Source: Brian Markham

The Landsat Calibration Working Group held their semi-annual meeting in Tahoe City, Nevada on December 2-3, 2009. USGS/EROS Image Assessment System (IAS), NASA/GSFC Landsat Project Science Office, South Dakota State University, Rochester Institute of Technology (RIT), University of Arizona, and NASA/JPL calibration personnel attended the meeting. Significant progress in a number of areas of Landsat calibration was reported at the meeting, including:
  • Continued good geometric and radiometric performance are observed for the Landsat 7 ETM+. Alignment updates typically occur quarterly and bumper parameters updates every 2-3 weeks. Two detectors in band 5 show an increase in the coherent noise levels, with magnitudes up to 1 DN.
  • Continued stability is observed in the radiometric calibration of the Landsat 7 ETM+ reflective bands with no required update to the calibration.
  • The August 13, 2009 Landsat 5 attitude anomaly resulted in an apparent warming of the TM cold focal plane, the removal of the window contamination and resetting of the cold focal plane band gains. The CPF was updated to account for this change.
  • The Landsat 5 TM geometric performance continues to be good, with sensor alignment updates and bumper mode parameter updates on a schedule similar to Landsat 7 ETM+.
  • Updates to the calibration of the thermal bands on both Landsat 5 and 7 have been developed. These adjustments are made possible by a new method of using NOAA buoys for vicarious calibrations and the recent addition of a warm water calibration site in the Salton Sea. In general, the calibration adjustments have very little impact on the radiometric calibration across typical water temperatures (4° C to 20° C), but will increase the derived temperature for warm targets (+1.6°C at 40° C for Landsat 7 and +1.3 ° at 40° for Landsat 5) and decrease the derived temperature for cold targets (-0.7° C at 0° C for Landsat 7 and -0.5° C at 0° for Landsat 5). The Landsat 7 update went into effect 1 January 2010 for all newly processed data; the current scheme of Landsat data production will result in some lag in the application to all data. Scenes that have already been processed to level 1T and remain in the on-line cache will still reflect the old calibration until refreshed. The Landsat 5 update will be effective April 1, 2010 and will affect all data acquired since launch.
  • The Landsat 5 TM relative gain trends are being studied in more detail with the large number of scenes that have now been processed through IAS. Refinements to the current models are anticipated to reduce the residual striping in TM data.
  • Two detectors in band 2 of the Landsat 4 TM showed degraded MTF performance; one is so poor that it has not been used in data production at all over the life of the mission. Analysis of the MTF of these two detectors has been performed and a correction algorithm developed. Initial results are promising showing a clear improvement in these two detectors. Refinement will continue.
  • An update to the radiometric calibration of the Landsat 1-4 MSS sensors to match the Landsat 5 MSS has been developed. This calibration brings data from the 4 sensors to within 5% of the Landsat 5 MSS calibration over the lives of all the missions. This update is currently in test. The multiple flavors of archived MSS data (X-format, P-format, etc) will result in a phased-in application of the new calibration coefficients. Initially P format data will be processed by LPGS with implementation expected late winter – early Spring 2010. As LPGS is expanded to cover the other formats of data, the improvements will expand to the rest of the archive.

—————–
One of Simon Hook’s thermal calibration buoys on Lake Tahoe was visited by the calibration working group team members. Though the water was too rough to board the buoy, the sensors were observed and their operation explained. On the last visit of the day the water was calm enough to change out one of the sensors.