Explore imagery - Spectral resolution

An important characteristic of imagery data is its resolution. There are four types of resolution: spatial, temporal, spectral, and radiometric, as illustrated in our growing tutorial series Explore imagery resolution. In this tutorial, you'll learn about spectral resolution.

You'll become familiar with several concepts such as spectral resolution, spectral bands, multispectral imagery, band combination, and spectral signatures. You'll examine and manipulate satellite imagery of different spectral resolutions in ArcGIS Pro, with a focus on the region of Brandenburg, Germany. Finally, you'll learn how to change the number of spectral bands in your imagery, a skill useful to prepare for further analysis.

This tutorial was last tested on January 9, 2026, using ArcGIS Pro 3.6. 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)

Visualize a multispectral satellite image

You'll set up an ArcGIS Pro project, learn about a few essential concepts related to spectral resolution, and start exploring satellite imagery in the Brandenburg region of Germany.

Set up the project

First, you'll download a project that contains all the data for this tutorial, and you'll open it in ArcGIS Pro.

1. Download the Brandenburg_spectral_resolution.zip file and locate the downloaded file on your computer.

   Note:

   Most web browsers download files to your computer's Downloads folder by default.

2. Right-click the Brandenburg_spectral_resolution.zip file and extract it to a location on your computer, such as a folder on your C: drive.

   

3. Open the extracted Brandenburg_spectral_resolution folder. Double-click Brandenburg_spectral_resolution.aprx to open the project in ArcGIS Pro.

   

4. If prompted, sign in with your ArcGIS account.

   Note:

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

   The project opens.

   

   It contains several maps that you'll use during the tutorial. The first map, named Spectral bands, is currently displayed. It is centered on an area south of Berlin in the State of Brandenburg, Germany. For now, it contains only the default World Topographic Map basemap.

Learn about the EM spectrum and spectral resolution

Before you start exploring satellite imagery for the Brandenburg area, you'll learn about some concepts essential to understanding spectral resolution.

The electromagnetic spectrum

The electromagnetic (EM) spectrum is made up of a range of wavelengths. Human beings can see in the visible light range because our eyes have receptors that are sensitive to red, green, and blue. Electro-optical sensors, such as satellite, aerial, and drone sensors, can capture the visible light range, as well as ranges that the human eye can't see, for instance near infrared and shortwave infrared.

This diagram represents the EM spectrum. The human eye can see only the visible light range, whereas imagery satellite sensors can capture a larger range that may include visible light, near infrared, shortwave infrared, and more.

Spectral bands

Imagery sensors capture portions of these wavelengths, named spectral bands. Examples of bands are Blue, Green, Red, Near infrared, and Shortwave infrared. Each spectral band is stored as an individual raster (or grid of pixels). Put together, the bands form a complete image. The following diagram shows three spectral bands from the EM spectrum being captured and stored as three individual rasters.

Spectral resolution

Spectral resolution refers to a sensor's ability to distinguish between different wavelengths. Sensors that capture more spectral bands produce images that have a higher spectral resolution. Usually, when a sensor captures more bands, these bands are also narrower, containing more precise information. There are three main types of images based on their number of bands, listed from lower to higher spectral resolution:

• RGB (standing for Red, Green, Blue) – 3 bands
• Multispectral – usually between 4 and 13 bands
• Hyperspectral – hundreds or even thousands of very narrow bands

The following diagram shows a stylized representation of (1) RGB, (2) multispectral, and (3) hyperspectral images.

For instance, a handheld camera usually captures RGB bands to produce traditional color photos. Many earth observation imagery sensors are multispectral. Some advanced earth observation imagery sensors are hyperspectral.

Images with higher spectral resolution contain more information and can provide many interesting insights into the landscape observed. For instance, they can be used to distinguish between different tree species or to detect whether some of these trees are healthy or diseased.

In the rest of this tutorial, you'll focus on multispectral imagery. You'll explore images from three satellite sensors: Sentinel-2, PlanetScope, and Landsat 8. You'll learn about their spectral resolution and how to work with their spectral bands.

Note:

While hyperspectral imagery is beyond the scope of this tutorial, you can learn more the topic on the Hyperspectral imagery in ArcGIS documentation page, or see an example of hyperspectral imagery preparation and analysis in the Map Oaks Using AVIRIS Hyperspectral Imagery article.

Inspect a Sentinel-2 multispectral image

You will now start exploring a Sentinel-2 image. Sentinel-2 images have 13 spectral bands. Some are visible to the human eye (Blue, Green, and Red) and some are not, such as Coastal aerosol, Red edge, Near infrared, and Shortwave infrared. The following diagram shows where the 13 bands fall on the EM spectrum, between 400 and 2,400 nanometers in wavelength.

Here is the list of these bands, including their number and name:

• Band 1—Coastal aerosol
• Band 2—Blue
• Band 3—Green
• Band 4—Red
• Band 5—Vegetation red edge
• Band 6—Vegetation red edge
• Band 7—Vegetation red edge
• Band 8—Near infrared
• Band 8A—Near infrared narrow
• Band 9—Water vapor
• Band 10—Shortwave infrared - Cirrus, used to detect cirrus clouds
• Band 11—Shortwave infrared
• Band 12—Shortwave infrared

Note:

Sentinel-2 is a satellite mission from the European Space Agency. It was launched in 2015 and produces imagery with a spectral resolution of 13 spectral bands, several of which have a spatial resolution of 10 meters. The images cover nearly the entire landmass of Earth, and every place is captured with a temporal resolution of at least every five days. Sentinel-2 images are freely available and can be downloaded through the Copernicus Data Space Ecosystem.

You’ll look at the Sentinel-2 image provided with the project and identify its spectral bands.

1. On the ribbon, click the View tab. In the Windows group, click Catalog Pane.

   

2. In the Catalog pane, expand Folders, Brandenburg_spectral_resolution, and Imagery.

   

   The Imagery folder contains three images. For now, you will work with Sentinel_2_2024_08_13.tif.

3. Expand Sentinel_2_2024_08_13.tif and observe the spectral bands it contains.

   

   For this image, the bands are named B + a band number. Upon inspection, notice that this image contains all the Sentinel-2 bands, except Band 10 - Cirrus, which is only used for cloud detection. So, 12 bands are present.

Visualize individual spectral bands

You'll visualize two of the Sentinel-2 spectral bands and compare them. You'll also learn about the concept of reflectance. First, you'll add the Red (B4) band to the map.

1. Right-click B4 and choose Add To Current Map.

   

   The Red band appears on the map, displayed in gray tones, ranging from white to black.

   

   Note:

   The original Sentinel-2 image is much larger, but it was clipped to a smaller extent for the needs of this tutorial.

   To better understand what you are seeing, it is useful to learn about reflectance. When the sunlight in a specific wavelength range reaches the surface of the earth, some of it is absorbed and some of it is reflected back into space. The imagery sensor captures the portion of the light that is reflected back, as illustrated in the following graphics.

   

   Graphics showing (1) the sunlight, (2) the light absorbed, (3) the light reflected and captured by the sensor.

   The amount of reflected light in each spectral band varies based on the physical and chemical properties of the material on the ground (vegetation, soil, rock, water, and so on). For example, the chlorophyll in vegetation, responsible for photosynthesis, strongly absorbs Blue and Red wavelengths, but reflects Green wavelengths and even more strongly Near infrared. The sensor captures the reflectance values for the light reflected in different bands throughout the entire area imaged.

2. Observe the Red band displaying on your map.

   In that band, areas that strongly reflect light in the red wavelength range have the highest reflectance values and appear in white and light grays. This is the case of built-up areas, such as the H-shaped airport in the upper center part of the image. Areas that absorb most of the light in the red wavelength range have the lowest reflectance values and appear in black and dark grays. This is the case of forests and water bodies. You will now add the Near infrared (B8) band to the map.

3. In the Catalog pane, right-click B8 and choose Add To Current Map.

   The Near infrared band appears on the map, also displayed in gray tones. However, this time the image looks quite different. The features that have the highest reflectance values are the grass and agricultural fields (white tones). The features with the lowest reflectance values are the water bodies (black tones).

   

   You'll compare both bands using the Swipe tool.

4. In the Contents pane, confirm that Sentinel_2_2024_08_13.tif_B8 is selected.

   

5. On the ribbon, click the Raster Layer tab. In the Compare group, click Swipe.

   

6. On the map, drag from top to bottom to peel off the Sentinel_2_2024_08_13.tif_B8 layer and reveal the Sentinel_2_2024_08_13.tif_B4 layer under it. Observe the differences between the two layers.

   

   The light reflects off materials in different amounts based on the wavelength range. As a result, spectral bands highlight different features and phenomena beyond what the human eye can see.

7. When you are finished exploring, on the ribbon, click the Map tab. In the Navigate group, click the Explore button to exit the swipe mode.

   

Visualize image composites

The full power of multispectral imagery comes when you add the entire image with all its spectral bands at once to the map. You'll do that next and learn how to form image composites by combining different spectral bands.

1. In the Catalog pane, right-click Sentinel_2_2024_08_13.tif and choose Add To Current Map.

   

   Note:

   There can be many different imagery file formats. In this case, the image is made of a single TIFF file that encapsulates all 12 spectral bands. Sometimes, every spectral band is provided as a separate TIFF file, and a metadata file—named for instance MTL.txt, MTL.xml, or MTD_MSILxx.xml—contains information to display the image properly. For such formats, you’ll add the image to the map by right-clicking the metadata file and choosing Add To Current Map.

   The image appears in the Contents pane and on the map.

   

   All the spectral bands are available. However, because of the human eye's limitations (trichromacy), it is not possible to visualize all of them at the same time. Three of the bands are combined to create a color image, named image composite.

   Note:

   Learn more about the way the human eye perceives light in the Bending light and blending light article.

   Images are displayed using Red, Green, and Blue (RGB) channels, and you can choose any set of three spectral bands to display them through these channels. The following diagram illustrates how this works:

   

   How an image composite is created: (1) choose any three bands from the multispectral image to display; (2) assign them to the Red, Green, and Blue channels in the order desired; (3) obtain an image composite.

   Currently, the image on your map displays by default the Red (B4), Green (B3), and Blue (B2) bands through the Red, Green, and Blue channels, respectively. This band combination, named Natural color, approximates how the landscape would appear to the human eye. You'll experiment with some other band combinations, but first, you'll adjust the general image appearance: you’ll increase its brightness, since it is currently somewhat dark.

2. In the Contents pane, confirm that the Sentinel_2_2024_08_13.tif image is selected.

   

   The layer's legend indicates the current band combination.

3. On the ribbon, click the Raster Layer tab. In the Enhancement group, for Layer Brightness, type 20 and press Enter.

   

   Note:

   The other Enhancement settings, Layer Contrast and Layer Gamma, can also be used to adjust the image appearance. Learn about these imagery rendering options and many others on the Imagery appearance and Change the symbology of imagery pages. You'll learn more about imagery rendering options in an upcoming Radiometric Resolution tutorial.

   The image updates to overall brighter tones.

   

   Next, you'll change the band combination.

4. In the Contents pane, right-click the Sentinel_2_2024_08_13.tif image and choose Symbology.

   

5. In the Symbology pane, for Primary symbology, confirm that RGB is selected. Confirm that the Red, Green, and Blue channels are currently displaying the spectral bands B4 (Red), B3 (Green), and B2 (Blue), as expected.

   

   Next, you'll display the Color infrared combination, composed of the B8 (Near infrared), B4 (Red), and B3 (Green) bands.

6. For Red, expand the drop-down list. In the list of 12 Sentinel-2 reflectance bands, pick the B8 band.

   

7. For Green, choose B4. For Blue, choose B3.

   

   On the map, the image updates to show vegetation in bright red tones, built-up or bare earth areas in bluish or brownish tones, and water in dark navy blue.

   

   When any bands beyond Red, Green, and Blue are displayed, the landscape can appear in unusual tones, referred to as false colors. This is a powerful technique for the human eye to visualize wavelength ranges that are usually invisible. The Color infrared band combination is particularly useful to highlight vegetation and monitor its health. Next, you'll try another combination, usually named Agriculture, composed of B11 (Shortwave infrared 1), B8 (Near infrared), and B2 (Blue).

8. In the Symbology pane, for Red, choose B11; for Green, choose B8; and for Blue, choose B2.

   

   This band combination is one of the most versatile and is excellent for distinguishing many different land cover types clearly: vegetation appears in bright green, water in dark blue, built-up in pink tones, and bare earth in light orange.

   

   You'll switch back the Natural color combination.

9. In the Symbology pane, form the B4, B3, B2 combination.

   Some features are hard to distinguish in Natural color but clearly differentiated with the Agriculture combination, such as the lake and the vegetation surrounding it.

   

   A lake and its surroundings shown in the Natural color (left) and Agriculture (right) band combinations.

   You experimented with the Natural color, Color infrared, and Agriculture band combinations, but there are many more possible combinations, such as Geology (B12, B11, B2) to highlight geological formations, or Bathymetric (B4, B3, B1) used for coastal studies.

So far, you learned about several important concepts: the EM spectrum, spectral bands, spectral resolution, different types of imagery (RGB, multispectral, and hyperspectral), image composites, and common band combinations, some of which in false colors. You also started exploring Sentinel-2 multispectral imagery, visualizing it, and experimenting with several band combinations to highlight different features of the landscape.

Learn about spectral signatures

You'll now explore spectral band variations interactively, and learn about spectral signatures and spectral profile charts. Then, you'll compare two images with different spectral resolutions.

Explore spectral band variation interactively

To become familiar with the variation in reflectance values across all spectral bands, you'll use the Image Information tool, which provides interactive information at the pixel level.

1. In the Contents pane, confirm that the Sentinel_2_2024_08_13.tif image is selected and that it is displaying with the Natural color band combination (B4, B3, B2).

   

2. On the ribbon, on the Imagery tab, in the Tools group, click Image Information.

   

   The Image Information pane appears.

3. On the map, point to a bright green agricultural field, full of healthy vegetation.

   

   In the Image Information pane, a graph appears showing the reflectance value for every band at that particular pixel location. For this agricultural field, the Near infrared (NIR) band has a very high value and the Red band (symbolized in red) a very low value, which is typical of plants rich in chlorophyll, as explained earlier. This type of graph is named a spectral profile chart.

   

4. On the map, point to various land covers, such as bare earth (beige or light brown), forest (dark green), and built-up (white or light gray). Observe how the reflectance values change interactively.

   For instance, for bare earth, the Red band reflectance value is comparatively higher and the NIR band reflectance value lower than what you saw for the agricultural field. The two Shortwave infrared (SWIR) bands also have higher values.

   

5. Look at several pixels representing agricultural fields throughout the image. In contrast, look at several pixels representing bare earth.

   You will notice that each land cover type seems to always produce a similar recognizable curve on the chart. This is because different materials tend to have unique patterns of reflectance values across different wavelengths of the EM spectrum. These recognizable patterns are referred to as spectral signatures, and could be described as the fingerprint of a material based on how it interacts with light. Spectral signatures make it possible to distinguish different land cover types (grass field, forest, water, built-up, and so on). Similarly, diseased crops or sparse forest would have a typical spectral signature that would distinguish them from healthy crops or dense forest. Most imagery analysis techniques take advantage of these predictable patterns to detect valuable information about the landscape.

Examine a spectral profile chart

Next, you'll compare the spectral signatures for different land cover types in more detail, using a spectral profile chart that was prepared for you. You'll switch to the second map of the project.

1. Above the map, click the Spectral profiles tab.

   

   This map contains the same Sentinel-2 image you have been exploring, but a spectral profile chart has been added to it.

2. In the Contents pane, under Charts, right-click Spectral Profile – Sentinel-2 and choose Open.

   

   The spectral profile pane appears, displaying the Spectral Profile – Sentinel-2 graph.

   

   This spectral profile chart contains the reflectance value curves for five pixels on the map: Forest, Water, Built-up, Grass field, and Bare earth. The x-axis lists the Sentinel-2 bands and the y-axis indicates the amount of reflectance captured by the sensor.

   You can see the corresponding points on the map.

   

   Here again, each land cover type has its own specific curve on the graph, and you can now observe it in more detail. You can recognize the Grass field with reflectance values that are very low for the Red band ( B4), very high for the NIR band ( B8), and relatively low for the SWIR bands (B11 and B12). In contrast, the Bare earth increases steadily in reflectance values from B1 to B12, and the Water has very low reflectance values for all bands.

   Note:

   Because Red edge and NIR wavelengths are so important for vegetation monitoring, Sentinel-2 images have several bands in these ranges. They are: B5, B6, B7 (Red edge bands), B8 (main NIR band), and 8A (narrow NIR band). These bands provide a wealth of information that can be used in various specialized applications in agriculture and environmental studies.

   Next, you'll add more points to the graph. First, you'll declutter the graph.

3. In the Chart Properties pane, under Spectral Profiles, uncheck Forest, Water, and Built-up.

   

   On the graph, only the curves for Grass field and Bare earth remain.

4. In the Chart Properties pane, under Define an area of interest, click the Point button.

   

   Your pointer changes to a crosshair shape.

5. On the map, click a grass or agriculture field of your choice (vivid green areas).

   

   A point is added to the map, and the spectral profile curve for that specific pixel is added to the graph.

6. Similarly, add a new bare earth point (beige areas).

   

   On the graph, the curves for your new grass field and bare earth points should be quite similar to the preexisting ones, confirming that the grass field and bare earth land cover types each have their own typical spectral signature.

   

   Note:

   The colors are assigned at random and may differ in your graph.

7. Optionally, add new points for other land cover types (Forest, Water, Built-up) and compare them to the original curves in the graph.
8. When you are done, close the spectral profile graph pane.

   

   Note:

   If you want to create a spectral profile chart from scratch for your own imagery, in the Contents pane, right-click the image, click Create Chart, and choose Spectral Profile. Learn more about Spectral profile options.

   To learn about how imagery analysis techniques take advantage of spectral signatures to detect valuable information about the landscape, try the following tutorials:

   • Assess hail damage in cornfields with satellite imagery (spectral indices and change detection)
   • Assess burn scars with satellite imagery (band combinations and spectral indices)
   • Calculate impervious surfaces from spectral imagery (supervised classification)
   • Classify land cover to measure shrinking lakes (unsupervised classification)
   • Extract high-resolution land cover with GeoAI (deep learning)

Compare images with different spectral resolutions

Next, you'll compare two satellite images with different spectral resolutions. You'll switch to the third map of the project.

1. Above the map, click the Compare images tab.

   

   This map contains the Sentinel-2 image you have been exploring, along with a new image, PlanetScope_2024_08_13.tif. This is an Analysis-Ready PlanetScope satellite image produced by the Earth-imaging company Planet Labs. It was captured on August 13, 2024, the same day as the Sentinel-2 image, and is clipped to the same extent.

   

   Note:

   Analysis-Ready PlanetScope images are produced by Planet Labs. PlanetScope is a collection of hundreds of satellites that have been deployed from 2014 onward and produces imagery with a 3 meter per pixel resolution. The images cover nearly the entire landmass of Earth, and each location is captured almost daily.

   You'll check what spectral bands are present in that image.

2. In the Contents pane, right-click the PlanetScope_2024_08_13.tif image and choose Symbology.
3. In the Symbology pane, for Red, expand the drop-down list.

   There are four spectral bands available: Blue, Green, Red, and NIR.

   

   Compared to the Sentinel-2 image, which has 12 bands, the PlanetScope image has a lower spectral resolution. For instance, it doesn't include any SWIR bands. Currently, it is displayed using the Natural color band combination (Red, Green, Blue), like the Sentinel-2 image.

4. In the Contents pane, turn the PlanetScope_2024_08_13.tif image on and off to compare the two images.

   

   In the Natural color band combination, the two images look very similar. However, since the PlanetScope image has only four bands, fewer band combinations are possible than with the Sentinel-2 image. Besides Natural color (Red, Green, Blue), the main other band combination is Color infrared (NIR, Red, Green). You'll switch to it.

5. In the Symbology pane, set the Red channel to NIR, the Green channel to Red, and the Blue channel to Green.

   On the map, the PlanetScope image updates to Color infrared. As you learned earlier, this is a useful combination to study vegetation health.

   

   Other band combinations, such as Agriculture, Geology, or Bathymetric, are not possible with only the Blue, Green, Red, and NIR bands. You'll switch back to Natural color.

6. In the Symbology pane, switch back to Natural color (Red, Green, Blue).

   

   You'll now compare the spectral profile charts for the two images.

7. In the Contents pane, under PlanetScope_2024_08_13.tif, right-click Spectral Profile – PlanetScope and choose Open.

   

8. Open the Spectral Profile - Sentinel-2 chart.

   You'll display the graphs side by side.

9. Drag the Spectral Profile - Sentinel-2 tab to the right-side docking target.

   Note:

   As you drag the pane—represented by a blue shadow—docking targets appear in the center of the table view. Each target represents an area where the pane can be positioned.

   

   The two graphs now display side by side.

10. If necessary, resize the graph panes, so that they are roughly of equal width.

    

11. Observe how the two graphs differ from each other.

    

    For easier comparison, remember that the PlanetScope bands correspond to the following four Sentinel-2 bands:

    • Blue—B2
    • Green—B3
    • Red—B4
    • NIR—B8

    Note:

    These two images were processed using slightly different methods, which accounts for the difference in values on the y-axis. However, their relative values are still useful for comparison.

    Both graphs display reflectance value curves for the same five pixels on the map. However, because of its lower spectral resolution, the PlanetScope image displays simpler curves that contain less information than the Sentinel-2 image. This means that it won't be able to support as many analysis workflows. For instance, workflows that rely on having SWIR or Red edge bands won't be possible.

    However, there are pros and cons to using higher versus lower spectral resolution imagery, as summarized in the following table:

    Higher spectral resolution imagery	Lower spectral resolution imagery
    Supports more band combinations Supports more sophisticated analysis Allows for more subtle differentiation of materials, vegetation species, and so on	Might take significantly less storage space Might have a higher spatial resolution (that is, show more details on the ground) Might have a higher temporal resolution (that is, a higher revisit cycle)

    For instance, PlanetScope has a higher spatial resolution than Sentinel-2: every pixel represents a square of 3 by 3 meters on the ground versus 10 by 10 meters. It also has a higher temporal resolution: every location is revisited almost daily versus about every five days.

    Choosing imagery with higher or lower spectral resolution depends on the intended use. It is also possible to use both types in conjunction with each other. For instance, you could perform a more sophisticated analysis every few months with higher spectral resolution imagery, and more frequent, quicker checks with lower spectral resolution imagery.

    Note:

    There is a great variety of sensors and each one has its own spectral resolution. To see the many satellite sensors supported in ArcGIS Pro and learn about the spectral bands they provide, refer to the Satellite sensor raster types documentation page. See a similar list for multispectral drone cameras. Many different aerial imagery types are also supported.

12. Close the two spectral profile charts.

In this part of the tutorial, you learned about spectral signatures and spectral profile charts. You explored spectral band variations interactively, examined a spectral profile chart in detail, and compared two images with different spectral resolutions.

Change the spectral resolution of your imagery

When you receive a new multispectral image, you should know how to gather information about its bands. You should also know how to transform its spectral resolution, that is, change its number of bands. You'll learn how to do that working with a satellite Landsat-8 image.

Explore a Landsat 8 image

First, you'll become familiar with the Landsat 8 image. You'll switch to the fourth map of the project.

1. Above the map, click the Extract bands tab.

   

   The Extract bands map contains Landsat_8_2024_08_31.tif, a Landsat-8 image that was captured on August 31, 2024. It is clipped to the same extent as the earlier images, and it is currently displayed in Natural color.

   

   Note:

   Landsat 8 is a satellite mission from USGS and NASA launched in 2013. It produces multispectral imagery with 11 spectral bands, most of them with a 30-meter spatial resolution. The images cover nearly the entire landmass of Earth, and every place is captured every 16 days (or every 8 days if combined with Landsat 9 images). Landsat is the longest-running satellite imagery acquisition program, providing more than five decades of continuous earth observation data.

   Landsat images are freely available. Learn how to download your own Landsat imagery.

   Here is the list of spectral bands for Landsat 8:

   • Band 1—Coastal aerosol
   • Band 2—Blue
   • Band 3—Green
   • Band 4—Red
   • Band 5—Near infrared (NIR)
   • Band 6—Shortwave infrared (SWIR) 1
   • Band 7—Shortwave infrared (SWIR) 2
   • Band 8—Panchromatic (large band covering most of the visible light range)
   • Band 9—Cirrus (used to detect cirrus clouds)
   • Band 10—Thermal infrared 1 (measures surface temperature)
   • Band 11—Thermal infrared 2 (measures surface temperature)

   Note:

   Learn more about Landsat 8 bands.

   The following graphic shows where the Landsat 8 bands are located on the EM spectrum compared to the bands of Sentinel-2 and PlanetScope.

   

   While Sentinel-2 has overall more bands than Landsat 8, including several in the Red edge and Near infrared areas, one of Landsat 8's strengths is that it has Thermal infrared bands that can measure surface temperatures.

   When you receive or download imagery, it is important to gather information about its bands. You would usually achieve that by reading some documentation about the sensor, as well as inspecting your image within ArcGIS Pro. One of the ways of doing the latter is to look in the Layer Properties window.

2. Right-click Landsat_8_2024_08_31.tif and choose Properties.

   

3. In the Layer Properties window, click Source. Expand Raster Information and identify the Number of Bands line.

   

   This Landsat 8 image has 7 bands.

4. Expand Band Metadata.

   

   In this Landsat 8 image, the full 11 bands are not available, because only what is considered as the 7 core reflectance bands (sr_band1 to sr_band7) are included. That's Coastal aerosol, Blue, Green, Red, NIR, SWIR 1, and SWIR 2, respectively.

   Note:

   The Layer Properties window contains valuable information about the image. To learn more about it, refer to the Raster dataset properties page.

   You have now learned two ways of gathering information about the number of imagery bands: through the Symbology pane and the Layer Properties window.

5. Close the Layer Properties window.

Change the number of bands

There are cases in which you might want to pare down the number of bands of your multispectral imagery. For instance, if you know that you'll perform an analysis that needs only specific bands, eliminating the other bands can save disk space. This is especially true if you aim to analyze a large number of multispectral images. The analysis tools you want to use might also be expecting only specific bands organized in a specific order.

Note:

For instance, when extracting information from imagery using deep learning, GeoAI pretrained models generally require input imagery similar to the data they were trained on. This is often 3-band imagery, Red, Green, and Blue, organized in that order. To learn more, see the Detect objects with a deep learning pretrained model tutorial.

Next, you'll learn how to make that change yourself. Starting from your 7-band Landsat image, you decide that you'll need only the following five bands: sr_band2 (Blue), sr_band3 (Green), sr_band4 (Red), sr_band5 (NIR), and sr_band7 (SWIR 1). You'll create an image that contains only these five bands using the Extract Bands raster function and the Export Raster tool.

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

   

2. In the Raster Functions pane, in the search box, type Extract Bands. Click the Extract Bands function.

   

3. Set the following Extract Bands parameters:
   • For Raster, choose Landsat_8_2024_08_31.tif.
   • For Method, choose Band Names.
   • For Combination, delete the current text.

   

4. Expand the Band drop-down list. Select sr_band2, sr_band3, sr_band4, sr_band5, and sr_band7.

   The Combination field is populated with these five band names.

   

5. Click Create new layer.

   A new layer, named Extract Bands_Landsat_8_2024_08_31.tif, appears in the Contents pane. You'll check to see whether the result is what you expected.

   Note:

   On the map, the new layer has a bluish tint, which is due to the default band combination. You'll remedy this issue later in the workflow.

6. In the Contents pane, under Extract Bands_Landsat_8_2024_08_31.tif, right-click the Red symbol.

   The list of bands available in the image appears.

   

   As expected, the five bands (sr_band2, sr_band3, sr_band4, sr_band5, and sr_band7) are listed.

   Layers created by raster functions are computed dynamically in memory. This makes the processing time very fast, but they are not saved on disk. In this case, you want to persist the resulting layer as a TIFF file on your computer. You'll do that with Export Raster.

7. Right-click Extract Bands_ Landsat_8_2024_08_31.tif, point to Data, and choose Export Raster.

   

8. In the Export Raster pane, for Output Raster Dataset, click the Browse button.

   

9. In the Output Location window, browse to Folders > Brandenburg_spectral_resolution > Imagery. For Name, type Landsat_8_2024_08_31_5bands.tif. Click Save.

   

10. In the Export Raster pane, accept the default values for all other parameters. Click Export.

    

    After a few moments, the new image is added to your map.

11. In the Contents pane, under Landsat_8_2024_08_31_5bands.tif, right-click the Red symbol and confirm that the expected five bands are listed.

    Currently, the bands are assigned to the RGB channels in the default increasing order:

    • Red channel—sr_band2 or Blue band
    • Green channel—sr_band3 or Green band
    • Blue channel—sr_band4 or Red band

    

    This order is not particularly helpful and results in the image having overall bluish tones. Instead, you'll form the Natural color band combination (sr_band4, sr_band3, sr_band2).

12. In the Contents pane, right-click the Red symbol and choose sr_band4.

    

13. Right-click the Blue symbol and choose sr_band2.

    

    The image updates to the Natural color combination. The bluish tones are gone, and the bare soil areas display in more natural brown and beige tones. Alternatively, you could also switch to other band combinations, such as Color infrared (sr_band5, sr_band4, sr_band3) or Agriculture (sr_band7, sr_band5, sr_band2), as needed to better explore the imagery.

    Note:

    Sometimes, you might acquire imagery that is delivered as a set of separate files, one for each spectral band. In that case, one option is to use the Composite Bands tool to gather these files into a single TIFF file composed of several bands, similar to the imagery you worked with in this tutorial.

Check your understanding

Optionally, try to answer the following questions about the key concepts you learned in this tutorial. Just use your own words. How would you explain these concepts to someone else? If you are not sure, go back up the tutorial to review the relevant explanations.

• What are spectral bands?
• What is spectral resolution?
• What is multispectral imagery? And what are some examples of satellite sensors that produce multispectral imagery?
• What are some examples of common band combinations? And what are the strengths of each one?
• What does a spectral profile chart represent?

Go further

Optionally, challenge yourself with more activities to continue to learn.

• Generate a spectral profile chart for the Landsat image. How does it compare to the other two charts you saw earlier in the tutorial?
• Create a map with the Sentinel-2 and Landsat images, and render both of them with the Agriculture band combination. Compare the two layers: can you spot some fields that were harvested between the two images? Can you spot some fields where the vegetation seems to have grown?
• Using the Sentinel-2 or Landsat image, create a spectral profile chart for sample points located in different water bodies in the area. Can you find a water body that has higher Green band values, indicating the presence of algae? And what do you think muddy water—that is, water mixed with soil—would do to the spectral profile?
• Check out some tutorials illustrating different types of multispectral imagery analysis:
  • Assess hail damage in cornfields with satellite imagery (spectral indices and change detection)
  • Assess burn scars with satellite imagery (band combinations and spectral indices)
  • Calculate impervious surfaces from spectral imagery (supervised classification)
  • Classify land cover to measure shrinking lakes (unsupervised classification)
  • Extract high-resolution land cover with GeoAI (deep learning)

In this tutorial, you became familiar with the concepts of spectral resolution, spectral bands, multispectral imagery, reflectance, image composites, band combinations, spectral signatures, and spectral profile charts. You visualized satellite imagery with various band combinations. You explored spectral profile charts and compared imagery of different spectral resolutions. Finally, you learned how to change the number of spectral bands in your imagery to be ready for more sophisticated analyses.

You can find more tutorials like this in our growing series Explore imagery resolution.

Acknowledgements

• Sentinel_2_08_13_2023.tif: Sentinel-2 satellite image, courtesy of the European Space Agency, downloaded from Copernicus Data Space Ecosystem.
• PlanetScope_08_13_2023.tif: Brandenburg Agriculture, an Analysis-ready PlanetScope satellite image, by Planet Labs, Inc., licensed under CC BY 4.0.
• Landsat_8_08_31_2023.tif: Landsat 8 satellite image, courtesy of the U.S. Geological Survey, downloaded from USGS Earth Explorer.
• World Topographic Map sources: Esri, TomTom, Garmin, FAO, NOAA, USGS, OpenStreetMap contributors, and the GIS User Community
• World Hillshade layer sources: Esri, Maxar, Airbus DS, USGS, NGA, NASA, CGIAR, N Robinson, NCEAS, NLS, OS, NMA, Geodatastyrelsen, Rijkswaterstaat, GSA, Geoland, FEMA, Intermap, and the GIS user community
• All Adobe Stock images are ©2025 Adobe Stock. All rights reserved.
• This tutorial was developed in collaboration with Planet Labs with special thanks to Matt Irons, Alana Belcon, and Ely Hienrich.