Lesson 6

Conduct the analysis

ALL YOUR PREPARATORY WORK has led to this stage: the analysis. Your original feature class of parcels was like an assembled 796,000-piece jigsaw puzzle covering the entire city, with each piece representing an individually owned property. In lesson 4, you cut this down to about 29,000 parcels (those that were vacant) and then to around 14,000 (those that were adjacent). Now you’ll continue taking pieces away, removing unsuitable parcels until you’re left with only those candidates that meet the criteria established in lesson 2:

•a vacant land parcel at least one-quarter acre in size,

•within the LA city limits,

•within half a mile of the LA River (the closer the better),

•at least a quarter mile from the nearest park,

•in a neighborhood in which

•population density is at least 8,500 people per square mile,

•at least 22 percent of the population is under 18 years old, and

•median household income is $50,000 or less, and

•preferably serving the most people within a quarter-mile radius.

But how does this translate into a method and workflow? How do you take these requirements and match them to appropriate geoprocessing tools in the right sequence? Planning a workflow means identifying the tasks, finding a suitable tool for each one, and understanding the inputs and outputs so you can order the operations in a logical, efficient way.

Most of the tasks in your analysis—and this is typical—fall into just a few categories: distance problems (what’s near what), topological problems (what crosses, touches, or contains what), attribute query problems (what has this or that value), and overlay problems (what areas are common to features in different layers). Each of these problem categories is associated with its own special set of tools.

Before beginning an involved analysis, you may find it helpful to draw a rough plan or diagram. Any medium will do—paper, an app, a whiteboard—as long as you’re prepared to make revisions as you go. A typical working plan (the one you’ll follow) could be sketched out to look something like the diagram shown here. This sketch depicts the broad general approach, although there are bound to be a few additional twists and turns along the way.

Looking at the sketch of the analysis, you can see that this plan represents only one of several possible approaches to the problem. The plan you’re following here is reasonable and efficient. It uses a variety of important geoprocessing tools and has the visual benefit of letting you see the analysis unfold, as land areas are progressively stripped away from consideration. It is certainly possible, however, to reach the same or similar conclusions using different combinations of tools.

Your analysis results depend on the state of the data in the project geodatabase. If you’ve successfully completed all the exercises up to now, you should be in good shape. If necessary, download the lesson 5 results from the book resource web page, at esri.com/UnderstandingGIS4.

Exercise 6a: Establish proximity zones

In this exercise, you have two objectives. First, you want to define the area of interest as a half-mile zone around the river. Any parcels falling outside this zone will be excluded from consideration. Second, you want to define quarter-mile zones around each park. Any parcels in the area of interest that lie within these internal exclusion zones will also be dropped. You’ll use one simple operation—buffer—to significantly reduce your hunting grounds.

Create a results geodatabase

When you generate outputs, you need to store them somewhere. In lesson 4, you decided to keep your input data in one geodatabase and your output data in another geodatabase. (See the sidebar “Project database considerations” in lesson 4.) It’s a matter of preference, but our thought is that it’s easier to keep the project organized if inputs and outputs are separated. You may want to share the input data with someone else so the analysis can be run independently. You may want to repeat the analysis with different parameters (something that you’ll do in lesson 7). The more feature classes you add to a geodatabase, the harder it is to keep track of what they represent and what purpose they serve.

1)Start ArcGIS Pro, and open your LARiverParkSite project.

2)Insert a new map, and name it Lesson6a.

3)Close any other open maps by closing their associated tabs above the map. You can always get back to these maps by double-clicking them in the Catalog pane, in the Maps group.

4)In the Catalog pane, expand the Databases folder. It should contain your LARiverParkSite project geodatabase.

5)Right-click the Databases folder, and click New File Geodatabase. By default, the dialog box should already be set to the project folder so that the new geodatabase will be created in this folder.

6)Name the geodatabase AnalysisOutputs, and click Save.

Note that the new geodatabase has been added to the list of databases in the Catalog pane.

You’ll want to make AnalysisOutputs the current workspace so that your output data is put here by default.

Set the current workspace

When working with geoprocessing tools in ArcGIS Pro, it is convenient to set a default location to save the outputs (in this case, the new AnalysisOutputs geodatabase). This default output location is known as the “current workspace” and can be set as an environment setting for the project. The “scratch workspace” is a second environment setting option for the output location of temporary data—data that you don’t need to keep and maintain. By setting both the current workspace and scratch workspace, you save yourself the step of browsing to a new output location every time you run a tool.

1)On the Analysis tab, in the Geoprocessing group, click the Environments button .

2)Click the browse button next to the Current Workspace box, and browse to the new AnalysisOutputs geodatabase. Click OK.

3)Set Scratch Workspace to the same geodatabase.

4)Confirm that the Output Coordinate System is still set to NAD_1983_StatePlane_California_V_FIPS_0405_Feet.

5)Compare to the figure, and click OK on the Environments dialog box.

Buffer the LA River

A buffer is a zone around a map feature measured in distance units. You’ll use the Buffer tool to create a half-mile proximity zone around the LA River. For more information about the Buffer tool and other analysis tools, see the sidebar “Essential GIS analysis tools.”

1)In the Catalog pane window, expand LARiverParkSite.gdb, and add LARiver to the map.

Note that the layer is likely symbolized in a color other than blue. During analysis, you’re more interested in geoprocessing results than map appearance. For that reason, you’re not going to symbolize input and output layers carefully at each step along the way. You’ll do it as the need arises, but often you’ll just accept the ArcGIS Pro defaults.

2)On the Analysis tab, in the Tools gallery, click Buffer . The Geoprocessing pane opens with the Buffer tool parameters displayed.

3)In the Geoprocessing pane, click the Input Features drop-down arrow, and click LARiver.

4)Name the output feature class LARiverBuffer. Because you’ve set the AnalysisOutputs geodatabase as the scratch workspace, the output will be saved to this location by default.

5)In the Distance text box, type 0.5. Click the drop-down arrow for linear units, and click Miles.

6)Compare your settings to the figure, and click Run.

When the tool is finished running, a message at the bottom of the Geoprocessing pane notifies you that the Buffer tool completed successfully, and the new LARiverBuffer layer is added to the map.

Also, note that a warning icon appears next to the Output Feature Class text box. Hovering over the icon notifies you that the feature class exists so that it will be overwritten if you rerun the tool.

7)Close the Geoprocessing pane.

8)Confirm the presence of the new feature class by expanding the Databases folder in the Catalog pane and looking in AnalysisOutputs.gdb.

Essential GIS analysis tools

Analysis tools

The tools shown here are by no means a complete list, but they include several of the ones used most often in GIS analysis. It’s hard (and probably unnecessary) to give an exact definition of what makes a tool an “analysis” tool. GIS practitioners solve problems of many different kinds, and most of these problems have facets that involve spatial relationships among geographic objects: how far A is from B, how many A’s are close to B, which areas are common to A and B, how you get from A to B, and so on. Analysis tools quantify these relationships among features and their attributes.

Query tools

Query tools answer questions of the form “Which features meet such-and-such condition?” Select By Attributes selects features according to an attribute value or combination of values. Select By Location selects features according to their spatial relationship to features in another layer (or sometimes the same layer). Spatial relationships include intersection, containment, adjacency, and distance.

Query tools can be accessed on the Map tab, in the Selection group, or as geoprocessing tools (Select Layer By Attribute and Select Layer By Location).

Proximity tools

The Buffer tool creates a feature class of polygons at a specified distance around input features. It is often used to draw an exclusion zone around features (no A’s should be allowed within a mile of B) or to define an area of interest (you want to look for A’s within only a mile of B).

The Create Thiessen Polygons tool creates a feature class of contiguous polygons around input point features. Each polygon’s shape is defined by proximity to the nearest point. Thiessen polygons can be used to define allocation areas (which areas are closer to A than to any other point, which are closer to B than to any other point, and so on).

The Near tool finds the nearest feature in one or more specified layers to each feature in the input layer. It writes the distance as a new attribute to the input layer table.

Overlay tools

The basic purpose of overlay tools is to find common areas between different layers. Overlays answer questions about where A and B, each with unique and important attributes, overlap. In contrast to queries, which return selections of existing features, overlay tools create new features that have the attributes of both input layers.

The different overlay tools—Intersect, Union, Identity, and Erase are the most common—perform variations on the same basic process. They differ with respect to how much area from the input layers is included in the output feature class: common area only (Intersect), all areas whether common or not (Union), one input layer’s area only (Identity), or one input layer’s area minus the common area (Erase).

Spatial Join

Spatial Join is like a Select By Location query with the added benefit of joining source layer attributes to target layer features.

Buffer the parks to a quarter mile

Now you’ll add the Parks data and draw a quarter-mile buffer around each park. These buffers will encompass areas that are close to existing parks and are therefore out of consideration.

1)From the LARiver geodatabase, drag Parks to the map window.

2)In the Contents pane, right-click LARiverBuffer, and click Zoom To Layer.

3)Turn off the LARiver layer.

You can already see a lot of overlap between existing parks and the river buffer zone. You must still extend the exclusion zone a quarter mile farther out on all sides of each park.

4)On the Analysis tab, in the Tools gallery, click Buffer.

5)Drag the Parks layer from the Contents pane into the Input Features text box of the Buffer tool.

6)Accept the default Output Feature Class name of Parks_Buffer.

7)In the Distance text box, type .25. In the Linear Unit drop-down menu, click Miles.

8)Click the Dissolve Type drop-down arrow, and then click Dissolve all output features into a single feature.

This setting dissolves boundaries between overlapping park buffers. The dissolution of boundaries isn’t strictly required, but it makes it a lot easier to interpret the result, which would otherwise be a dense tangle of lines. A side effect of dissolving the buffers is that the output will consist of one multipart feature, essentially devoid of attributes. That’s okay, because all you need from this layer is its geometry.

9)Check your settings against the figure, and click Run.

When the tool is finished running, a Parks_Buffer layer is added to the map, and a feature class is created in AnalysisOutputs.

10)Close the Geoprocessing pane.

11)Turn off the Parks layer.

12)Change the transparency of the Parks_Buffer layer to 50 percent.

How? In the Contents pane, select the Parks_Buffer layer, and change the transparency using the slider on the Appearance ribbon, in the Effects group.

The portion of LARiverBuffer not covered by the Parks_Buffer layer represents the areas that are still under consideration. The layers are purple and tan in the figure, but your colors may differ.

Now you can think about what analysis you have accomplished so far. One layer, LARiverBuffer, contains all the area within a half mile of the river. You want to exclude the Parks_Buffer areas from that area, but those areas are stored in a different layer. Essentially, you want to subtract one layer from another.

Erase park-accessible areas

The tool for doing this subtraction is named Erase. This tool is a member of the family of spatial overlay operations that help make GIS such a powerful technology. Unlike the Buffer tool, the Erase tool is not available in the Tools gallery, so you’ll find it using a search.

1)On the Analysis tab, click the Tools button.

2)In the Search box, type Erase.

3)Click Erase (Analysis Tools) to open the tool.

4)For the Input Features parameter, click LARiverBuffer.

5)For the Erase Features parameter, click Parks_Buffer.

6)Name the Output Feature Class parameter ProximityZone.

7)Check your settings against the figure, and click Run.

When the process is completed, the ProximityZone layer is added to the map.

8)Close the Geoprocessing pane.

9)In the Contents pane, turn off all layers except ProximityZone and the basemap.

10)Symbolize ProximityZone in a medium blue (for example, Cretan Blue). This color will help the layer stand out against the basemap.

You’re left with a reduced portion of the original buffer zone. Your search for a park is now confined to this shape.

11)Change the transparency of the ProximityZone layer to 50 percent. (Use the slider on the Appearance tab with the layer selected in the Contents pane.)

12)Zoom in on some of the “holes” in the ProximityZone layer.

You can see in the figure the effect that the existing parks (visible on the basemap) had on the creation of the ProximityZone layer.

13)Pan along the river, and look at a few more examples.

If you see some anomalies—such as a hole with no park—consider that the topographic basemap may not correspond exactly to the Parks layer. You can turn the Parks layer on and off in the Contents pane for reference.

14)When you’re finished, zoom to the ProximityZone layer.

15)Save your project.

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

In the next exercise, you’ll evaluate the neighborhood demographics within the area of interest.

Exercise 6b: Apply demographic constraints

You’ve isolated areas that meet two of your requirements: distance to the river and distance from parks. Still ahead are the tasks of factoring in neighborhood demographics (population density, income, and age) and evaluating parcels by size and total population served. You’ll deal with the demographics in this exercise, and with the parcels and population served in exercise 6c.

Within the proximity zone, you want to find areas that have the right demographic criteria. The problem is that your blue blob, if you can call it that, doesn’t include those attributes; they are found only in the BlockGroups feature class. What you want to do then is make a feature class that has the spatial area of ProximityZone and the attribute values of BlockGroups.

This combination again calls for an overlay. In exercise 6a, you subtracted one layer from another, but overlay operations are typically more additive than subtractive. Usually, you want to find common ground between layers that have different attributes. The important attribute in ProximityZone is its distance to the river. The important attributes in BlockGroups are population density, age, and median income. By overlaying these two layers, you’ll get all these important attributes in a single layer.

The overlay will cause splitting of any block group features that lie partly inside and partly outside the proximity zone. What happens to the attribute values of these features? If a block group with 2,000 people is halfway in the proximity zone, should the output feature defined by the overlapping half get all 2,000 of those people, or just 1,000? By default, attribute values are copied rather than apportioned in overlay operations, and copying can lead to anomalies in which large attribute values are assigned to small slivers of geometry. This problem has been dealt with ahead of time by using block group attribute values that are statistically homogenized and assumed to be uniform across the feature. It’s also possible, however, to redistribute attribute values by area during geoprocessing.

Overlay block groups on the proximity zone

You want to keep all the geometry of the ProximityZone layer and only as much BlockGroups geometry as is spatially coincident with the proximity zone. This type of overlay is called Identity: it keeps all the geometry of layer A, the source layer, and only the coincident geometry of layer B, the target layer.

1)If necessary, start ArcGIS Pro, and open the LARiverParkSite project.

2)Make a copy of the Lesson6a map, and rename it Lesson6b. Open the new Lesson6b map, and close the Lesson6a map by closing its tab above the map.

3)In the Catalog pane, add BlockGroups to the map from the project geodatabase.

4)Move the BlockGroups layer under the ProximityZone layer so that you can visually see the overlap of the two layers.

5)On the Analysis tab, click Tools, and search for Identity.

6)In the Tools list, click “Identity (Analysis Tools)” to open the tool.

7)For Input Features, click ProximityZone.

8)For Identity Features, click BlockGroups.

9)Change the Output Feature Class name to ProximityZoneDemographics.

10)Check your settings against the figure, and click Run.

When the tool is finished running, the ProximityZoneDemographics layer is added to the map. A corresponding feature class is created in the AnalysisOutputs geodatabase.

11)Close the Geoprocessing pane.

12)Zoom to the ProximityZoneDemographics layer, and turn off all other layers in the Contents pane except for the basemap.

What you see in the figure is a fractured version of the original proximity zone. ProximityZoneDemographics has the geometry of BlockGroups clipped to the boundaries of ProximityZone. As you’ll see, it has the attributes of both input layers.

Examine attributes

Looking at the attributes of both your input and output layers will help you understand what you’ve accomplished by overlaying them using the Identity tool.

1)Open the ProximityZone layer attribute table. The ProximityZone layer consists of a single record. Its NAME and BUFF_DIST (buffer distance) attributes are inherited from the LARiverBuffer layer.

2)Open the attribute table of the BlockGroups layer. Scroll across its attributes.

This layer has 6,417 records. It has the attributes you specified and created in lesson 4.

3)Open the attribute table of the ProximityZoneDemographics layer.

4)Drag the ProximityZoneDemographics table by its title bar, position it over the lower docking target, and release.

5)Resize the window so you can see both BlockGroups on the top and ProximityZoneDemographics on the bottom.

The ProximityZoneDemographics table has 206 polygons. It has all the attributes of both input layers (ProximityZone and BlockGroups), plus two ID fields that trace the features back to the input features with which they are spatially coincident.

6)Zoom in on the western end of the ProximityZoneDemographics layer.

A map scale of around 1:24,000 is good.

7)Turn on the BlockGroups layer.

8)On the Map tab, in the Selection group, click the Select tool .

9)Use the mouse to draw a small box inside the feature indicated in the figure to select all features intersecting the drawn box. (A single click selects only features in the topmost layer.)

You can see in the map and in the tables that two features are selected: one in ProximityZoneDemographics and one in the BlockGroups layer underneath.

10)Click the Show Selected Records button at the bottom of each of the tables.

11)Scroll horizontally across the two tables so that you can see all the demographic fields.

All the block group values have been copied to the spatially corresponding features in the ProximityZoneDemographics layer. This distribution of values is what you want—these cut-up block groups (some are cut up, some are intact) are the “neighborhoods” that you’ll query for suitable demographics.

Note the importance of analyzing attributes, such as density and percentage, rather than raw numbers. The Pop Under 18 value of the selected block group is 194. You would reasonably expect this value to be smaller in the ProximityZoneDemographics feature (because it doesn’t include the whole block group), but because the attribute values are copied, it’s the same. It’s fair to assume, however, that the % Under 18 value would remain the same even when the block group is cut into smaller pieces. This assumption is not the case, however, for the Total Pop attribute, and you might consider apportioning this value on the basis of the percentage of the area that was cut. Your analysis will not rely on apportioning, because you’ll be using the more detailed census block units to derive the park access population (as opposed to the block groups that you just used for the other demographic variables). For more information on apportioning, see the sidebar “Apportioning attribute values.”

12)On the Map tab, in the Selection group, click Clear.

13)Close the open attribute tables.

14)Turn off the BlockGroups layer.

15)Zoom to the ProximityZoneDemographics layer.

Apportioning attribute values

Splitting feature geometry has implications for attribute values. In example 1, a census block group is split by a buffer in an overlay operation, resulting in two output features. By default, ArcGIS Pro copies the attribute values of the input feature to both output features. That’s okay for POPDENSITY, which is already a ratio: the value applies to the parts as well as the whole. It’s not okay for TOTALPOP, a count value, because it doubles the population, as shown.

In example 2, you want to count the population inside the buffer, which covers parts of three block groups. A reasonable approach is to analyze the percentage of each block group’s area that falls inside the buffer and assign the same percentage to its population value. In this case, 1% of block group 1’s area is inside the buffer, so its contribution to the buffer population should be 1% of 1,755. Likewise for block group 2 (21% of 4,445) and block group 3 (38% of 6,722).

The examples are variations of the same problem, which can be solved with a tool named Make Feature Layer. This tool makes a layer from a feature class (essentially the same thing that happens automatically when you add data to ArcGIS Pro). It has a very useful parameter, however: a “ratio policy” check box for each attribute in the input feature class. When you geoprocess a layer made with this tool, the ratio policy is applied to any split features, dividing their attribute values according to area.

Select areas by demographic attributes

The ProximityZoneDemographics layer now contains the demographic attributes of interest: population density, the percentage of children, and median household income. You want to consider areas only if they meet your thresholds for all three values:

•population density greater than or equal to 8,500 people per square mile,

•percentage of children greater than or equal to 22 percent, and

•median household income less than or equal to $50,000.

You’ll use an attribute query to select features that satisfy these conditions.

1)On the Map tab, in the Selection group, click Select By Attributes.

2)At the top of the Select Layer By Attribute tool, click the Layer drop-down arrow, and click ProximityZoneDemographics. The drop-down menu defaults to the selected layer in the Contents pane, so the correct layer may already be selected.

3)Make sure the Selection type is set to New selection.

4)Click Add Clause, and create a clause for Pop Density is Greater Than or Equal to 8500. Then click Add.

5)Click the Add Clause button again, and create another clause for And % Under 18 is Greater Than or Equal to 22. Then click Add.

6)Click the Add Clause button again, and create another clause for And Median HH Income is Less Than or Equal to 50000. Then click Add.

If your clause is getting too long to read unhindered, hover over it, and you’ll see the entire thing.

Combining the queries in a single statement is more efficient than making a series of selections and subselections. The And operator ensures that features will be selected only by passing all the tests.

7)Check your tool against the figure, and click Run.

If you need to edit an expression, double-click on the clause.

8)Close the Geoprocessing pane when the selection runs successfully and the resulting parcels are highlighted on the map.

The figure shows the features that satisfy the query. There aren’t that many, and most of them lie toward the ends of the proximity zone.

9)Open the ProximityZoneDemographics attribute table, and click the Show Selected Records button.

10)Confirm that there are 27 selected records, and review the values of the demographic fields to confirm that they meet the park criteria.

11)Close the table, but keep the records selected.

Create a new feature class from the selected features

You’ll save these selected features to a new feature class. Previously, in lesson 4, you used the Copy Features tool from the Tools list for this purpose. You’ll use that same tool but a more convenient method to get to it.

1)Confirm that the parcels are still selected, and then right-click ProximityZoneDemographics in the Contents pane, and click Data > Export Features. This command opens the Copy Features tool with the Input Features parameter already filled out.

2)Change the Output Feature Class parameter to GoodZones.

3)Check your settings against the figure, and click Run.

When the tool is finished running, the new GoodZones layer is added to the map, and a new feature class is added to the AnalysisOutputs geodatabase.

4)Close the Geoprocessing pane.

5)Turn off all layers except the new GoodZones layer and the basemap.

6)On the Map tab, in the Selection group, click the Clear button.

7)Zoom to the full extent of the GoodZones layer.

8)Save your project.

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

Exercise 6c: Select suitable parcels

You’ve narrowed your focus to a layer named GoodZones. Having covered the demographic criteria, you’ll turn your attention to the parcels. Thanks to your work in lesson 4, you have a dataset consisting of all, and only, vacant parcels. From this dataset, you’ll select the parcels that lie within good zones. From that selection, you’ll further select parcels that are one-quarter acre or larger: that selection will comprise your set of candidate park sites. Then you’ll do further analysis to compare the sites by total population served and their distance to the river.

Select vacant parcels within good zones

You’ll select the parcels using a spatial query.

1)If necessary, start ArcGIS Pro, and open the LARiverParkSite project.

2)Make a copy of the Lesson6b map, and rename it Lesson6c. Open the new Lesson6c map, and close the Lesson6b map by closing its tab above the map.

3)Add the VacantParcels feature class to the map from the project geodatabase.

4)On the Map tab, in the Selection group, click Select By Location.

5)Set Input Feature Layer to VacantParcels.

6)Confirm the Relationship drop-down menu is set to Intersect.

7)Set the Selecting Features parameter to GoodZones.

8)Leave the Search Distance box blank.

9)Confirm that the Selection type is set to New selection.

Parcels will be selected if they lie partially or entirely within good zones.

Some parcels may cross good zone boundaries because of how the zones are clipped. Should you include parcels that straddle good zones? There’s no right or wrong answer. By choosing the “intersect” spatial relationship method, you’re going with a generous interpretation—you’ll select a parcel if any part of it lies within a good zone, because you want to maximize the candidate pool. If you wanted to be restrictive (the parcel must be entirely within a good zone), you’d choose a method such as Within.

10)Compare your tool to the figure, and click Run.

11)Close the Geoprocessing pane after the Selection tool runs successfully.

12)Open the VacantParcels attribute table. You should have just 55 of 14,269 parcels selected.

Select the subset of parcels that are one-quarter acre or larger

Of the 55 selected parcels, you want to keep only those that are at least one-quarter acre. You’ll select this subset using an attribute query on the Acres field using the features that are already selected (on the basis of the spatial query created in the last section).

1)On the Map tab, click the Select By Attributes button above the table.

2)In the Select Layer By Attribute tool, use the drop-down menu to change the Selection type to Select subset from the current selection. This setting restricts the scope of the query to records that are already selected.

3)Add a clause for the following:

Acres >= 0.25.

4)Check your tool against the figure, and click Run.

5)Close the Geoprocessing pane.

6)Look at the attribute table. Click the Show selected records button. You’re down to six selected records: just six parcels meet all your conditions.

At this scale, it’s hard to see the selected parcels on the map. You’ll take a closer look in step 7.

7)In the Contents pane, right-click VacantParcels and click Selection > Zoom To Selection.

8)Close the attribute table.

Export selected features

These six parcels are the critical ones you want: they represent the culmination of the process, and you want to save them as a feature class. Again, you’ll use the convenient Export Data method.

1)In the Contents pane, right-click VacantParcels and click Data > Export Features.

2)In the Copy Features tool, change the Output Feature Class name to SixSites.

3)Check your settings against the figure, and click Run.

When the tool is finished running, the new SixSites layer is added to the map.

4)Close the Geoprocessing pane, and close any tables.

Inspect the candidates

Now you’ll zoom in and take a close look at the six sites against basemap imagery.

1)On the Map tab, change to the Imagery basemap.

2)In the Contents pane, remove VacantParcels. Turn off all layers except SixSites and Imagery. The six site polygons may still be difficult to see with the current scale and symbology.

3)In the Contents pane, open the Symbology pane for the SixSites layer by clicking the color patch below the layer name.

4)In the Symbology pane, switch to the Properties view if necessary.

5)Set the fill color to No color and the outline color to Medium Apple.

6)Make the outline width 2 pt.

7)Click Apply.

8)Close the Symbology pane.

9)Open the attribute table of the SixSites layer, and then resize the table so that the map is large and all six rows are visible.

10)In the table, select the third record. Right-click the gray cell next to the third record, and click Zoom To.

The map zooms to one of your candidate parcels. Although a portion of the parcel is being used as a parking lot, the parcel is adjacent to a quiet street to the west and could be a good park.

11)In the drop-down scale menu below the map, change the map scale to 1:24,000.

12)If necessary, pan south so you can see the river.

13)In the Contents pane, turn on the ProximityZone layer.

14)Optionally, turn on the Parks layer and, if necessary, symbolize it in a shade of green.

The view confirms that the site is near the outer edge of the proximity zone. Even more interesting is that the site has two other parks in the immediate vicinity that are both closer to the river. That’s why human interpretation of analysis results is so important. You considered proximity to existing parks, but it didn’t occur to you that a park might be more than a quarter mile from a candidate site and yet closer to the river. It doesn’t invalidate your candidates, but it’s something to think about.

15)Turn off the ProximityZone and Parks layers.

16)In the table, select the first record. Right-click the gray square next to the first record, and click Zoom To.

17)Zoom to a map scale of 1:5,000.

These two sites are close to the river, which is ideal. Notice a couple of other things. First, the river bottom is natural here, which could make these sites more scenic than others. Your analysis doesn’t take aesthetic considerations into account, but the views, sounds, and smells of a park affect the people who use it.

But, these sites are also close to a freeway. This location has an aesthetic impact and probably a health impact as well should kids be exposed to all that car and truck exhaust.

Investigate the remaining sites

1)Zoom to the remaining sites and investigate them:

•Zoom and pan to see the site in relation to the river.

•Turn other map layers on and off.

•Make observations that add context to the analysis.

2)When you’re finished, click the Clear button to remove any selections in the table.

3)Turn off all layers except SixSites and the basemap.

Leave the SixSites attribute table open. Confirm that there are no selected records in the table. If necessary, click the Clear button above the table.

At this point, two items remain from your list of analysis criteria: you want the new park to serve the maximum number of people, and you want it to be as close to the river as possible.

Calculate distance to the river

You know that all the sites are within a half mile of the river, but knowing the exact distance from each site would be a useful metric to help decision-makers prioritize the sites. You could use the Measure tool to do that, of course, but you’d face the same problem of repetitively writing the results to a table. Also, you might not be sure of making the shortest possible measurement. The Near tool will make these calculations for you automatically.

1)On the Analysis tab, in the Tools gallery, open the Near tool .

2)In the Near tool, set Input Features to SixSites.

3)Set the Near Features parameter to LARiver.

Note that you can add multiple layers to the list of Near features. Typically, the tool is used to discover which feature, in one or more layers, is closest to a feature of interest. In this case, you want only one distance measurement.

4)Using the drop-down menu, change Method from Planar to Geodesic.

This method gives you the most accurate distance, regardless of the map projection. For more complicated datasets with more features, processing using this method may take longer because of the more complex geodesic distance calculations. But again, in this case, you want only one distance.

Note that, unlike most of the other tools you’ve used in this lesson, the Near tool doesn’t have an Output Feature Class parameter. The Near tool does not create a new feature class. Instead, it will add a couple of fields to the SixSites table.

5)Check your settings against the figure, and click Run.

6)If necessary, open the SixSites layer attribute table.

The NEAR_DIST field stores the distance, in feet, of each site to the river. This distance unit is dependent on the coordinate system. Because your coordinate system uses feet as the linear unit, your distance units are in feet.

7)Sort the attribute table by NEAR_DIST in ascending order to see how the sites rank in terms of proximity to the river.

8)Close the table.

Add block centroids to the map, and visualize the analysis

You must find the population within a quarter mile of each potential park site. The basic approach of counting the block centroids around each park and summing their populations was introduced in lesson 2 and sketched in the analysis plan at the beginning of this lesson.

You can do that using the Spatial Join tool. You might think of a spatial join as a spatial query plus a table join: attributes from one table are joined to another on the basis of a spatial relationship between layers rather than possession of a common attribute.

Now you can get a visual sense of what you’ll accomplish with the spatial join by examining the population data within a quarter mile of a park site.

1)Zoom in on one of the sites.

2)Change the scale to 1:10,000 using the drop-down scale menu below the map.

3)Add BlockCentroids to the map from the project geodatabase. With the imagery basemap, you can clearly see the blocks that the point centroids represent.

4)Open the BlockCentroids attribute table, and find the POP2010 attribute. This is the attribute that you’ll use to find the population within a quarter mile.

5)Turn off all layers (including the imagery basemap) except BlockCentroids and SixSites.

6)On the Map tab, in the Inquiry group, click the Measure tool. Confirm that Measure Distance is selected by clicking the drop-down arrow under the button.

7)Change the distance units to miles by clicking Options from the Measure panel in the upper-left corner of the map.

8)Click the center of the site and move your pointer to measure out a quarter mile. Double-click to stop measuring.

In the next section, you’ll use the POP2010 field values and sum the values within this quarter-mile radius as a simple means of determining how many people are within easy walking distance of the park.

9)Change the active Measure tool back to Explore by clicking the Explore button on the Map tab, in the Navigate group.

10)Turn off the BlockCentroids layer.

11)Close the BlockCentroids table.

Determine population around the park site

In lesson 2, you decided to sum the population in terms of a quarter-mile distance from the park site. This amount is an oversimplification of accessibility, but it has the virtue of being easy to analyze. You’ll use block centroids for this measurement, because they are the most detailed geographic unit maintained by the census. Because of their small size, the block centroid population values are not as easily estimated and are therefore only updated with the decennial census every 10 years. This method will be acceptable for this analysis but is definitely something to consider for fast-growing areas.

1)Right-click the SixSites layer, and click Zoom To Layer. You probably can’t see the sites anymore, but that’s okay.

2)Clear any selections by clicking the Clear button on the Map tab, if necessary.

3)On the Analysis tab, in the Tools gallery, click Spatial Join .

4)In the Geoprocessing pane, set Target Features to SixSites.

5)Set Join Features to BlockCentroids.

6)Change the Output Feature Class name to SixSitesAccessPop.

7)Leave the default Join Operation parameter as Join one to one.

The Field Map of Join Features area defines which fields will be in the output. The only fields you need in the output are the ACRES, NEAR_DIST, and POP2010 fields, so you’ll remove all the others. You can also do some cleanup of field names at this point, so you’ll change the name of the NEAR_DIST attribute to something more descriptive.

8)In the Output Fields list in the Field Map of the Join Features area, click Shape_Length, and then click the Remove button that appears on the right of the field name.

9)Repeat step 8 for all other fields in the list except ACRES, NEAR_DIST, and POP2010.

If you remove too many fields by mistake, you can reset the Output Fields list by clicking the Reset button in the upper-right corner of the list.

10)In the Output Fields column, click POP2010, and then click Sum in the Merge Rule drop-down menu, in the Source area on the right.

The Merge Rule setting is crucial. It tells ArcGIS Pro that for each target feature (park site), you want to sum the population values of the join features (block centroids). Note that other merge rules are available to select.

11)Click the Properties tab (just above the Merge Rule drop-down arrow), and change the Field Name to AccessPopulation. Confirm that the alias changed to AccessPopulation.

12)In the Output Fields column, click NEAR_DIST, and change the Field Name to RiverDistance. Confirm that the alias changed to RiverDistance.

13)In the Match Option drop-down box, click Within a distance.

14)Enter 0.25 for Search Radius, and confirm that the distance is set to miles.

15)Compare your tool to the figure, and click Run.

When the tool is finished running, a SixSitesAccessPop layer is added to the map.

16)Close the Geoprocessing pane.

17)Open the attribute table of the SixSitesAccessPop layer.

The AccessPopulation field has the total population within a quarter mile of each site. You’ve also taken the opportunity to clean up the table, so it’s easier to interpret the results.

18)Close the table.

Add the demographic attributes to SixSitesAccessPop

It would be nice to store all the relevant site-selection attributes in a single table with the candidate sites. These attributes include the following:

•Acreage

•Distance to the river

•Total population within a quarter mile

•Demographic variables

Doing so will make it easy to evaluate and compare the sites at a glance—for you and for anyone you share the data with. You already have the acreage, distance, and population attributes in the SixSitesAccessPop table. In this section, you’ll use the Identity tool to get the demographic attributes into the table. These attributes are found in the BlockGroups layer.

1)Open the Identity tool (browse or search from the Geoprocessing pane).

2)For Input Features, click SixSitesAccessPop.

3)For Identity Features, click BlockGroups.

4)Rename Output Feature Class as RecommendedSites.

5)Check your settings against the figure, and click Run.

When the tool is finished running, the RecommendedSites layer is added to the map.

6)Close the Geoprocessing pane.

7)Open the RecommendedSites attribute table, and scroll across it.

It has all the attributes of interest, plus several attributes that you don’t need along with some ID attributes (back links to features in other tables) that you don’t need for this exercise. In the next exercise, you’ll format this table to make it presentable.

8)Close the table.

9)In the Catalog pane, confirm that your AnalysisOutputs geodatabase looks like the figure.

10)Save your project.

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

Exercise 6d: Clean up the map and geodatabase

Analysis projects tend to leave clutter behind. It’s normal to find your working map and geodatabase filled with a mixture of data you want to preserve and data you can now discard. It’s tempting to leave the mess behind, but sooner or later someone will return to the scene and wish that you’d taken the time to put things in order. In this exercise, you’ll simplify your map, which contains several unnecessary layers. You’ll also format the SixSites attribute table and save it as a layer file. Finally, you’ll remove intermediate data from the geodatabase and add metadata to the datasets you want to keep.

Clean up the map

This map isn’t the one you’ll use for your final layout, but you still want it to be intelligible. With that in mind, you’ll keep just a few important layers in it.

1)If necessary, start ArcGIS Pro, and open the LARiverParkSite project.

2)Make a copy of the Lesson6c map, and rename it Lesson6d. Open the new Lesson6d map, and close the Lesson6c map by closing its tab above the map.

3)In the Contents pane, remove all layers except the four listed as follows. Put the layers in the order shown, from top to bottom:

•LARiver

•RecommendedSites

•LARiverBuffer

•World Imagery (the basemap)

4)Turn all four layers on, and zoom to the LARiverBuffer layer.

5)If necessary, symbolize RecommendedSites with a fill color of No color, an outline color of Medium Apple, and an outline width of 2 pt.

6)Symbolize the LARiver layer with the color Big Sky Blue and a width of 3 pt.

7)Symbolize LARiverBuffer with a fill color of No color, an outline color of Medium Lilac, and an outline width of 3.

Your map should look like the figure.

Add layer descriptions

When you add a layer to a map, a description—taken from the item description of the source data—is added to the layer properties. This description explains what the layer represents. Layer descriptions are helpful to anyone using the map and are required if you share layers as layer packages. The LARiver layer already has a description, as does the World Imagery sublayer of the basemap; the other two layers, created during geoprocessing, do not.

1)Open the layer properties of RecommendedSites (right-click the layer), and click the Metadata tab on the upper left under General.

2)In the drop-down menu, switch from Show metadata from source (read-only), which is the default, to Layer has its own metadata.

3)In the Title box, type: Recommended Sites.

4)Add the tags Los Angeles and Recommended Sites.

5)In the Description box, type These are the six sites recommended for a new park near the Los Angeles River. Click OK.

6)Open the layer properties of LARiverBuffer, and switch from Show metadata from source (read-only) to Layer has its own metadata.

7)Change Title to LA River Buffer (with spaces).

8)Add Los Angeles and LA River Buffer as tags.

9)In the Description box, type This is a half-mile buffer around the Los Angeles River. Click OK.

Delete unnecessary fields

In exercise 6c, you got all the important attributes into the RecommendedSites table. You also have several attributes you don’t need, along with several FID attributes that you don’t need to maintain.

1)Open the RecommendedSites attribute table.

2)In the table, right-click the FID_SixSitesAccessPop field heading, and click Delete.

3)Click Yes on the prompt to confirm the field deletion.

The field is deleted from the table.

Be careful when you delete a field, because you can’t undo it.

4)In the same way, delete all the remaining unnecessary fields from the table:

•Join_Count

•TARGET_FID

•FID_BlockGroups

•Total Pop (this was the block group’s population, not the more detailed block sum)

•Pop Over 18

•Sq Miles

•Pop Under 18 (you need only the percentage)

The table should look like the figure. If you have more fields than those listed, go ahead and delete them.

Add a SiteID field

You’re going to add an identifier field that you can manage. You’ll have a reason to use this field in lesson 8.

1)Click the Add Field button at the top of the table.

2)Name the field SiteID, give it an alias of SiteID, and change Data Type to Short.

3)Check your settings against the figure, and save your field changes by clicking Save on the Fields ribbon, in the Changes group.

4)Close the fields view to return to the attribute table.

5)Scroll horizontally to the end of the table. Right-click the SiteID field heading, and click Calculate Field.

6)In the Calculate Field tool, double-click OBJECTID in the list of fields to add it to the Expression box.

This expression will assign the same values of 1 to 6 to the new field. The difference is that SiteID is a field you can manage, whereas OBJECTID is not.

7)Check your settings against the figure, and click Run.

The field is populated with the values from the OBJECTID field.

8)Close the Geoprocessing pane.

Format fields

Most of the fields in the table should be given an alias or formatted in other ways to make them more readable. The steps are provided in detail for one field, and you can do the rest with general instructions.

1)Open the fields view for RecommendedSites by clicking the Menu button to the right of the table and then clicking Fields View.

2)Confirm that the Current Layer drop-down box above the table is displaying RecommendedSites (Lesson6d).

3)In the list of fields, in the Visible column, click to clear Shape_Length and Shape_Area to hide these fields. In the Fields ribbon, click Save.

4)At the bottom of the list, press and hold the gray square to the left of the SiteID row, and drag it to the top of the list just below the Shape field.

Moving this field affects only this layer. In the feature class, fields will continue to be stored in the order in which they’re added. If you drag this feature class to this map or any other, the fields will be displayed in their original order.

5)Find the row for the PCTUNDER18 field, and scroll to the right to find the Number format column.

6)Double-click in the cell that currently contains the text Numeric, and click the ellipse button to the right of the cell.

The Number Format dialog box appears to more specifically define how the attribute values should be displayed in the map.

7)In the Category drop-down menu, click Percentage.

8)Confirm that Number already represents a percentage is selected.

9)Change the number of decimal places to 0.

10)Compare to the figure, and click OK.

11)Format the remaining fields per table 6-1. Some of these fields may already be correct in your table, but depending on the field, you may have to

•change or add an alias,

•change the number format,

•change the number of decimal places, or

•select the check box to show thousands separators.

Table 6-1. Field changes
Field	Alias	Number Format
ACRES	Acres	Numeric (one decimal place)
RiverDistance	Feet to LA River	Numeric (zero decimal places; show thousands separators)
AccessPopulation	1/4 Mile Population	Numeric (show thousands separators)
MEDHINC	Median Household Income	Currency
POPDENSITY	People per Square Mile	Numeric (zero decimal places; show thousands separators)

12)On the Fields ribbon, click Save.

13)Close the fields view to return to the attribute table. The attribute table should now look like the figure.

14)Close the attribute table.

Save the layer as a layer file

All these format settings are layer properties. If you want to make them available in other maps, you must save RecommendedSites as a layer file, as you did with the city boundary of Los Angeles in lesson 1.

1)In the Contents pane, right-click RecommendedSites, and under Sharing, click Save As Layer File.

2)In the Save Layer(s) As LYRX File dialog box, navigate, if necessary, to the MapAndMore folder.

3)Name the file RecommendedSites and click Save.

Exercise 6e: Evaluate your results

The recommended sites meet your criteria and look good on basemap imagery, but ultimately, there’s no substitute for on-site inspection. How do these candidate locations really look in the context of their surroundings? If you had the opportunity, it would be worthwhile to visit the sites and record your impressions. In this exercise, some photographs of the sites are included to help give you a sense of that experience.

Compare your guesses from lesson 1

In lesson 1, you made some guesses about likely areas for parks on the basis of a preliminary look at the data. Now you can see how those guesses turned out.

1)If necessary, start ArcGIS Pro, and open the LARiverParkSite project.

2)Make a copy of the Lesson6d map, and rename it Lesson6e. Open the new Lesson6e map, and close the Lesson6d map by closing its tab above the map.

3)Open the Lesson1b map by double-clicking it in the Catalog pane, under Maps.

4)Drag the Lesson1b tab to the right-hand docking target that appears in the middle of the map.

The Lesson1b map is placed to the right of the Lesson6e map so that you can compare them side by side.

5)On the View tab, click the drop-down arrow below the Link Views button, and click Center And Scale. A link icon appears on both map tabs to show that the map views are linked.

The two maps are linked, so that as you navigate one, the other map’s center and scale will be synchronized. Linking views links the views of all the open maps, so this same method can be used to compare as many maps (and 3D scenes, which display data on a globe) as you want.

6)Click the Lesson1b tab to make it the active map. The active map determines which layers are displayed in the Contents pane.

7)Zoom to an area where you placed some potential park locations. Both maps will zoom to that area for closer inspection.

Copy layers between maps

Another way to make this comparison is to add the layer that has your guesses (the Map Notes you created in the Lesson1b map) to the Lesson6e map.

1)If necessary, make Lesson1b the active map by clicking its tab above the map.

2)Right-click Bright Map Notes, and click Copy.

3)Activate the Lesson6e map by clicking its tab above the map.

4)In the Contents pane, right-click Lesson6e, and click Paste. The Bright Map Notes group layer (containing point, line, and polygon sublayers) is added to the map.

5)On the View tab, in the Link group, click the Link Views button again to turn it off. The link icon will disappear on both map tabs.

6)Close the Lesson1b map.

7)Zoom and pan around the Lesson6e map to see how your guesses (purple stars) compare with the analysis results.

Examine park candidates

You’ll take a short virtual tour of the six candidate sites and create bookmarks of each site.

1)Open the RecommendedSites attribute table, and zoom to Site 1. Create a new bookmark, Site 1, on the Map tab, under Bookmarks.

2)Zoom to Site 2, and create a Site 2 bookmark.

3)Zoom out until you see both Sites 1 and 2. As noted earlier, both sites are close to the river, which is ideal. Both sites have steep slopes but do have good access.

4)Zoom to Site 3, and create a Site 3 bookmark.

Site 3 is on a quiet street and is level, but part of the site is being used as a parking lot. This would require some additional work to create a park at this location.

5)Zoom to Site 4, and create a Site 4 bookmark.

Site 4 has good access and has a bike path across the street to the east. It appears it is being used for storage.

6)Zoom to Site 5, and create a Site 5 bookmark.

As with Sites 1 and 2, Sites 4 and 5 are also close together.

7)Zoom out until you see both sites. Site 5 is on a corner lot, which would create a more open feel to a potential park. It also has good demographics, serving the lowest median household income of all the sites and the highest density of people.

8)Zoom to Site 6, and create a Site 6 bookmark.

Site 6 is clear and ready for development. And it has the advantage of having a new Metrolink station right across the street.

9)Close the attribute table, and zoom to the LARiverBuffer layer.

10)Close any open maps.

11)Save your project, and close ArcGIS Pro.

This section wraps up your analytical work. You started with a set of (fictional) guidelines from the city council. Those guidelines were vague in many cases, so you ended up supplying your own working definitions. Obviously, these guidelines could be questioned. The guidelines also excluded many factors that might be important: whether the land is actually available, its environmental condition, barriers to access such as freeways, and several more. Finally, what the council did not give you—and you have no way to evaluate on your own—is the relative importance of the criteria. You can’t rank your candidates because you don’t know, for example, if low household income outweighs a high percentage of children in making a site desirable, and if so, by how much. You have to present the options neutrally, because you don’t know what the council’s priorities are. (The council members may not know either until they begin to study the results.)

Suppose somebody asked you how the results would change if you set the income requirement lower or the percentage of children requirement higher. How could you answer a question like that without rerunning the entire analysis from scratch? Read on to lesson 7 for the solution.