While Landsat satellites are well-designed and calibrated prior to launch, continuous re calibration is required to offset degradation that may be caused by mechanical or electrical effects, or exposure to UV radiation. Calibration requires a comparison between the measuring instrument and an “absolute” reference standard of known accuracy.
The calibration of Landsat sensors is supported by pre-flight, post-launch onboard, and ground reference data, which is continually researched and updated by the USGS EROS Cal/Val Center of Excellence, NASA Cal/Val Team, along with funded university partners working in vicarious calibration.
Radiometric Characterization and Calibration
Reflectance, radiance, and atmospheric conditions all affect the digital number (DN) values of Landsat data. Radiometric calibration employs algorithms and processes that improve Landsat data. This is done by converting the DN values of the data to spectral radiance (at the sensor), and then to reflectance (also at the sensor). This is followed by the removal of atmospheric effects, which are due to absorption and scattering, to perform atmospheric correction (reflectance at the surface).
Pseudo Invariant Calibration Sites
Pseudo Invariant Calibration Sites (PICS) are considered benchmarks for on- orbit calibration, as they are:
- Essentially invariant over time
- Are spatially very uniform, have stable spectral response s over time
- Atmospheric effects on upwelling radiance is minimal due to high surface reflectance
- Are in regions where rainfall is extremely limited:
- Prevents vegetative growth
- Very sparse human populations
(Ref: Absolute Calibration of Optical Satellite Sensors Using Libya 4 Pseudo Invariant Calibration Site, Remote Sensing 2014, 6(2), 1327-1346; doi:10.3390/rs6021327.)
Validating the geometric accuracy of Landsat data incorporates the known processing levels, the root-mean-square error (RMSE) information, and the use of Ground Control Points (GCPs), which are points on the surface of the earth of known location used to geo-reference Landsat Level-1 imagery. In 2014, improvements to GCP files began, and include removing outdated files, creating new, time-specific and seasonal GCP’s. The Landsat Geometry page provides more information on the work done to create high-quality data.
An ephemeris is a set of data that provides the assigned places of a celestial body (including a manmade satellite) for regular intervals.
New Landsat 7 ETM+ acquisitions are processed immediately using a predicted ephemeris. After one to two days, definitive ephemeris (DE) is applied to the data. DE shows the position and velocity of the satellite in one-minute intervals. There is no DE for Landsat 8, since the Payload Correction Data (PCD) for the satellite is very accurate. Telemetry information in the form of Two-Line Element (TLE) Set — which lists the orbital elements that describe the state (position and velocity) of the spacecraft — is listed on http://www.n2yo.com/satellite/?s=39084.
Calibration Parameter Files
Calibration Parameter Files (CPFs) provide radiometric and geometric coefficients needed for processing of raw, uncorrected Landsat image data. Each timeframe-specific CPF contains calibration coefficients that have been adjusted to correct for the time varying performance of the sensor.
The Calibration Parameter File page allows the searching and downloading of CPFs specific to each Landsat sensor, and displays definition documents and notices about changes that affect the files.
For Landsat 8, Bias Parameter Files (BPF) and Response Linearization Look Up Tables (RLUT) are also available. BPFs supply radiometric correction parameters during Level-1 processing of Landsat 8 data products. RLUTs are optional additional files that accompany Calibration Parameter Files (CPF) and contain a mapping look up table to linearize the output of the OLI detectors.
Calibration notices describe changes made to the calibration that affects Landsat data products.
Landsat Calibration/ Validation News Archive
February 8, 2018 - Landsat TM "no-Payload Correction" Data (PCD). PCD provides critical information when correcting for geometric distortions inherent to the imaging system. Over 250,000 Landsat 5 Thematic Mapper (TM) scenes acquired between January 1986 and December 1999, originally downlinked to international ground stations, lack the necessary PCD. Until recently, these scenes failed Landsat Level-1 data processing and could not be added to the Landsat data inventory. However, as of early February 2018, processing algorithm changes make it possible to process the majority of these scenes into the Landsat Collection 1 data inventory and make them available for download from EarthExplorer and GloVis.
The Landsat TM no-PCD page provides more technical details about these data.
Solar Illumination and Sensor Viewing Angle Coefficient Files
Angle Coefficient Files are included in all Landsat Collection 1 Level-1 data products, and serve to provide sensor viewing angle model coefficients that can be used to compute the solar and sensor viewing angles for every pixel. The Solar Illumination and Sensor Viewing Angle Coefficient Files page provides angle creation tools and technical details.
Known Issues in Landsat Data
Landsat data are geometric, radiometric, and terrain corrected to provide the highest quality data to the user communities. Occasionally, anomalies occur and artifacts are discovered that require research and monitoring.
The Landsat Calibration and Validation (Cal/Val) team investigates and tracks anomalous data. Details about a number of anomalies that have been discovered and investigated can be found on the Landsat Known Issues page.These artifacts vary widely between the MSS, TM, ETM+ and OLI/TIRS sensors caused by specific sensor characteristics and anomalies identified after launch.