Predict deforestation in the Amazon rain forest

The Amazon rain forest spans nine countries and millions of square kilometers. Due to deforestation, current estimates indicate only about 80 percent of the original rain forest remains. In this tutorial, you'll use ArcGIS Pro to predict how much deforestation would be caused by a road proposed to be built across the Brazilian state of Rondônia. First, you'll open a map of roads and deforestation in the state. You'll create a buffer around the roads and intersect it with deforestation to calculate the percentage of area around roads that is deforested. Then, you'll use that percentage to estimate the area of deforestation around the proposed road. Throughout the workflow, you'll learn how to use multiple basic ArcGIS Pro geoprocessing tools to analyze data.

This tutorial was last tested on June 4, 2026, using ArcGIS Pro 3.7. If you're using a different version of ArcGIS Pro, you may encounter different functionality and results.

Requirements

• ArcGIS Pro (see options for software access)

Compare roads and deforestation

To determine how much deforestation a proposed road may cause, you first need to understand existing patterns of deforestation around roads. You'll analyze the relationship between roads and deforestation and quantify the percentage of land that is deforested within a certain distance from roads. Later, you'll use this value to predict deforestation around the proposed road.

Explore deforestation

First, you'll download road and deforestation data for the state of Rondônia, Brazil. You'll explore the data visually to better understand it before you analyze it.

1. Download the Amazon_Deforestation project package.
2. Browse to the downloaded file and double-click it to open the project in ArcGIS Pro. If necessary, sign in using your licensed ArcGIS account.

   Note:

   If you don't have access to ArcGIS Pro or an ArcGIS organizational account, see options for software access.

   The project contains a map of Rondônia, Brazil.

   

   The map contains several layers, identified with a legend in the Contents pane. Other than the state boundary and the boundary of the Amazon ecoregion, the only other visible layer shows roads throughout the state. The Roads layer is symbolized to show two types of roads. Official roads are thicker and white. They represent roads built and managed by the government and generally connect major population centers. Unofficial roads or roads under assessment are thinner and purple. They are roads that were built informally, usually by loggers or farmers.

   The map also contains a layer showing the proposed road and a layer showing deforestation, but these layers are turned off.

   Note:

   This map already has data layers that have been symbolized. To learn how to add data to a map and symbolize it, as well as how to perform several other basic tasks, try the tutorial Get started with ArcGIS Pro.

3. In the Contents pane, check the Deforestation check box.

   

   The layer turns on. Due to the density of roads, it's difficult to see at this zoom level. You'll navigate to a bookmark that has been prepared for you to see the deforestation in more detail.

4. On the ribbon, click the Map tab. In the Navigate group, click Bookmarks and choose Deforestation.

   

   The map zooms in to one specific area of deforestation.

   

   Deforestation typically occurs in small parcels, not large unbroken swaths. On the map, it has a patchy appearance and tends to follow roads.

   Typically, small strips of forest are deforested through a process called slash-and-burn agriculture. In slash-and-burn-agriculture, farmers cut and burn plots of forest to create fields. The burned biomass serves as fertilizer for agriculture on the cleared land. This agricultural technique has been practiced for centuries throughout the world, including the Amazon. In small amounts, slash-and-burn agriculture can be sustainable. But when widespread, massive areas are cleared in a small amount of time and it can drastically affect an ecosystem.

5. On the map, pan and zoom as necessary to explore other areas of deforestation.

   There is a strong relationship between roads and deforestation. In fact, 95 percent of deforestation in the Amazon rain forest occurs within 5.5 kilometers of a road. Roads allow access to the otherwise impenetrable rain forest and facilitate the transportation of lumber. Unlike official roads, which connect cities, unofficial roads access deeper areas of the rain forest and connect rural properties.

   Note:

   The roads and deforestation data in this tutorial is from 2014, so you may see differences between it and the more recent imagery in the basemap.

6. On the ribbon, on the Map tab, click Bookmarks and choose Rondônia.

   You navigate back to the full extent of the data.

Create a buffer around roads

You want to know the percentage of deforestation within a certain distance of roads. First, you'll create a buffer around the roads. A buffer is a feature that extends a specified distance from other features. It's useful for finding out whether one type of feature (in this case, deforestation) is close to another type (in this case, roads). Since almost all deforestation occurs within 5.5 kilometers of roads, you'll use this distance to create your buffer.

1. On the ribbon, click the Analysis tab. In the Geoprocessing group, click Tools.

   

   The Geoprocessing pane appears. This pane contains a large number of tools you can use to analyze data.

2. In the Geoprocessing pane, in the search box, type Pairwise Buffer. In the list of search results, click Pairwise Buffer.

   

   Note:

   The Pairwise Overlay toolset contains tools that provide an alternative to a number of classic overlay tools for functional and performance considerations. See Comparison of classic overlay tools to pairwise overlay tools for details.

   The Pairwise Buffer tool appears. To run a geoprocessing tool, you must first set parameters to determine how the tool will run. First, you'll choose the input features to create a buffer around.

3. For Input Features, choose Roads.

   Next, you'll choose the name of the output feature class that the tool will create.

4. For Output Feature Class, delete the existing text and type Roads_Buffer_Default.

   

5. For Distance, type 5.5. Change the unit of measurement to Kilometers.

   

   The only other parameter you need to change is the dissolve type. By default, the tool creates a buffer for each feature in the input layer. Because the Roads layer has many features, the tool would create a large number of overlapping buffer features. By changing the dissolve type, the tool will create a single feature as its output.

6. For Dissolve Type, choose Dissolve all output features into a single feature.

   

   Tip:

   If you're unsure what a parameter does, point to the parameter and point to its information button.

7. At the bottom of the Geoprocessing pane, click Run.

   The tool runs. When it finishes, the Roads_Buffer_Default layer is added to the map and the Contents pane.

   

   The buffer universally extends 5.5 kilometers from all roads. The buffers have significant overlap with the Deforestation layer, although not uniformly. The northern part of the buffer has many areas that are near roads but have relatively little deforestation.

   Additionally, the buffer extends past the study area in some places. You'll zoom in to learn more.

8. On the map, zoom in to the southern part of the state.

   

   Your roads and deforestation data only extends to places within the state of Rondônia (the thick purple boundary) that are also within the Amazon ecoregion (the dashed green boundary). Though there are roads and deforestation outside of this area, as evidenced by the imagery in the basemap, it's not accounted for in your data.

Intersect the buffer and the study area

Because the buffers extend past the study area and the deforestation data doesn't, calculating the percentage of deforested land within the buffer would lead to inaccurate results. To account for this problem, you'll run another geoprocessing tool, the Intersect tool, to create a buffer layer that only contains area inside the state and Amazon ecoregion.

1. In the Geoprocessing pane, click the Back button.

   

2. Search for Intersect. In the list of search results, click the Intersect tool.

   The Intersect tool takes two or more input layers and creates an output layer that contains the places where all of the input layers intersect. In this case, it'll find buffer areas that intersect the state and ecoregion boundaries.

   Note:

   You'll use the Intersect tool instead of the Pairwise Intersect tool because the Intersect tool allows for more than two inputs.

3. For Input Features, click the Add Many button.

   

4. Check the Roads_Buffer_Default, Amazon Ecoregion, and Rondônia check boxes and click Add.

   The three layers are added as inputs.

5. For Output Feature Class, delete the existing text and type Roads_Buffer.

   

6. Click Run.

   The tool runs and the layer is added to the map. You'll remove the original buffer layer, which you no longer need.

7. In the Contents pane, right-click Roads_Buffer_Default and choose Remove.

   

   The layer is removed from the project. On the map, only the intersected buffer is shown. It does not extend past either the state boundary or the ecoregion boundary.

   

8. On the ribbon, click the Map tab. In the Navigate group, click Bookmarks and choose Rondônia.

Clip deforestation to the buffer

Your goal is to calculate the percentage of land within the buffer that is deforested. To do so, you'll also need a layer that only contains deforestation within the buffer. Although most deforestation happens within 5.5 kilometers of roads, there are some small patches of deforestation outside the buffer area. You'll clip the deforestation layer to only include areas within the buffer.

1. In the Geoprocessing pane, click the Back button.
2. Search for Pairwise Clip. In the list of search results, click Pairwise Clip.

   The Pairwise Clip tool is similar to the Intersect tool, but removes data from a single input layer instead of combining multiple input layers. In many cases, such as this one, the tools can be used interchangeably, but you used the Intersect tool previously because it allowed for more than two inputs. In this case, you only need to clip deforestation to one layer (the buffer layer), so you'll use this tool instead.

3. For Input Features, choose Deforestation. For Clip Features, choose Roads_Buffer.
4. For Output Feature Class, delete the existing text and type Deforestation_Clip.

   

5. Click Run.

   The tool runs and the new layer is added to the map. Because it is so similar to the original Deforestation layer, the map looks similar to before.

6. In the Contents pane, uncheck the original Deforestation layer.

Calculate deforested area

You've created two layers. One, Roads_Buffer, shows the area within 5.5 kilometers of roads in your sample area. The other, Deforestation_Clip, shows deforestation within that buffer. Your next goal is to calculate the percentage of the buffer area that is deforested. You can perform this calculation by comparing the area of each of the two layers. All polygon feature classes automatically calculate their area. The area value is contained in the layer's attribute table.

1. In the Contents pane, right-click Deforestation_Clip and choose Attribute Table.

   

   The attribute table appears. This table displays all the data, or attributes, associated with features. The Shape_Area field contains the field's area.

   The unit of measurement for the area value is not shown in the table, but specified by the layer's projection. The projection of the data is South America Albers Equal Area. Output layers made from geoprocessing tools use the same projection as their input. For the South America Albers Equal Area projection, the linear unit is meters, so the Shape_Area field in the attribute table is in square meters.

   Note:

   To learn more about projections and how to determine the projection of your data, try the tutorial Choose the right projection.

   Though square kilometers would be a more appropriate unit of measurement for deforestation, you're only trying to find the percentage of deforested area, so all that matters is that this layer and the buffer layer have the same unit of measurement. In this case, they do.

2. Right-click the Shape_Area value and choose Copy.

   

   Next, you'll compare the deforestation area to the buffer area.

3. In the Contents pane, right-click Roads_Buffer and choose Attribute Table.

   The buffer layer's attribute table appears, replacing the deforestation layer's attribute table.

   Tip:

   You can switch between the two tables by clicking the tabs above them.

   The buffer layer also has a Shape_Area field. You'll create and calculate a new field that compares it to the deforestation area.

4. On the attribute table toolbar, click Calculate.

   

   Options appear to create and calculate a field. First, you'll choose the field type and name. The field type determines the type of data in the field. You'll choose the double field type, which contains large numbers with decimal points.

5. Click the Text Field Type button and choose Double (64-bit floating point).

   

   Next, you'll set the field to calculate. You can either calculate an existing field by choosing it from a drop-down menu or create a new field by typing the name of a field that doesn't exist. You'll create a new field.

6. For Target Field (Existing or New), type Percent_Deforested.

   

   Next, you'll create an expression to calculate the field. To find the percentage, you'll divide the area of the Deforestation_Clip layer by the area of the Roads_Buffer layer and multiply the result by 100.

7. In the Enter an expression to calculate field values box, paste the value you copied from the Deforestation_Clip table.

   

   Tip:

   If the area value is no longer in your clipboard, you can return to the Deforestation_Clip table by clicking the tabs above the table and copy it again. Alternatively, you can paste the value 77701073236.492386.

   Next, you'll create an expression to divide this value by the area of the buffer layer and multiply the result by 100 to make it a percentage. Rather than copying and pasting the buffer layer's Shape_Area value, you can refer to the field directly using specific notation.

8. After the pasted value, type the division sign (/) and type !Shape_Area!. Type the multiplication sign (*) and type 100.

   The final expression reads 77701073236.492386 / !Shape_Area! * 100.

   

9. Click Calculate.

   

   The new field is created and calculated using the expression you created.

10. If necessary, scroll to the end of the table to see the new field.

    

    Based on your calculation, 49.6 percent of land within 5.5 kilometers of a road is deforested. If a new road were built in this sample area, you could predict that a similar percentage of land within 5.5 kilometers of that road would become deforested.

    Creating a new field involves editing the data. Before you continue, you'll save the edits. If you don't, it might not be possible to use the data for future analysis.

11. On the ribbon, click the Edit tab. In the Manage Edits group, click Save.

    

12. If the Save Edits window appears, click Yes.

    Your edit is saved. Now that you have the Percent_Deforested value, the Deforestation_Clip layer is no longer necessary. You'll remove it. You'll keep the Roads_Buffer layer because it has the percentage value in its attributes, but you'll turn off the layer so it's hidden on the map.

13. Close the Roads_Buffer and Deforestation_Clip tables.
14. In the Contents pane, uncheck Roads_Buffer. Right-click Deforestation_Clip and choose Remove.
15. On the Quick Access Toolbar, click the Save Project button.

    

You've calculated the impact of deforestation in Rondônia due to the road network. Next, you'll use this percentage to predict the impact of the proposed road.

Predict the impact of a proposed road

Now that you've calculated the percentage of deforested area within 5.5 kilometers of roads, you'll use that percentage to estimate the area in square kilometers that would be deforested if a proposed road were built. You'll create a buffer around the road, remove any already deforested land from the buffer, and calculate the predicted deforestation.

Buffer the proposed road

First, you'll create a buffer around the proposed road to represent the area where deforestation might occur nearby. Because the percentage deforestation value you calculated used a 5.5-kilometer buffer area, you'll keep the distance the same.

1. In the Contents pane, check the Proposed Road check box.

   The proposed road appears on the map. It connects the center of the road network to the western edge, cutting across the rain forest.

   

   Note:

   This proposed road is fictitious for the purposes of this tutorial and does not reflect an actual road proposal.

2. In the Geoprocessing pane, click the Back button. Search for and open the Pairwise Buffer tool.
3. For Input Features, choose Proposed Road. For Output Feature Class, delete the existing text and type Proposed_Buffer_Default.
4. For Distance, type 5.5. Set the unit of measurement to Kilometers.

   

   Because there is only a single feature to buffer this time, you don't need to change the dissolve type.

5. Click Run.

   The tool runs. The new layer is added to the map.

6. On the map, zoom in to the proposed road.

   

Remove deforested area

Some of the area within the buffer may have already been deforested. Since deforested area cannot be deforested again, you'll remove these areas from the buffer.

1. In the Contents pane, turn on the Deforestation layer.
2. If necessary, on the map, zoom in to compare the layers.

   

   There are some areas where deforestation has already occurred. You don't want to include these areas in your analysis, so you'll remove them with the Pairwise Erase tool. This tool subtracts parts of one layer that overlap another layer.

3. In the Geoprocessing pane, click the Back button. Search for and open the Pairwise Erase tool.
4. For Input Features, choose Proposed_Buffer. For Erase Features, choose Deforestation.
5. For Output Feature Class, type Erased_Buffer.

   

6. Click Run.

   The tool runs and the layer is added to the map.

7. In the Contents pane, right-click Proposed_Buffer and choose Remove.

   The Erased_Buffer layer is a single multipart feature. Areas of deforestation have been erased from it. The erased areas will no longer be counted in your analysis.

   

8. If necessary, zoom back out until you can see the entire proposed road buffer.

Estimate deforestation

To find out how much of the buffer area would become deforested if the proposed road were built, you'll create and calculate a field that multiplies its area by the percentage value you obtained previously.

1. In the Contents pane, right-click Erased_Buffer and choose Attribute Table.

   Like the other layers, this layer contains a Shape_Area field. The field is in square meters. Though the unit of measurement didn't matter when calculating the percentage of area deforested around roads, you now want to predict the amount of area that will be deforested, not a percentage. While calculating the field of predicted deforestation, you'll divide the area by 1,000,000 to convert it from square meters to square kilometers.

2. On the attribute table toolbar, click Calculate.
3. Click the Text Field Type button and choose Double (64-bit floating point). For Target Field (Existing or New), type Potential_Deforestation.

   

   Next, you'll create the expression to calculate the field. First, you'll convert the area from square meters to square kilometers.

4. For Enter an expression to calculate field values, type (or copy and paste) !Shape_Area! / 1000000.

   

   Next, you'll multiply this value by the percent of deforested area you calculated earlier. This value is in the attribute table for the Roads_Buffer layer.

5. In the Contents pane, right-click Roads_Buffer and choose Attribute Table.
6. In the Roads_Buffer attribute table, scroll to the Percent_Deforested field. Right-click the value (49.60041) and choose Copy.
7. Close the Roads_Buffer table and return to the Erased_Buffer attribute table.
8. At the end of the expression, add a multiplication sign (*) and paste the value you copied (49.60041).

   

   Because the value you're multiplying by is a percentage, you'll divide it by 100 to turn it into a fraction.

9. After the percentage value, add a division sign (/) and type 100. Add parentheses around the first and second part of the expression (each side of the multiplication sign).

   The full expression reads (!Shape_Area! / 1000000) * (47.07733539681277 / 100).

   

10. Click Calculate.

    The field is calculated and added to the end of the table.

    

    According to this analysis, approximately 671 square kilometers would become deforested if the proposed road were built.

11. On the ribbon, on the Edit tab, in the Manage Edits group, click Save. If the Save Edits window appears, click Yes.
12. Close the attribute table and the Geoprocessing pane. Save the project.

In this tutorial, you performed analysis to answer a geographic question. Using the Buffer, Intersect, Clip, and Erase tools, you compared layers of roads and deforestation to determine how much deforestation occurs near roads and predict the deforestation that would occur if a proposed road were built.

The analysis you performed in this tutorial was relatively simple and did not account for other factors that might influence deforestation, such as protected areas or the elevation of the terrain. By incorporating more layers, you could perform an even more robust analysis with a more sophisticated prediction. This tutorial's goal was only to introduce you to some of the basic geoprocessing tools in ArcGIS Pro, but there are hundreds of tools available. To learn more about the available tools, see ArcGIS Pro geoprocessing tool reference. For a more advanced analysis workflow, try the tutorial Determine the most dangerous roads for drivers.

You can find more tutorials in the tutorial gallery.

Acknowledgements

• The Amazon Ecoregion layer is from the World Wildlife Fund (WWF) and used in this tutorial with permission from Aurélie Shapiro of WWF Germany.
• The Rondônia layer is derived from the World Administrative Divisions layer and shared under the Esri Master License Agreement.
• The Roads layer is from Imazon and used in this tutorial with permission from Carlos Souza of Imazon.
• The Deforestation layer is from the Instituto Nacional de Pesquisas Espaciais (INPE) and shared under the CC BY-SA 4.0 license.
• The Proposed Road layer is fictitious data created by the tutorial author.
• World Imagery layer sources: Esri, Vantor, Earthstar Geographics, and the GIS User Community