Next, you'll open the project in ArcGIS Pro.

The project opens.

On the map, two layers are currently turned on. The Study Area layer shows the area of interest for this workflow, represented as a rectangular polygon (in black). The World Topographic basemap provides general geographic context. You can see that the study area is located in the greater Los Angeles area, in the United States. It includes the several communities, such as San Fernando Valley, Burbank, and Santa Clara, as well as several mountain ranges spread out throughout the area. Next, you will explore the mountain lion habitat areas.

The study area contains four core mountain lion habitats, represented in purple. These are the habitats that you want to connect by proposing locations for wildlife corridors. Note that mountain lions can be present in a larger zone around each core habitat, but the purple polygons show the areas of highest mountain lion concentration.

The three habitats identified as Los Padres, Santa Monica, and San Gabriel represent true core mountain lion habitat. They are set in spacious, mountainous natural areas and contain significant mountain lion populations. The Santa Susana location, which was established by the National Park Service and is located just outside of Pico Canyon Park, doesn't represent a significant core area, but it serves as an important geographic gateway for linking the other three geographically disconnected habitat patches to each other.

Instead of managing the mountain lions present in the four habitats as four separate populations, you want to manage them as a single metapopulation—a population of populations. Ensuring that this overall population of mountain lions can move around from one core habitat patch to another will help it remain healthy and genetically diverse. And if the mountain lions go locally extinct in one of the patches, the lions from other patches can recolonize it. This is why it is essential to establish wildlife corridors to connect these existing habitat patches.

You'll turn the labels off as you won't need them any longer.

On the map, the labels disappear.

The Elevation layer appears.

This is a raster layer where each cell represents elevation above sea level. Cells with lower elevation appear in dark tones and cells with higher elevation appear in lighter tones. You can see mountainous elevated areas in white and light gray and flatter low-lying valleys in black and dark gray.

You'll zoom in to better visualize how the raster layer is made of cells.

You'll display the elevation value for some of these cells.

This option ensures that you see information about the layer that is currently selected.

In the following example image, the cell represents an elevation of about 890.8 meters.

As such, the layer represents elevation and not ruggedness. Later in the workflow, you need to preprocess it to derive a new layer that can effectively represent the ruggedness criterion. You'll then assign the lowest cost to the highest ruggedness values.

Next, you'll review the dataset that you'll use for the second criterion, dense vegetation land cover.

The Land Cover layer appears.

Land Cover is another raster layer, although it doesn't have a simple black-to-white color gradient like the Elevation layer. To understand what this layer shows, you'll investigate its symbology.

This layer represents different land cover types, like Evergreen Forest or Cultivated Crops. Each of these land types is symbolized with a color: the red colors are developed urban areas, while the green colors correspond to forests or grasslands. The brown colors are shrubland or herbaceous areas. This layer doesn't need to be preprocessed and can be used directly in the least-cost analysis, where you'll assign the lowest cost to dense vegetation cover types, such as forest and shrub.

Next, you'll review the layer you'll use for the third criterion: protected areas.

The Protected_Status layer appears on the map.

This raster layer indicates the level of protection for all lands contained in the Protected Areas Database of the United States (PAD-US): from highly protected in dark green to not protected in light green. An example of highly protected land would be a national forest. An example of not protected land would be a recreation area. Areas without a color on the map are not tracked in PAD-US, and usually correspond to urban, agriculture, or other forms of privately exploited areas—they will usually be the least favorable to mountain lions. This layer doesn't need to be preprocessed and can be used directly in the least-cost analysis, where you'll assign the lowest cost to the highest levels of protection.

Last, you'll review the dataset related to the fourth criterion: distance from roads.

On the map, the Roads layer appears.

The Roads layer is not a raster layer; it's a vector layer that represents features as lines. (Other types of vector layers can contain points or polygons.) Each line represents a road, ranging from small roads to major highways, although you will not distinguish between these different types for most of the workflow. To use this layer in the least-cost analysis, you will need to preprocess it and derive from it a raster representing the distance from any cell to the nearest roads. You'll then assign the lowest cost to the highest distance-to-road cell values.

The Hillshade raster function appears.

A new Hillshade_Elevation layer appears on the map.

The layer makes it clear to the human eye where mountain ranges and valleys are located throughout the study area. While you won't use this layer in the analysis itself, you will use it as a helpful background to visualize the various layers you'll generate throughout the analysis.

You will rename the Hillshade_Elevation layer.

Since this layer will be used mostly as a background, you will move it below the data layers in the Contents pane.

You'll save your project.

The Options window appears.

These options will ensure that you can repeat your analysis smoothly.

Settings changed through the Options window, as you just did, will be saved across all your projects in ArcGIS Pro. The rest of the settings you'll choose are specific to this project, so you must choose them within the project itself.

The Environments window appears. First, you will ensure that the default workspace (the location where new layers created by geoprocessing tools are saved) is the geodatabase connected to your project.

Both workspaces indicate the location where data is saved, but the scratch workspace is intended for data you don't want to keep. For this exercise, you will keep the same location for both.

The coordinate system automatically changes to the coordinate system of the Elevation layer, NAD 1983 UTM Zone 11N. By choosing a default coordinate system, you ensure that all the output layers you create will use the same coordinate system.

This parameter guarantees that all output layers will automatically be clipped to your study area. By reducing the size of your output, you lower the processing time and ensure that your results cover only the area you are interested in.

Extent updates to the coordinates of the Study Area's corners.

The Elevation layer and other raster layers in this project have a resolution of 30 meters. This means that each cell represents an area of 30 square meters in the real world. The Cell Size parameter ensures that all output raster layers generated during the analysis will have the same cell size.

The Snap Raster parameter causes all output raster layers to match the cell alignment of the chosen layer, meaning that the cells will overlap exactly.

A new, blank model appears. First, you'll add the Elevation layer to it. The layer is listed in the map's Contents pane, so you can drag it from there.

The layer is displayed in the model as a blue oval with the layer's name. Blue ovals are input data variables: they represent the initial data that you will process with a geoprocessing tool.

Next, you'll add the geoprocessing tool. For this workflow, ruggedness will be defined by large and dramatic shifts in elevation. You'll use the Focal Statistics geoprocessing tool to generate a layer that shows the amplitude in elevation variation. For each cell in the Elevation raster, the tool will look at the cells surrounding it (also known as its neighborhood) and compute the range between the lowest and highest cell values. For instance, if the lowest and highest values in a cell's neighborhood are 430 and 450 meters, the range will be 20 meters, and that will be the value assigned to that cell. Higher range values correspond to more pronounced ruggedness. First, you'll locate the Focal Statistics tool.

The Geoprocessing pane appears.

The tool is currently depicted as a gray rectangle, connected to an output that is also gray. Both elements are grayed out because they are currently inactive. You must connect them to an input layer to make them active.

When you release the mouse button, the two model elements are connected and a list of connection options appears.

This means that the Elevation element will play the role of the input raster in relation to the tool. The tool becomes yellow and the output element becomes green, indicating that they are now active and the tool can be run.

Before you run it, you will set additional parameters to fine-tune your results.

You'll change the name of the output and choose the statistics type to compute the neighborhood range.

The name of the output element changes to Ruggedness. Last, you will modify the output element, so that the output layer is added to the map when the tool runs.

You'll name the model and save it.

The model is saved. You will now run the model.

The model runs. A window appears and informs you about the status of the process. After a few moments, the resulting Ruggedness layer is added to the Contents pane.

A gray drop shadow behind the Focal Statistics and Ruggedness elements also indicates that the tool has finished running.

The Ruggedness layer displays on the map.

The Ruggedness layer depicts the ruggedness level in black to white tones. Darker cells have the least ruggedness, while lighter cells have the most.

The Symbology pane appears for the Ruggedness layer.

The map updates to the new symbology.

The green areas are the most rugged, and the red areas the least. To add more context, you'll make the Hillshade layer show through.

The map updates, showing the Ruggedness layer overlayed on the Hillshade layer.

Looking at the ruggedness in relation to the terrain relief, it's clear that the high, rocky mountain peaks tend to be more rugged, and the low-lying valleys tend to be flatter and smoother. More ruggedness signifies a lower cost for the mountain lions, in terms of stress and danger. Notice that the mountain lions core habitat patches are, unsurprisingly, located in highly rugged areas.

To calculate distance from roads, you'll use the Distance Accumulation tool, which will calculate the straight-line distance from each cell to the closest road.

This tool has several outputs, but you are only interested in the distance accumulation raster, so you'll ignore the other output rasters.

The tool and the Output Distance Accumulation Raster element become active, but the other output rasters do not. This is because these output rasters are optional and don't get created by default. You'll open the tool parameters to change the name of the distance accumulation output.

You'll save the model.

Next, you'll run the tool. However, you don't want to run the Focal Statistics tool again. You can run only part of a model by selecting the tool you want to run.

The tool runs. After a few moments, the resulting layer is added to the Contents pane.

You'll change the layer's symbology.

The map updates to the new symbology. You'll now let the Hillshade layer show through.

The map updates.

In the Contents pane, the legend for the layer indicates that the cell values range from 0 to over 10,000 meters (approximately 7 miles). Note that most of the study area is close to a road (in red) and that the farther areas (in light yellow or green) are few and fragmented. This indicates how widespread human development has become in the area. The few areas away from roads tend to be on higher mountainous peaks. You can see that the mountain lions core habitat patches are for the most part located in these rare areas away from roads. Further distance to roads signifies a lower cost for the mountain lions in terms of stress and danger. However, when roads can be found almost anywhere, they can't be entirely avoided.

The tool and its output become active. Next, you must specify the output name and the transformation function.

Some of the parameters were populated automatically, based on the Ruggedness layer input. Reading the Lower Threshold and Upper Threshold parameters, you can see that the ruggedness values can vary from 0 to 223.39.

• For Output raster, change the name of the output layer to Ruggedness_Cost.
• For Transformation function, choose MSLarge.
• For From scale, type 10. For To scale, type 1.

Ruggedness values close to 0 will be transformed into the highest cost of 10. Ruggedness values close to 223 will be transformed into the lowest cost of 1.

The parameters are saved.

You'll save the model.

Next, you'll run the tool.

After a few moments, the resulting layer is added to the Contents pane.

You want to invert the symbology of the new layer, so that the low cost values appear in green and the high cost values in red.

The map updates.

The Ruggedness_Cost layer looks relatively similar to the Ruggedness layer, but it represents cost values varying between 1 and 10.

In the following example image, the cell has a cost value of about 3.39.

The tool is added as Rescale by Function (2) to differentiate it from the first one you used for Ruggedness.

You can see that the original Distance_to_Roads values vary from 0 to 10,961.62.

• For Output raster, change the name of the output layer to Distance_to_Roads_Cost.
• For Transformation function, choose MSLarge.
• For From scale, type 10. For To scale, type 1.

Distance_to_Roads values close to 0 will be transformed into the highest cost of 10. Distance_to_Roads values close to 10,961 will be transformed into the lowest cost of 1.

Next, you'll run the tool.

The tool runs.

You want to invert the symbology of the new layer, so that the low cost values appear in green and the high cost values in red.

The map updates. For more context, you'll make the Hillshade layer show through.

The map updates.

Here again, the Distance_to_Roads_Cost layer looks relatively similar to the Distance_to_Roads layer, but it represents cost values varying between 1 and 10.

First, you'll add the Land Cover layer as an input data variable.

Next, you must specify the output name and the one-to-one transformations.

• For Reclass field, verify that ClassName is selected.
• Below the Reclassification table, verify that Unique is selected.

In the Reclassification table, the Value column is populated with the land cover types. In the New column, you will replace the defaults with 1 to 10 cost values based on your assessment of each land cover type. Mountain lions feel the most safe in forest and shrub land cover types, although they can also make use of agricultural or open areas. Developed lands are not safe and should receive a high cost value. Finally, water-covered areas will get the highest cost value, since they are inhospitable for mountain lions.

Next, you'll run the tool.

After a few moments, the resulting layer is added to the Contents pane.

The new Land_Cover_Cost layer displays, but the Distance_to_Roads_Cost layer shows underneath in the ocean portion of the study area.

You'll turn that latter layer off.

The map updates.

The colors representing the cost values were assigned at random. You will change the symbology to make it more meaningful.

You will invert the symbology of the new layer.

The map updates.

The lower cost land cover types are mostly located in the mountain ranges with a lot of forest and shrubland. The higher cost land cover types tend to be located in the flatter low-lying valleys where most of the human development has taken place. The mountain lions core habitats are clearly nested in midst of low cost areas. This makes sense, since the lions would naturally gravitate toward low stress land cover.

• For Reclass field, verify that Protected_Status is selected.
• Below the Reclassification table, verify that Unique is selected.

In the Reclassification table, the Value column is populated with the Protected_Status levels. In the New column, you'll replace the defaults with cost values between 1 and 10, based on your assessment of each protection status category. Mountain lions feel the least stress in the most protected areas, so these should receive the lowest cost value. Unlike with land cover, you'll also reclassify NODATA cells, which correspond to locations that are not tracked in Protected Areas database and are usually very high stress areas like urbanized and other development-prone locations. NODATA cells should receive the highest cost value.

Next, you'll run the tool.

After a few moments, the resulting layer is added to the Contents pane.

The colors representing the cost values were assigned at random. You will change the symbology to make it more meaningful.

The map updates.

The lowest cost areas (in dark green and light green) in terms of protection status are located in the mountain ranges. Many areas are not protected (in orange or red). It doesn't mean that mountain lions couldn't go through them at all, but they would tend to be higher stress.

You will connect the four cost layers to the tool.

Next, you'll edit the tool parameters to provide your chosen weight for each input.

The four transformed layers are listed as input rasters. The default weight for each one is 1, which means they all have the same weight. You will need now change some of those weights. You decide that the land cover criterion should play a heavier role, as you don't want the mountain lions passing through urban areas. Similarly, you decide that the ruggedness criterion is very significant, as mountain lions don't do well in flat, non-rugged areas. As a result, you will increase the weights of these two criteria to ensure that the mountain lions avoid urban areas and avoid flat areas: the higher weights will make these unwanted areas significantly more costly. Because you created a model of the workflow, you can always go back and adjust the weights afterward to fine-tune your results.

Next, you'll run the tool.

After a few moments, the resulting layer is added to the Contents pane.

You'll invert the symbology of the new layer, so that the low cost values appear in green and the high cost values appear in red.

The map updates. For more context, you'll make the Hillshade layer show through.

The map updates.

This Cost_Surface layer simulates how the mountain lions will experience the landscape as they move through it. The green areas have the lowest cost in terms of stress and danger, while the red areas have the highest cost, and the beige areas have a medium cost. You can see that almost all low and medium cost areas are located in mountainous regions. The core mountain lion habitats are all located in areas that are low cost (dark green or light green). This makes sense, since mountain lions would naturally want to live in low stress and low danger areas. It is a good indication that, although it's your first attempt and your model could be refined further, your results are good.

Next, you'll add the Optimal Region Connections tool.

The input regions that you want to connect can be either raster data or vector data. The Core Mountain Lion Habitats layer is a polygon vector layer.

You'll enter the names for the tool's two output layers.

This is the primary output layer. It will contain the optimal network of paths (or lines) for mountain lions to move from one habitat patch to any other patches (possibly passing through another patch).

This is a secondary optional output identifying all paths from each patch to each of its closest (or lowest cost) neighbors. The paths to lowest cost neighbors can possibly provide alternative paths to reach specific patches.

You'll run the Optimal Region Connections tool and explore the results.

After a few moments, the resulting two layers, Mountain_Lion_Paths and Mountain_Lion_Paths_Neighbors, are added to the Contents pane.

The new layers are line vector layers. The default symbology might be difficult to see, so you'll change it.

The map updates.

You'll first examine the Mountain_Lion_Paths layer.

The Mountain_Lion_Paths layer represents the optimal network of paths to connect the four habitat patches with the least cost. You can observe that the paths primarily go through darker green and lighter green areas of the Cost_Surface layer. This makes sense since these are the lower cost areas. The paths only go through red areas when there are no other alternatives. This network is the most efficient way of connecting the four patches, and it represents the optimal locations for creating mountain lion corridors.

This layer contains the same paths as the previous layer, but it offers additional paths that connect each habitat patch to its lowest cost neighbors. These extra routes are not necessary, but they provide supplementary options. Similarly, you can see that they primarily go through dark green and light green areas, whenever possible.

The model redraws to fit entirely in the window.

The positioning of your model elements might not be perfectly aligned and equally spaced. You'll use the Auto Layout functionality to reorganize the model optimally.

The model redraws to optimize the placement of the elements.

To better show the model's structure, you'll reorganize it into three sections or groups:

• Preprocess datasets
• Create cost surface
• Find least-cost paths

The elements you selected are now encapsulated into a group. You'll rename it.

The first group is complete, and you'll now create the second one. First, you'll move the Core Mountain Lion Habitats element out of the way.

The second group is complete. You'll now create the third one.

Your model is now finalized.

The model you built is located there.

The model reopens and is ready to be used, either to run it or edit it further.

Validate verifies that all data elements and parameter values are valid. It also resets the model status, so that it's ready to run again.

You could now run the entire model as many times as you want by clicking the Run button. However, note that the symbology styles you applied to the result layers earlier in the workflow would be lost. You won't run it now.

You saw earlier that the proposed paths go through the lowest cost areas within the cost surface raster, whenever possible. When there is no other alternative, they might be passing through a few short higher-cost stretches. Overall, these paths look like good options for developing wild corridors to connect the core mountain lion habitats. You will now look at the four individual criteria cost layers individually to check how the paths fit each one.

The paths are almost only through high ruggedness terrain. From this criteria's standpoint, the paths show great promise for high quality mountain lion corridors.

Even though the paths are as low cost as possible, as expected, there is no way to avoid near-road locations in this study area. Later in the workflow, you will make a more fine-grained distinction between highways and small roads and learn about possible approaches to tackle this challenging high-road density situation.

The paths are almost only through low-cost land cover. From this criteria's standpoint, the paths show again great promise for high quality mountain lion corridors.

The optimum paths do their best to go through protected areas, but much of the valleys and some of the mountainous regions are unprotected, and the paths can't avoid going through them. Later in the workflow, you will learn how this situation could be improved by changing the protection status in the stretches of land along the paths.

• For Input Features, choose Mountain_Lion_Paths.
• For Output Feature Class, type Wildlife_Corridors.
• For Distance, type 1000, and choose Meters.
• For Dissolve Type, choose Dissolve all output features into a single feature.

The Wildlife_Corridors buffer appears on the map in semitransparent blue.

On the Protected_Status_Cost layer, the areas that are not protected appear in red and orange.

These are the areas that would need to be turned into protected land, whenever possible.

A copy of the Roads layer is added. You'll rename it.

You'll create the definition query.

The Carto attribute stores a numeric value that encodes the road types from largest (1) to smallest (6). The query will keep only the roads with Carto values of 1 and 2, which are highways and highway ramps.

The map should look like the following example image.

You'll choose a more visible symbol for the Highways layer.

On the map, the highways now display as thick red lines. You can identify four locations where a proposed path crosses a highway.

This number of intersections between paths and highways is not surprising, as the Los Angeles area is known for its extensive highway system. Examining the map, it's clearly not possible for the paths to avoid crossing these highways, even if considering taking longer detours. Therefore, the construction of overpasses should be considered at these four intersections. You'll look closer at one of these locations.

You'll switch to an imagery basemap to get a better sense of the situation on the ground.

The map updates displaying a photographic view of the area along with some cartographic indications.

You can see that the path needs to cross not only Route 126 (or Henry Mayo Dr.), but also a railroad from the Southern Pacific company. A single overpass could probably go over both at the same time, since they are very close to each other. A bit further south, there is also the Santa Clara River that would need to be reviewed, to see if it's passable by mountain lions. If it isn't, some type of passage would also need to be arranged there.