Monitor forest change over time

Forests are an important natural resource, providing wood products, wildlife habitat, clean water and air, carbon sequestering, recreational opportunities, and more. Monitoring forest disturbances, such as tree logging activities, forest fires, and pest infestations, is an important task for forest management. The West Cascades ecoregion in Oregon contains large forests that include both commercial timber production land and federal and state preserved land. The state plans to use ArcGIS Pro time series image analysis capabilities to drive forest management practices. For the data, it will rely on Landsat satellite imagery archives spanning several decades.

In this tutorial, you'll use tools based on the LandTrendr algorithm (Landsat-based detection of Trends in Disturbance and Recovery) to better understand the spatial and temporal patterns of timber harvest.

This tutorial was last tested on May 26, 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)
• ArcGIS Image Analyst

Explore the data

An image time series is a set of images taken at different times in the same location, which is useful for understanding how an area has changed over time. For example, an image time series of a forested area might show when trees were cut down and later regrown, or when and where fires or pest infestations occurred. Because these disturbances usually accompany abrupt changes in the pixel values of the images, image time series analysis uses algorithms to detect these changes.

The ArcGIS Image Analyst extension offers two time series change detection algorithms for imagery. One is the LandTrendr algorithm, which relies on Landsat imagery and is commonly used for forest applications. Because Landsat-TM imagery offers continuous coverage of Earth's land surface since 1984, it's possible to use it to monitor how land cover has changed over a period of many years.

First, you'll become familiar with a multidimensional raster of the West Cascades area in Oregon. A multidimensional raster is the dataset model used to represent an image time series. Then, you'll prepare the image time series for analysis and explore how land cover has changed at a single pixel in the image.

Explore a multidimensional raster

First, you'll download a project that contains all the data for this tutorial and open it in ArcGIS Pro. Then, you'll familiarize yourself with the multidimensional raster dataset you'll use throughout the tutorial.

1. Download the Forest_Disturbance_Analysis 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 imagery of the West Cascades area in Oregon.

   

   The Contents pane lists the layers in the project. The WestCascade.crf layer contains a time series of 78 satellite images of the Landsat Analysis Ready Imagery (ARD) type, collected between 1984 and 2020 over the West Cascades. Each image is represented as a multiband raster, which contains blue, green, red, infrared, and shortwave infrared spectral bands. The images were all collected in summer so the trees have leaves and are in a similar growth stage, which is important due to the requirements of the LandTrendr algorithm you'll use in this tutorial. Furthermore, only images with a cloud cover of less than 10 percent were included. All of these rasters are gathered into a single multidimensional raster stored in the CRF format.

   Note:

   To apply the workflow in this tutorial to your own study area, you can create a similar multidimensional raster by downloading Landsat ARD images from USGS EarthExplorer for your area of interest. To learn how to clean up the images and generate a multidimensional raster, read Clean up your Landsat imagery: removing cloud and cloud shadow.

   Next, you'll browse the 78 images included in WestCascade.crf and see when each one was collected.

3. In the Contents pane, click the WestCascade.crf layer to select it.

   

   When a multidimensional dataset is selected, the Multidimensional tab appears on the ribbon. This tab contains tools to work with multidimensional rasters.

4. On the ribbon, click the Multidimensional tab.

   

5. In the Current Display Slice group, expand the StdTime drop-down list. Click the second value in the list, starting with the date 1984-07-19.

   

   The map updates to display the image taken on July 19, 1984. In an imagery time series like WestCascade.crf, each image is called a time slice. The image you just displayed is the second time slice in the series.

   

   In the Contents pane, the layer's legend shows the spectral bands used to display the image.

   

   The image is displayed using the natural color band combination, where Band 1 is blue, Band 2 is green, and Band 3 is red. This band combination approximates how the landscape would appear to the human eye. The image also contains an infrared band (Band 4) and two short wave infrared bands (Band 5 and Band 6). You'll use these bands later, though they are not currently displayed.

   Note:

   To learn more about multispectral imagery and band combinations, see the tutorial Get started with imagery.

   The image mostly displays forested areas (in green), but there are also some areas in the center of the image where the trees were cut (in beige).

   

   You can also animate the time series to view all of the images in a sequence.

6. On the Multidimensional tab, in the Current Display Slice group, click the Play Slices Along StdTime button.

   

   The time series plays as an animation, with a new image shown every few seconds. The forest sometimes changes significantly between images and sometimes changes only a little.

   Note:

   Some images have NoData areas that appear as holes in the layer. These areas correspond to cloud pixels that have been masked out and will not affect the analysis.

7. When finished, click the Play Slices Along StdTime button to stop the animation.

   Next, you'll review the layer's properties.

8. In the Contents pane, double-click the WestCascade.crf layer.

   The Layer Properties window appears.

9. Click the Source tab. Expand the Multidimensional Info, SR and StdTime sections.

   

   The Count value indicates that the dataset contains 78 time slices. The pixel value, SR, corresponds to the surface reflectance. The temporal Extent values range from 1984-07-03 to 2020-08-23.

10. Close the Layer Properties window.

    In summary, the WestCascade.crf layer contains 78 multispectral images. Each image contains 6 spectral bands. In total, that means the dataset contains 468 rasters. The following image represents this multidimensional raster structure:

    

Apply an index formula

To analyze the image time series, you'll use the LandTrendr algorithm. LandTrendr works with a multidimensional raster where each time slice is a single band raster. However, your time slices contain 6 bands. To derive a single band raster for each time slice, you'll apply an index formula to your time series. You'll use the Normalized Burn Ratio (NBR) index, which is good for distinguishing healthy forest from forest disturbances and is often used to detect forest burn scars.

Note:

To learn more about the NBR index, see the tutorial Assess burn scars with satellite imagery.

1. In the Contents pane, ensure the WestCascade.crf layer is selected.
2. On the ribbon, click the Imagery tab. In the Tools group, click Indices.

   

3. In the list of indices, under Landscape, click NBR.

   

   The NBR window appears. The NBR index combines the Near Infrared and Shortwave bands using a mathematical formula. In your layer, those bands correspond to Band 4 and Band 6, respectively.

4. In the NBR window, for Near Infrared Band Index, choose 4 - Band_4. For Shortwave Infrared Band Index, choose 6 - Band_6.

   

5. Click OK.

   A layer named NBR_WestCascade.crf is added to the map. This layer is still a multidimensional raster, but it contains only one NBR raster band per time slice.

   Note:

   The NBR tool is a raster function that generates a new raster layer dynamically. This approach saves processing time, but the new layer exists only in your computer's memory. If you remove the layer from your project, it will be deleted and you'll need to recreate it.

   By default, the layer is displayed using the black and white stretch renderer. You'll change the symbology to highlight the differences between healthy vegetation and disturbances (the absence of healthy vegetation).

6. In the Contents pane, right-click the NBR_WestCascade.crf symbol and expand the color ramp drop-down list. Check the Show names check box and choose the Pink-Green (Continuous) color ramp.

   

   On the map, the symbology updates. The highest NBR pixel values represent healthy forest and display in green. The lowest NBR pixel values represent disturbances and display in purple. The NBR pixels with a medium value are represented in white or light green.

   

   The new layer is also a time series.

7. In the Contents pane, ensure that the NBR_WestCascade.crf layer is selected.
8. On the ribbon, click the Multidimensional tab. In the Current Display Slice group, for StdTime, expand the drop-down list and choose the time slice for 1984-07-19.

   The map updates to display the image for that date, which is symbolized the same way.

9. Set StdTime back to the first time slice (1984-07-03).

Chart change for one pixel

Next, you'll use the NBR time series to start exploring how the forest has changed over time. Throughout this tutorial, you'll monitor forest change in many different ways. First, you'll monitor change for a single pixel using the Pixel Time Series Change Explorer pane.

Your map currently displays the NBR raster for July 3, 1984. If you picked a single pixel on this raster and observed how its values changed from 1984 to 2020, across all 78 time slices, you could represent the values on a graph, like in the following image:

Then, you could use the LandTrendr algorithm to find the overall trends in the graph. LandTrendr will find the inflection points where the pixel trajectory changes and fit the data into a piecewise linear model. The result will be a fitted curve made of several line segments that go from one inflection point to the next, like in the following image:

Each segment represents the change that occurred in a time period, characterized by a start time, an end time, a duration, a slope, and a magnitude (that is, how much change has occurred between the start and end times). In the example image, there are three segments:

• The first segment represents a healthy forest (high and steady NBR values).
• The second segment indicates a disturbance such as logging (rapidly decreasing NBR values).
• The third segment represents the forest recovery process, as the trees grow back (slowly increasing NBR values).

You'll generate a similar graph for your data.

1. In the Contents pane, ensure that NBR_WestCascade.crf is selected.
2. On the ribbon, on the Multidimensional tab, in the Analysis group, click the Temporal Profile drop-down arrow and choose Pixel Time Series Change Explorer.

   

   An empty chart view and the Chart Properties pane appear.

3. In the Chart Properties pane, on the Data tab, for Change detection method, choose LandTrendr.

   

   Note:

   A warning message may appear, stating that building a transpose (which is additional storage optimized for time series analysis) will improve performance. The NBR_WestCascade.crf raster function layer does not have a transpose, but the underlying WestCascade.crf multidimensional raster does. You may ignore the warning.

   For convenience, the point you'll graph is marked in the Pixel location layer.

4. In the Contents pane, check the Pixel location check box.

   

   A green point appears on the map. You'll navigate to its location using a bookmark.

5. On the ribbon, click the Map tab. In the Navigate group, click Bookmarks and choose the Pixel point bookmark.

   

   The map navigates to the pixel marked by the point. You'll graph this pixel.

6. In the Chart Properties pane, under Define a pixel location, click the Point button.

   

7. On the map, click the pixel of interest.

   

   A gray location point is added on the map. In the Chart Properties pane, a row is added to the pixel location list. Before generating the chart for that pixel location, you'll choose some styling options.

8. In the Chart Properties pane, in the pixel location row, under Symbol, click the point symbol and choose Cretan Blue.

   

   Note:

   To see an image name, point to it.

9. Set the size to 3.

   

10. Under Model Parameters, for Show fitted curve with, click the Color line and choose Electron Gold.

    

11. Change the line thickness to 3.

    

12. Under Model Parameters, click Fit and create chart.

    In the chart pane, a chart appears.

    

    The blue points are the pixel values for the 78 time slices. The orange line is the fitted curve calculated using the LandTrendr algorithm that shows the general change trends.

Explore the pixel change chart

Next, you'll explore the chart.

1. In the chart pane, point to one of the blue points.

   

   A pop-up shows the year of the time slice and its NBR value.

   The orange fitted curve is composed of several segments, each corresponding to an important stage in the pixel's history. The chart legend identifies the start and end times of each fitted curve segment.

   

   Tip:

   Click any item in the chart legend to turn the corresponding segment on and off on the chart.

   Currently, there are five line segments. This may not be the optimal number of segments for understanding the data. You'll change the number of segments in the fitted curve. Finding the optimal number of segments can be achieved through trial and error and will depend on your dataset.

2. In the Chart Properties pane, expand the Model Parameters group. Change Maximum Number Of Segments to 4.

   

   Note:

   For more information about each model parameter, see Analyze Changes Using LandTrendr.

3. Click Fit and update chart.

   The chart is updated to have only four segments:

   • The first segment has high NBR values and indicates stable mature forest.
   • The second segment shows a strong disturbance. The amount of healthy forest rapidly dropped to 0, probably because the area was logged.
   • The third segment shows a quick recovery curve, as tree seedlings were planted and grew relatively quickly.
   • In the fourth segment, the recovery continues more slowly, as the young trees develop to full maturity.
4. On the chart, point to the curve segments.

   

   A pop-up shows the fitting model for the segment. The fitting model is represented by a formula in the format ax + b, where a represents the slope and b the intercept. The RMSE (Root Mean Square Error) value is also shown.

   To better understand the changes in your area of interest, you'll examine the corresponding raster images.

5. On the chart view toolbar, click the Legend button.

   

   The legend turns off, expanding the chart.

6. On the chart, double-click a blue vertex to select a time slice.

   Tip:

   Alternatively, drag the pointer to draw a box around the vertex.

   

   The map updates with the image for the time slice you selected.

   Note:

   If selecting a specific point on the graph is difficult, you can zoom in to the chart. In the chart pane toolbar, click the Zoom Mode button and draw a box around the point of interest.

7. Visualize the four time slices highlighted in the following image:

   

   You see the following four time slices on the map:

   

   These time slices illustrate the changes at the pixel location, from mature forest to logging plot to young trees to mature forest again.

8. Close the chart view and the Chart Properties pane.
9. In the Contents pane, uncheck the Pixel location check box.
10. Right-click the NBR_WestCascade.crf layer and choose Zoom To Layer.

    

    You return to the full extent of the data.

11. On the Quick Access Toolbar, click the Save Project button.

    

    The project is saved.

You've set up the project in ArcGIS Pro, become familiar with the multidimensional raster, and explored forest change at a single pixel. Next, you'll expand your understanding of forest change to the entire image.

Detect disturbance and recovery

To analyze change for all pixels in the image, you'll first generate a change analysis raster. Then, you'll create a forest disturbance map to identify events like logging and fires throughout the forest's history. Finally, you'll create a forest recovery map to assess how long it took for disturbed areas to return to mature forest.

Create a change analysis raster

You'll use the Analyze Change Using LandTrendr tool to generate a change analysis raster. This tool uses the same LandTrendr algorithm as the Pixel Time Series Change Explorer tool, but it applies it to all the pixels in the multidimensional raster instead of only one pixel. The output will be a change analysis raster that captures information about the fitting curves for every pixel.

1. In the Contents pane, ensure that the NBR_WestCascade.crf layer is selected.
2. On the ribbon, click the Multidimensional tab. In the Analysis group, for the tool gallery, click the More button.

   

3. Click the Analyze Changes Using LandTrendr tool.

   

   The Geoprocessing pane appears with the tool.

4. For Output Multidimensional Raster, type WestCascade_change_analysis.crf.

   

   You'll leave the other parameters unchanged.

   Note:

   The output of this tool is a multidimensional raster with only one time slice per year. Before doing the LandTrendr change analysis, the tool chooses the best pixel values available in the input raster for each year. For a given year, the best pixel values are the ones from the time slice that is closest to the Snapping Date parameter (by default, 6-30, or June 30). If the best pixel value happens to be NoData because of a cloud, the tool takes the pixel value from the next closest time slice.

5. Click Run.

   Note:

   The process may run for 15 minutes or longer. It also may not show progress until it is completed.

   The process runs. When finished, the output raster appears in the Contents pane and its first raster band displays on the map.

   The WestCascade_change_analysis.crf layer is a multidimensional raster with 34 time slices, which is close to one per year (some years did not have any data). Each time slice is a multiband raster, where the bands store information about the LandTrendr fitting model. You'll explore and display these raster bands.

6. In the Contents pane, click the WestCascade_change_analysis.crf symbol.

   

   The Symbology pane appears.

7. In the Symbology pane, for Band, expand the drop-down menu.

   

   If the formula ax + b represents the fitting model for one segment of the fitted curve, the following apply:

   • The Slope band stores value a, which describes the change direction as either increasing or decreasing.
   • The Intercept band stores value b.
   • The Fitted_Value band stores the value interpolated from the fitting model for a given time. In this dataset, it is an approximation of the input NBR value for each time slice.
   • The RMSE band stores the root mean squared error of the fitted curve.
   • The Change_Magnitude band stores the change magnitude, or the difference between the fitted values at the beginning and end of the change period.

   Note:

   All the bands listed have a name that starts with Band_1, such as Band_1_Slope, because they were derived from the NBR raster, which has a single band per time slice named Band_1.

   Using this change analysis raster, you can extract change information such as date of change, change duration, and magnitude. In the rest of the tutorial, you'll use this multidimensional raster as the input to perform various change detection tasks.

   Currently, the Band_1_Slope band is displayed on the map. You'll display another band instead.

8. In the list of bands, choose Band_1_Fitted_Value.

   The map updates.

   

   The dark green areas represent the lowest NBR fitted values and the deep purple areas represent the highest. You can also visualize the Band_1_Fitted_Value band for different time slices using the StdTime option on the Multidimensional tab.

9. Close the Symbology pane.

Map disturbances

Now that you have a change analysis raster, you can use it to map other information. First, you'll map forest disturbances over time with the Detect Change Using Change Analysis Raster tool. You'll run the tool with parameters that will extract occurrences of abrupt decreases in the fitted NBR values, signifying a sudden change from healthy forest to an absence of trees.

1. In the Contents pane, confirm that the WestCascade_change_analysis.crf layer is selected.
2. On the ribbon, on the Multidimensional tab, in the Analysis group, expand the tool gallery and choose Detect Change Using Change Analysis Raster.

   

   The tool appears in the Geoprocessing pane.

3. For Output Raster, type Disturbance.crf.

   You'll set the change direction so only decreasing changes are detected, ensuring you capture areas where healthy forest was disturbed. You'll also track the time of earliest change, so you know when disturbances began.

4. For Change Direction, choose Decreasing. For Change Type, choose Time of earliest change.

   

   Before you run the tool, you'll also filter the results to designate events that last no longer than 4 years (signifying an abrupt change), with a decrease from a high NBR between 0.7 and 1 (healthy forest) to a low NBR between -1 and 0.35 (absence of trees). These pixel value ranges were chosen by reviewing typical values for healthy forest and logged or burnt areas in the Band_1_Fitted_Value raster bands.

5. Expand the Filter By Attributes section. Set the following parameters:
   • Check the Filter by Duration check box.
   • For Maximum Duration (in years), type 4.
   • Check the Filter by Start Value check box.
   • For Minimum Start Value, type 0.7.
   • For Maximum Start Value, type 1.
   • Check the Filter by End Value check box.
   • For Minimum End Value, type -1.
   • For Maximum End Value, type 0.35.

   

6. Click Run.

   The Disturbance.crf layer is created and added to the map. It is a single-band raster and not multidimensional. Each pixel value represents the date when a logging or fire event started.

7. Close the Geoprocessing pane.
8. In the Contents pane, turn off WestCascade_change_analysis.crf and NBR_WestCascade.crf.
9. On the map, click a disturbance area.

   A pop-up appears, displaying the start date of the disturbance event.

   

10. Close the pop-up.

Fill missing values

In the Disturbance.crf raster, there are some NoData pixels within the disturbance areas where there was not enough information to compute a valid model. However, it is likely that these areas have the same disturbance value as the surrounding pixels. You'll fill these NoData areas with the Statistics raster function, using the majority value within the nearby pixels.

1. On the ribbon, click the Imagery tab. In the Analysis group, click the Raster Functions button.

   

   The Raster Functions pane appears.

2. In the Raster Functions pane, search for Statistics. In the Statistical group, click Statistics.

   

   You'll set the tool parameters so that, for each NoData pixel, the tool will look at a box of 3 by 3 pixels around it and choose the value that occurs most frequently in that box.

3. For Raster, choose Disturbance.crf.
4. For Statistics Type, choose Majority. Confirm Number of Rows and Number of Columns are both set to 3.
5. Check the Only fill NoData pixels check box.

   

6. Click Create new layer.

   A new raster layer, Statistics_Disturbance.crf, is added to your map. This layer is smoother and cleaner, with fewer gaps in the disturbance areas. You'll rename the layer and change its symbology.

7. Close the Raster Functions pane.
8. In the Contents pane, click Statistics_Disturbance.crf to make its name editable. Type Smoother_Disturbance.crf and press Enter.

   

9. Click the Smoother_Disturbance.crf symbol.

   The Symbology pane appears. Rather than set the symbology yourself, you'll import a layer file prepared for you and included in your project.

10. In the Symbology pane, click the options button and choose Import from layer file.

    

    The Import Symbology window appears.

11. In the Import Symbology window, double-click Folders, Forest_Disturbance_Analysis, commondata, and userdata. Select Disturbance_symbology.lyrx.

    

12. Click OK.

    The new symbology, with a white to purple color ramp, is applied.

    

    Dark purple areas represent disturbances that happened in recent years, while white or light pink areas are disturbances that happened in earlier years. Most of the disturbances resulted from logging activities that started in the center of the region and expanded north and south over time. There were no logging activities in large areas to the east and west.

13. Close the Symbology pane.
14. In the Contents pane, right-click Disturbance.crf and choose Remove.

Inspect the disturbances

Next, you'll review the forest disturbance layer to better understand the logging and fire events that occurred in the West Cascades forest. First, you'll compare the events with the boundaries for protected forest areas managed by the federal and state governments.

1. In the Contents pane, turn on the Forest management boundaries layer.

   The areas without logging activities are in protected areas managed by the government.

   

2. On the ribbon, click the Map tab. In the Navigate group, click the Explore drop-down arrow and confirm Topmost Layer is chosen.

   

   This option ensures you'll only display pop-ups for the topmost layer listed in the Contents pane. In this case, that layer is Forest management boundaries.

3. On the map, click the large protected forest area on the east side of the map to display its pop-up.

   

   This area is a state park. You'll explore it in more detail.

4. Close the pop-up.
5. On the ribbon, on the Map tab, in the Navigate group, click Bookmarks and choose State park.

   The map navigates to the park. Some pixels in the middle of the park are symbolized in dark purple, which indicates a disturbance happened in recent years.

   

   Could this be a case of illegal logging in a protected forest? You'll learn more about that specific event.

6. In the Navigate group, click the Explore drop-down arrow and choose Visible Layers.

   

   This option will ensure that the pop-ups show the information for all the layers displayed, not just the top layer.

7. Click the disturbed area to display the pop-up.

   

   In the pop-up, the date of the disturbance is listed under Smoother_Disturbance.crf. This disturbance occurred in 2017.

8. Close the pop-up.

   You'll review the imagery corresponding to this area using a side-by-side display.

9. In the Catalog pane, expand Maps. Right-click Source_Images and choose Open.

   Note:

   If you don't see the Catalog pane, on the ribbon, click the View tab. In the Windows group, click Catalog Pane.

   

   The Source_Images map appears. It shows WestCascade.crf, the original imagery multidimensional raster.

10. Drag the Source_Images tab and dock it on the right side of the Disturbance_Analysis map.

    

    You'll link the two maps so that they show the same extent.

11. On the ribbon, click the View tab. In the Link group, click the Link Views drop-down arrow and choose Center And Scale.

    

12. Zoom in and pan to the disturbed area in the state park.

    The two maps update in sync. You'll view the imagery for a date not long after the disturbance event.

13. Click the Source_Images map tab to make it the active map.

    

14. In the Contents pane, select WestCascade.crf.
15. On the ribbon, click the Multidimensional tab. For StdTime, choose the 2018-07-17 time slice.

    The imagery updates. The area doesn't look bare, but instead appears brownish, suggesting the trees were not logged but affected by fire.

    

    Next, you'll review how some past or recent disturbance events display in the most recent imagery.

16. On the ribbon, click the Map tab. Click Bookmarks and choose Linked view.
17. Ensure that the Source_Images map tab is active. On the ribbon, on the Multidimensional tab, for StdTime, choose the most recent time slice, 2020-08-23.

    From the linked maps, it's possible to make the following observations:

    • Many of the oldest disturbance areas (white or light pink) are now filled with fully grown trees because the forest recovered.
    • Where disturbances occurred a few years ago (medium purple), there are young trees in a stage of regrowth.
    • For the most recent disturbance events (dark purple), there is mostly bare earth on the imagery.

    

18. Close the Source_Images map.
19. On the ribbon, click the View tab. Click the Link Views button to unlink the views.

Measure forest recovery

A forest is a dynamic system that includes cycles of disturbance and recovery. For instance, trees are cut for timber harvest and new trees are planted. Recovery is usually much slower than the disturbance event. To estimate the length of the recovery process for replanted trees to grow to maturity, you'll run the Detect Change From Change Analysis tool again. This time, you'll extract the start and end dates of the change periods where NBR values increased instead of decreased. Then, you'll calculate the number of years between the start and end dates. The full workflow requires the following steps:

• Find the change period for recovery and extract the starting date of the change.
• Run the tool again to extract the ending date of the change.
• Calculate the difference between the two dates (the number of days for recovery).
• Convert the number of days to the number of years by dividing by 365.25.

To save time, you won't perform all these steps manually. You'll run the whole workflow as a single geoprocessing model:

1. In the Catalog pane, expand Toolboxes and Forest_Disturbance_Analysis.tbx. Double-click Calculate Recovery Duration.

   

   The Geoprocessing pane appears, showing the model as a tool.

2. For Input Change Analysis Raster, choose WestCascade_change_analysis.crf. Turn off Use the filtered records.
3. For Output Recovery Raster, delete all existing text and type Recovery_years.crf.

   Note:

   If you see a warning symbol next to Output Recovery Raster, it's because the output location doesn't exist. To ensure the correct output location is used, make sure you delete all text before typing the output name.

   

4. Click Run.

   After a few moments, the output raster is created and added to the map. It is a single band raster where the pixel values are the number of years for the recovery.

   

5. In the Contents pane, right-click Recovery_years.crf and choose Zoom To Layer. Turn off Forest management boundaries and Smoother_Disturbance.crf.

   You can now see the layer on the map.

   

   Many clear-cut areas are dark pink, meaning they took 11 to 15 years or more to recover. While a disturbance tends to happen quickly, it can take a relatively long time for the forest to grow back.

6. Close the Geoprocessing pane.
7. Save the project.

Using a change analysis raster, you've mapped disturbances and measured the length it takes a forest to recover. Next, you'll analyze how the forest has changed by classifying the change analysis raster.

Classify forest change

To get a more complete picture of how the West Cascades forest changed over the years, you'll create a forest classification map for each time slice. The classification process will assign one the following three classes to every pixel:

• Healthy forest
• Disturbance
• Recovery

Classifying an image time series entails classifying every pixel in every time slice. Instead of classifying the original image time series, WestCascade.crf, you'll classify the derived change analysis raster, WestCascade_change_analysis.crf, that you generated earlier in the tutorial. The classification tool will use the information stored in all the bands of the change analysis raster, including Fitted_Value, Slope, Intercept, and Change_Magnitude, to identify the current status of the forest. The workflow for classifying a time series is similar to classifying a single image, but each training sample needs to be created on a specific time slice and labeled with the date of the time slice.

Note:

To learn more about classifying a single image, see the tutorial Calculate impervious surfaces from spectral imagery.

You'll perform the following workflow:

• Load the classification scheme.
• Collect training samples using the time-enabled Training Samples Manager tool.
• Choose a classifier and train it on the training samples. In this tutorial, you'll use a random trees classifier.
• Classify the time series using the trained classifier.
• Create a chart to show a summary of your classification results.

Examine the change analysis raster

To better understand the classification process, you'll examine the bands of the change analysis raster with RGB symbology. First, you'll set up a new map.

1. In the Catalog pane, in the Maps section, right-click the Classification map and choose Open.

   A map opens with only the basemap and WestCascade.crf. You'll also add the change analysis raster.

2. In the Catalog pane, expand Folders, Forest_Disturbance_Analysis, and commondata. Right-click WestCascade_change_analysis.crf and choose Add To Current Map.

   

   By default, the first time slice (1984-06-30) is displayed. You'll change the time slice and the symbology.

3. In the Contents pane, ensure that WestCascade_change_analysis.crf is selected.
4. On the ribbon, click the Multidimensional tab. For StdTime, choose the 1987-06-30 time slice.
5. In the Contents pane, click the WestCascade_change_analysis.crf symbol.

   The Symbology pane appears.

6. On the Symbology tab, set the following parameters:
   • For Primary Symbology, choose RGB.
   • For Red, choose Band_1_Slope.
   • For Green, choose Band_1_Fitted_Value.
   • For Blue, choose Band_1_Change_Magnitude.

   

   The map updates. The RGB symbology combines the three bands as a single red, green, and blue composite image. This symbology is a useful way to identify how different Slope, Fitted_Value and Change_Magnitude values might indicate different forest statuses.

7. On the ribbon, click the Map tab. Click Bookmarks and choose RGB details.

   

   Based on the values of the three bands, the color of different areas varies from bright green to bright pink. The following list explains what each color means:

   • Healthy mature forest—light green: Fitted_Value is high and Slope is more or less flat (close to 0). The area is covered with stable healthy forest.
   • Healthy mature forest but soon to be harvested—bright green: Fitted_Value is high and Slope is strongly negative. A patch of healthy forest that will be harvested in the next few time slices.
   • Disturbance—dark green: Fitted_Value is medium to very low and Slope is strongly negative. A disturbed area where the trees are in the process of being cut.
   • Disturbance, but soon to recover—bright pink: Fitted_Value is very low and Slope is strongly positive. A disturbed area that will be replanted and start recover in the next time slices.
   • Recovery—light pink: Fitted_Value is medium and Slope is positive. An area currently in recovery, with young trees growing into mature ones.
8. Close the Symbology pane.
9. In the Contents pane, right-click WestCascade_change_analysis.crf and choose Zoom To Layer.

   The classification process will use the information stored in Fitted_Value, Slope, and other bands to decide how to classify each pixel.

Load the classification schema

Now that you better understand how the classification process will use the band information, you'll start the classification workflow by loading the classification schema.

Note:

This tutorial provides a classification scheme file that was prepared for you. To create your own scheme file, refer to the Training Samples Manager documentation.

1. In the Contents pane, ensure that WestCascade_change_analysis.crf is selected.
2. On the ribbon, click the Imagery tab. In the Image Classification group, click Classification Tools and choose Training Samples Manager.

   

   The Image Classification pane appears, displaying a default schema. You'll load the tutorial-specific schema.

3. In the Image Classification pane, click the Classification Schema button.

   

4. In the Load Schema window, double-click Folders, Forest_Disturbance_Analysis, commondata, and userdata. Click WestCascadeForestTypes.ecs.

   

5. Click OK.

   The schema appears in the Image Classification pane. It lists the three target classes, which match the three forest types.

   

Collect training samples

Next, you'll collect training samples for your image classification. First, you'll change the symbology of the WestCascade_change_analysis.crf layer to make it easier to identify good sample locations.

1. In the Contents pane, right-click WestCascade_change_analysis.crf and choose Symbology.
2. In the Symbology pane, set the following parameters:
   • For Primary Symbology, choose Stretch.
   • For Band, choose Band_1_Fitted_Value.
   • For Color Scheme, choose the Greens (Continuous) color ramp.

   

   The map updates to show the NBR fitted values. The lowest values are white, while the highest are bright green.

   You should collect samples illustrating all three target classes, with examples in different time slices. The first sample you'll create will be for healthy forest in the time slice that is currently active (1987-06-30).

3. Close the Symbology pane. In the Image Classification pane, click the Healthy forest class to select it.
4. Click the Polygon button.

   

5. On the map, zoom to a patch of healthy forest (bright green).
6. Click the map to sketch polygon vertices inside the healthy forest. Double-click the last point to complete the polygon.

   

   Note:

   To better identify the sample locations, you can also look at the same time slice of the original imagery WestCascade.crf as a supplement to the fitted value view.

   In the Image Classification pane, the training sample is added to the list of samples.

   

   Next, you'll create a Disturbance training sample in the 1992-06-30 time slice.

7. On the ribbon, click the Multidimensional tab. For StdTime, choose the 1992-06-30 time slice.
8. In the Image Classification pane, click the Disturbance class to select it. Click the Polygon button.
9. On the map, identify a patch to disturbed forest (white) and sketch a polygon.

   Note:

   If you need to pan to a different location showing a disturbed forest area, press the C keycc and pan and zoom with the mouse.

   

   The second training sample is added to the list.

   In a real-life workflow, you would continue creating training samples from different time slices to accumulate good examples for all three classes. However, to save time in this tutorial, you'll use a set of training samples already prepared for you.

10. In the Image Classification pane, click the Load Training Samples button.

    

11. In the Training Samples window, double-click Databases and Forest_Disturbance_Analysis.gdb. Choose WestCascade_training_samples.

    

12. Click OK.

    The training samples from the file are added to the pane.

    

    The list contains about 80 samples, each with a class name and a time slice date.

13. Double-click the row button next to any training sample in the list.

    

    On the map, the corresponding polygon appears, displayed in the correct time slice.

14. Close the Image Classification pane. In the warning window asking if you want to save the modification, choose No.

    You don't need to make or save any modifications to the training samples.

15. In the Contents pane, right-click WestCascade_change_analysis.crf and choose Zoom To Layer.

    Note:

    These training samples were collected from a change analysis raster generated with exactly the same parameters as yours. It would not be applicable if you used a different change analysis raster, such as one generated with a different value for the Snapping date parameter.

Perform classification

Next, you'll train a classifier using the training samples. You'll apply it to the time series to classify every pixel of every time slice. First, you'll use the Train Random Tree Classifier tool to do the training.

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

   

   The Geoprocessing pane appears.

2. In the search box, type Train Random Trees Classifier. In the list of search results, click Train Random Trees Classifier.

   

3. For Input Raster, choose WestCascade_change_analysis.crf. Turn off Use the filtered records.
4. For Input Training Sample File, click the Browse button.

   

5. In the Input Training Sample File window, click Databases and double-click Forest_Disturbance_Analysis.gdb. Choose WestCascade_training_samples and click OK.
6. For Dimension Value Field, choose StdTime.
7. For Output Classifier Definition File, delete the text and type WestCascade_trained_classifier.ecd.

   

8. Click Run.

   The tool generates an Esri classification definition (.ecd) file, which contains statistics and classification information for each class. Next, you'll use the trained classifier to classify the entire time series with the Classify Raster tool.

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

   

10. Search for and open the Classify Raster tool.
11. For Input Raster, choose WestCascade_change_analysis.crf. Turn off Use the filtered records.
12. For Input Classifier Definition File, click the Browse button. Expand Folders, Forest_Disturbance_Analysis, and commondata.
13. Choose WestCascade_trained_classifier.ecd and click OK.
14. For Output Classified Raster, type WestCascade_classification.crf.

    

15. Click Run.

    Using the information in the trained classifier, the tool classifies every pixel of every time slice from the change analysis raster into one of the three target classes. The multidimensional CRF raster is added to the map. It is a time series, where each time slice is made of a single band raster containing classification values. By default, the 1984-06-30 time slice is displayed.

    The legend in the Contents pane lists the three classes:

    • Areas in dark green are classified as Healthy forest.
    • Areas in bright pink are classified as Disturbance.
    • Areas in light green are classified as Recovery.

    

16. On the ribbon, click the Multidimensional tab. In the Current Display Slice group, click Play Slices Along StdTime to animate the time series.

    As the time series animates, patches of healthy forest become disturbance areas and then recover.

17. When you are done watching the animation, click Play Slices Along StdTime to stop it.

Summarize the changes

Next, you'll compute summary statistics for the three classes in each time slice using the Summarize Categorical Raster tool. This tool will count the number of pixels assigned to each class.

1. In the Geoprocessing pane, click the Back button.
2. Search for and open the Summarize Categorical Raster tool.
3. For Input Categorical Raster, choose WestCascade_classification.crf. Turn off Use the filtered records.
4. For Output Summary Table, type Forest_class_table.

   

5. Click Run.

   The tool runs. In the Contents pane, under Standalone Tables, the output table appears. Next, you'll create a bar chart to display the information contained in the table, focusing on the Disturbance and Recovery classes.

6. Close the Geoprocessing pane.
7. In the Contents pane, right-click Forest_class_table, point to Create Chart, and choose Bar Chart.

   

   The Chart Properties pane and an empty chart view appear.

8. In the Chart Properties pane, for Category or Date, choose StdTime. For Aggregation, choose Sum.

   

9. For Numeric field(s), click Select. Check the check boxes for Disturbance and Recovery and click Apply.

   

   A chart appears. You'll format it and add labels to improve the display.

10. In the Chart Properties pane, click the Series tab.

    

11. In the list of fields, for Disturbance, click the symbol and choose Peony Pink.

    

12. For Recovery, click the symbol and choose Light Apple.

    

13. In the Chart Properties pane, click the General tab. Set the following parameters:
    • For Chart title, type Forest Classification (1984-2020).
    • For X axis title, type Year.
    • For Y axis title, type Number of pixels.

    

    The chart updates.

    

    This chart shows the trends of forest dynamics over a span of 35 years. It gives a sense of the balance between the disturbance and recovery of the forest ecosystem. An evident trend is the complementary nature of the forest dynamics; when forest recovery is high, forest disturbance is low, and vice versa. For example, in the 10-year interval between 1992 and 2002, the graph shows an inverse relationship where forest recovery was high while forest disturbance was relatively lower. In the intervals 1985–1990 and 2014–2020, disturbance increased relative to recovery due to timber harvesting, as confirmed by the imagery for these years.

14. Save the project.

Forest cover change analysis is crucial for assessing forest health and biodiversity and to support forest management, production of wood products, monitoring potential illegal logging activities, and more. Meeting these needs requires monitoring large areas at frequent time intervals and for a reasonable cost. LandTrendr is an effective mapping tool to detect and monitor forest disturbances and recovery. It relies on a time series of Landsat imagery to automatically detect spatial changes in forest land cover.

In this tutorial, you used the Pixel Time Series Change Explorer to examine forest changes at the pixel level. You then used the Analyze Changes Using LandTrendr and Detect Change Using Change Analysis Raster tools to create maps that revealed patterns and trends in forest disturbance and recovery. Finally, you classified the change analysis raster into three classes, using a random trees classifier and you created a chart showing the trends in forest dynamics over the 35-year span.

You can find more tutorials such as this in the Introduction to Imagery & Remote Sensing tutorial collection.

You can find more tutorials in the tutorial gallery.

Acknowledgements

• WestCascade.crf: Analysis-ready Landsat imagery from the United States Geological Survey (USGS), downloaded via EarthExplorer.
• Forest management boundaries: Derived from the USA Protected Areas - Manager Type layer in ArcGIS Living Atlas of the World, originally from the PAD-US dataset from USGS. See PAD-US Manager Name for a similar dataset.
• Kennedy, R. E., Z. Yang, and W. B. Cohen . "Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1. LandTrendr — Temporal segmentation algorithms," Remote Sensing of Environment, 114, no. 12 (2010): 2897-2910. https://doi.org/10.1016/j.rse.2010.07.008.
• World Topographic Map sources: Esri, TomTom, Garmin, FAO, NOAA, USGS, OpenStreetMap contributors, and the GIS User Community
• World Hillshade layer sources: Esri, Vantor, Airbus DS, USGS, NGA, NASA, CGIAR, N Robinson, NCEAS, NLS, OS, NMA, Geodatastyrelsen, Rijkswaterstaat, GSA, Geoland, FEMA, Intermap, and the GIS user community