DN2REFLECTANCE

Description

DN2REFLECTANCE converts to absolute reflectance pixel values with radiometric calibration information. Thermal channels are converted to thermal brightness temperature in degrees Kelvin. The supported input calibration types are radiance and reflectance. The input reflectance can be relative (still affected by the effects of the solar zenith angle) or absolute, with the solar elevation angle effects removed. The absolute reflectances are transferred as-is, only with pixel values scaling, if required.

Parameters

Name	Type	Length	Value range
Input: Input data to be converted *	Raster port	1 - 1024
Input: Input Solar Zenith Angles *	Raster port	1 - 1024
Output type	String	0 - 1	8U| 16S| 16U| 32R Default: 16S
Output Scaling Offset	Float	0 - 2
Output Scaling Factor	Float	0 - 1
Output Thermal Scaling Offset	Float	0 - 1
Output Thermal Scaling Factor	Float	0 - 1

* Required parameter

Parameter descriptions

Input: Input data to be converted

The image layers to convert to reflectance. The layers are converted if they have the required radiometric calibration information, provided by the RadiometricTrans metadata entries (Gain, Offset, Quantity and Units). Layers that do not have the required metadata tags are transferred to the output file as-is, but only if the output file type is PIX. For all other output file types only the radiometrically calibrated (spectral) channels are processed, and a warning is issued with the channels that are omitted from processing.

Note: an exception is made for the Landsat-8 Quality channel. It is transferred to the output file (with the data type conversion), except when the output type is 8U, or the output file format is JP2.

Input: Input Solar Zenith Angles

The input raster layers with the per-pixel solar zenith angles. The file is typically generated with the SOLVIEWZAZ module. Each layer must have the same number of lines and pixels and georeferencing as the input raster layers.

Note that the solar zenith angle is only required if the input layer contains pixel values that represent radiance and relative reflectance. For input data in absolute reflectance this parameter should be defaulted, and is ignored if specified.

Output type

The data type of the output channel(s) to create.

When the number of output channels is greater than one, the specified data type is used for each output channel.

This parameter is optional.

The default value is 16S.

Note: the specified data type (except 8U) is used also for the Landsat-8 quality channel, if selected for processing. All other non-spectral channels selected for processing are transferred with their original data types, but only to the output files in PIX format.

Output Scaling Offset

Specifies scaling offset for converting the computed reflectance values to Digital Numbers (DNs) to store in the output image.

If RefBias is specified, the RefGain parameter must be specified as well, with the same number of values.

The RefGain and RefBias parameters are used jointly to convert the computed reflectance values to DNs as follows:

                            DN = Reflectance * RefGain + RefBias 
                        

Note: the radiometric gain and bias values stored in channel metadata are mutually related to the RefGain and RefBias values as follows:

                            Gain = 1 / RefGain 
                            Bias = - RefBias / RefGain 

                            RefGain = 1 / Gain
                            RefBias = -Bias / Gain
                        

Output Scaling Factor

Specifies scaling factor for converting the computed thermal values to Digital Numbers (DNs) to store in the output image. The specified value(s) must be positive (greater than zero).

If ThGain is specified, the ThBias parameter must be specified as well.

This parameter is optional. Please refer to the SCALOFFS parameter for more details, and for the list of data type-dependent default values.

Output Thermal Scaling Offset

Specifies scaling offset for converting the computed thermal values to digital numbers (DNs) to store in the output image.

If ThBias is specified, the ThGain parameter must be specified as well, with the same number of values.

The RefGain and ThBias parameters are used jointly to convert the thermal values to DNs as follows:

                            DN = Reflectance * ThGain + ThBias 
                        

Note: the radiometric gain and bias values stored in channel metadata are mutually related to the ThGain and ThBias values as follows:

                            Gain = 1 / ThGain 
                            Bias = - ThBias / ThGain 

                            ThGain = 1 / Gain
                            ThBias = -Bias / Gain
                        

Output Thermal Scaling Factor

Specifies scaling factor for converting the computed thermal values to Digital Numbers (DNs) to store in the output image. The specified value(s) must be positive (greater than zero).

If ThGain is specified, the ThBias parameter must be specified as well.

This parameter is optional. Please refer to the SCALOFFS parameter for more details, and for the list of data type-dependent default values.

Details

The module DN2REFLECTANCE performs conversion from radiometrically calibrated input data to absolute reflectance. Input data in FILI can represent radiance, relative reflectance or absolute reflectance (in which case they are copied as-is). Relative reflectance values still contain the effects of the solar zenith angle, while absolute reflectance values have these effects already removed.

The input file channels to process are specified in the DBIC parameter. By default all input channels are processed. If a channel selected for processing has no radiometric calibration metadata, it is copied to the output file without any modifications, as long as the output channel data type (specified by the FTYPE parameter) is PIX. For all other file types only the radiometrically calibrated channels are processed, and a warning is issued with the numbers on non-calibrated channels omitted from processing.

Solar angles are not required if the input file contains absolute reflectances. The conversion of radiances and relative reflectances requires that the solar effects are removed, and therefore these input data types require the knowledge of solar zenith angles. If the per-pixel angles are available, they can be used by specifying the name of the angle file in the FILEANG parameter. The angle channels to use are identified automatically for all input image channels in the DBIC parameter, based on the angle channel metadata tags Quantity (SolarZenithAngle) and BandDescription (AppliesToAllBands, or the standard image band description of individual bands). If an angle channel cannot be identified, the corresponding input image channel is copied to the output image as-is. The solar (and sensor) angle file is usually generated by the SOLVIEWZAZ module, with FILI used as its input. If the solar angle file is not provided, the nominal solar zenith angle at the scene center is used, if required.

The derived absolute reflectance values are written to a new output file, with its name specified in the FILO parameter. The type of output channels is specified by the DATATYPE parameter, and all output radiometrically calibrated channels (and the Landsat-8 Quality channel, if selected for processing) will be of the same type. Some output channel types may require that the computed values are scaled and offset, and the parameters SCALFACT and SCALOFFS can be used to define the required linear transformation. There may be up two values provided for each of the two parameters, with the first value used for all non-thermal bands, and the second value used only for the thermal bands. The content of each processed channel is determined automatically from its metadata. The derived and scaled pixel values (DNs) are clipped to be within the value range in the output channel, as shown in the description of the DATATYPE parameter. Input image channels without the radiometric calibration metadata or the corresponding angle channel (if required) are transferred as-is. However, the non-calibrated channels are appended at the end of the file only for output files in the PIX format. For all other output file formats they are omitted from processing, and their numbers are shown in a warning.

The absolute reflectances should nominally be within the [0, 1] range of values. However, due to inaccurate calibration and other artefacts of processing the actual computed values may lie outside of this range. Depending on the output data type the scaled pixel values may be truncated to be within the range of values supported by the output data type. Note that the truncation never results in the computed value being replaced by a no-data pixel value.

Thermal channels are converted to the brightness temperature in degrees Kelvin, and can never be negative. The highest values may reach hundreds or even thousands of degrees over active fires or volcanoes.