views.

6)Close the Lesson3b Web Mercator map.

Reminder: it is still in the project under the Maps folder.

Add another layer

Now you can get back to your true area of interest.

1)Open the Lesson3a map, copy the LARiver layer, and paste the layer into Lesson3b.

2)Zoom to the LARiver layer and turn off the State layer.

A coordinate system meant for Alaska (or the world) isn’t appropriate for Southern California. Local mapping needs across the United States are served by a system called the state plane coordinate system. This isn’t a single coordinate system for the entire country, but a patchwork of systems, each of which covers a state or part of a state. Together, they ensure that for whichever part of the country you want to map, you get uniformly low distortion. California is divided into six state plane zones, with Los Angeles falling into zone 5.

The state plane coordinate system divides states into zones and applies a locally optimized coordinate system to each zone. Vertical zones are based on the Transverse Mercator projection. Horizontal zones are based on the Lambert Conformal Conic projection.

Change the map’s coordinate system

You’ll change the map’s coordinate system again, this time to State Plane California Zone 5.

1)Open the map properties of Lesson3b.

2)On the Coordinate Systems tab, expand the Layers folder.

Within this folder are the three native coordinate systems of the layers in the map.

3)Select the coordinate system NAD 1983 StatePlane California V FIPS 0405 Feet.

This is the coordinate system used by the LARiver layer. (Strictly speaking, it is the coordinate system of the LARiver shapefile.)

This resets the current coordinate system of the map. The state plane coordinate system for California zone 5 is a projected coordinate system based on a Lambert Conformal Conic map projection.

Its underlying geographic coordinate system is North American 1983. This is different from the WGS 1984 system used by both the State and basemap layers.

4)Under General, change the Display units to US Feet to match this new projection.

5)Click OK.

Your map should look like the figure. The state plane coordinate system for California zone 5 is the one you’ll use for your analysis. Notice the coordinate values below the map have changed from decimal degrees to US Feet, consistent with the new coordinate system.

6)Close all the Lesson3 maps and any open tables.

7)Save your project.

8)Continue to the next lesson or close ArcGIS Pro. Save your changes if prompted.

In lesson 4, you’ll resume work on the project by taking care of the preparation tasks you listed in the data requirements table.

Datums

A geographic coordinate system is defined by three things: an angular unit of measure (usually degrees), a prime meridian (usually Greenwich), and a datum. The datum is the part that gives people trouble. To understand it, start with the shape of the earth.

The earth isn’t a perfect sphere, or even a mathematically regular spheroid. It’s a lump with an uneven shape owing to different concentrations of mass (and therefore unequal gravity) over its surface. In addition, it has topographic features such as mountains and valleys.

When the spatial positions of features are determined—as was formerly done by survey, and is now mostly done by satellite—they are first determined on the earth’s surface. These raw measurement values are then mathematically “leveled” to a geoid. A geoid is the (still gravitationally lumpy) shape that the earth would have if it was covered by a mean sea level surface—in other words, if it had no topography.

The shape of the geoid, however, is too complex to be a working model. So the next step is to move the measurements from the geoid to a spheroid: a model with a regular, nonlumpy shape.

That’s where the datum comes in. The datum is two things: first, it’s a chosen spheroid, which could be WGS 1984, GRS 1980, Clarke 1866, Bessel 1841, or a number of others. (The world is standardizing on the GRS 1980 spheroid but isn’t all the way there yet.) Second, it’s a mathematical orientation, or “fit,” of the geoid to the spheroid. In the transfer of measurements from geoid to spheroid, some error will be introduced because the lumps must be smoothed out. How that error is distributed is the “fit.” One approach is to make the fit really good for one part of the world, such as North America, and not to worry about the rest. That’s a local datum. It’s designed to maintain high accuracy for measurements over a limited area. The other approach is to average the error over the whole surface. That’s an earth-centered datum. It’s designed to maintain high accuracy for the world as a whole.

When two geographic coordinate systems are different, it’s usually because the datums are different (which, in turn, is either because the spheroids are different or the fit is different). When you get a coordinate system warning in ArcGIS Pro, one possibility is to ignore it and leave the data slightly out of alignment. Depending on your needs for accuracy, this may be an entirely sensible choice. The amount of misalignment depends on the datums involved and the part of the world being mapped, but in the mismatch that North Americans usually encounter (between the World Geodetic System of 1984 and the North American Datum of 1983), it typically doesn’t exceed a few feet. At most scales, the difference isn’t noticeable.

The other option is to reconcile the systems through a geographic transformation. Transformations are often done in conjunction with a coordinate system projection. Like projections, they can be permanently applied to datasets with data processing tools, or they can be done on the fly in ArcGIS Pro. Transformations require some expert knowledge. There are default methods to convert one spheroid to another, but there aren’t default fits, because the right fit depends on your area of interest. The table on “geographic (datum) transformations: well-known IDs, accuracies, and areas of use,” at https://desktop.arcgis.com/en/arcmap/latest/map/projections/pdf/geographic_transformations.pdf, can help you find the right fit for an area of interest.