This example demonstrates how to prepare imagery from most multispectral sensors for use in FLAASH. It uses Landsat 8 imagery as a case study, but the general concepts apply to other sensors such as QuickBird, WorldView, IKONOS, and others.

These steps do not use a specific data file for illustration; they are general steps that you can use with your own imagery.

See the following tutorials for related topics:

  • Preprocessing AVIRIS Data (demonstrates the use of hyperspectral data in FLAASH)
  • Using ASTER Data with FLAASH (available in the ENVI Classic Help)

Start FLAASH

  1. Type the word "flaash" in the search field of the Toolbox.
  2. Double-click the FLAASH Atmospheric Correction tool name that appears. The Data Selection dialog appears.
  3. Select a file. The FLAASH - Rigorous Atmospheric Correction dialog appears, with the Main tab selected.

Set Time and Output Options

  1. Go to the Layer Manager, right-click on the original Landsat scene filename, and select View Metadata.
  2. Select the Time category on the left side of the Metadata Viewer and note the Acquisition Time.

  3. In the Main tab of the FLAASH dialog, use the Acquisition Time from the metadata to set the Acquisition Date/Time in the drop-down lists.

  4. In the Output Raster field, specify a filename and path.

Set Sensor and Geometric Options

  1. Select the Sensor tab, then select Multispectral from the Sensor Type drop-down list.
  2. Select the Geometric tab, then enter an estimated Ground Elevation (km) value for the center of the scene. You can use a tool such as Google Earth to locate the approximate elevation.

Set Water Vapor and Aerosol Options

The proper water and aerosol retrieval settings are less critical for multispectral sensors than they are for imaging spectrometers such as AVIRIS and EO-1 Hyperion. Most multispectral sensors do not have the spectral resolution to perform accurate water and aerosol retrieval. The following steps provide some general recommendations. If you leave the remaining fields at their default values, it would still yield a relatively accurate atmospheric correction, suitable for most image-processing applications. However, the following steps will teach you about these options in greater detail.

  1. Select the Model tab. The Atmospheric Model setting relates to the average water vapor amount for the scene. Since Landsat 8 does not have any water vapor bands and do you likely do not have water vapor information available for the scene, you can choose one of these model atmospheres. They are just generalizations based on approximate geographic location. Select the atmospheric model that is closest to the geographic location and season for your scene.
  2. Select the Water tab. For Water Absorption Wavelength, keep the default selection of Automatic Selection.
  3. Because most multispectral sensors do not have the appropriate bands to retrieve water vapor information, the minimum and maximum default values of 0.01 and 1 is sufficient for the Water Column Multiplier field.
  4. Select the Aerosol tab. The choice of Aerosol Model is not critical if value of the Initial Visibility field on the Misc tab is 40 km. If the visibility is lower than 40km, select one of the following options:
    • High-Visibility Rural: Represents aerosols in areas not strongly affected by urban or industrial sources. The particle sizes are a blend of two distributions, one large and one small, where dust-like aerosol makes up the majority of the course particles. For clear to very clear conditions (visibility) the vertical distribution of the aerosol extinction coefficient is exponential.
    • Low-Visibility Rural: Represents aerosols in areas not strongly affected by urban or industrial sources. The particle sizes are a blend of two distributions, one large and one small, where dust-like aerosol makes up the majority of the course particles. For hazy conditions (visibility) the vertical distribution of the aerosol extinction coefficient is assumed to be independent of height up to 1 km with a pronounced decrease above that height.
    • Maritime: Represents the boundary layer over oceans, or continents under a prevailing wind from the ocean. It is composed of two components, one from sea spray and another from rural continental aerosol (that omits the largest particles).
    • Urban: A mixture of 80% rural aerosol with 20% soot-like aerosols, appropriate for high-density urban/industrial areas.
    • Tropospheric: Applies to calm, clear (visibility greater than 40 km) conditions over land and consists of the small-particle component of the rural model.

  5. For Use Aerosol, keep the default selection of Automatic Selection.

  6. Select the Misc tab. The Initial Visibility field is set to 40 km, which indicates clear atmospheric conditions throughout the scene. You can lower this value if the scene is in a dense urban area with atmospheric pollution, or if weather reports indicate hazy conditions that day.
  7. ENVI determines the best bands to use based on your selection, along with a recommended upper channel reflectance value and reflectance ratio. The default values are typically accurate, but you can experiment with different values.

Create the Reflectance Image

  1. Click the OK button.
  2. When processing is complete, open the Data Manager and find reflectance file was just created.
  3. Right-click on the filename and select Load True Color.
  4. One way to validate that apparent reflectance results are accurate is to view a Spectral Profile. Right-click on the reflectance filename in the Layer Manager and select Profiles > Spectral.
  5. FLAASH scaled the reflectance values by 10,000. The reason for the scaling is to convert the reflectance data to integers, which takes up less disk space than floating-point data. If you prefer to have the reflectance data range from 0 to 1.0, you can use the Band Math tool to divide the reflectance pixel values by 10,000.

Areas of dark shadow or water may have negative reflectance values. This is just an artifact of pixels that represented low radiance values, and they do not model atmospheric conditions well.