Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Calibration of Landsat 5 thermal infrared channel: updated calibration history and assessment of the errors associated with the methodology
Ist Teil von
Canadian journal of remote sensing, 2010-10, Vol.36 (5), p.617-630
Ort / Verlag
Taylor & Francis
Erscheinungsjahr
2010
Link zum Volltext
Quelle
Alma/SFX Local Collection
Beschreibungen/Notizen
The Landsat 5 thermal band lifetime calibration is being updated based on an improved calibration method that uses water temperatures observed by
buoys at deep water sites and thermal and radiative transfer models. An uncertainty propagation analysis was constructed to determine the expected
uncertainty in temperature (one standard deviation) at the sensor for this historic vicarious calibration process. The historical calibration effort
fused environmental data sources that feed a forward modeling vicarious calibration process. The process consists of three major modeling efforts:
subsurface temperature to water skin temperature, atmospheric radiative transfer, and sensor noise modeling. Each modeling effort was investigated
uniquely, and the results were combined to derive the total process error. The uncertainty propagation results indicate that the historic vicarious
calibration process has an expected uncertainty of ±0.5 K. This conclusion is consistent with the observed root mean square error (RMSE)
between observed and predicted values using this method. After the instrument calibration was updated, the difference between instrument-derived
radiance (observed data spanning a 23 year period) and radiance estimated using the subsurface buoy temperatures was 0.6 K RMSE. This
demonstrates that the residual error in the observed calibration results and the expected process uncertainty are essentially comparable. Using the
error analysis results, the data from the historical buoy temperature study were combined with data from traditional surface temperature and radiance
studies (1999-2000) to generate a lifetime calibration update for the Landsat 5 instrument. The final calibration uses data from the long
established thermal calibration sites on the Great Lakes (Erie and Ontario), the Salton Sea, and Lake Tahoe, as well as a number of additional deep
water sites where National Oceanic and Atmospheric Administration (NOAA) bouys and atmospheric sounding data provide adequate ground truth for the
historical calibration approach. This updated calibration has been implemented in the U.S. Geological Survey (USGS) - National Aeronautics and
Space Administration (NASA) processing system. These results indicate that the image data is calibrated to better than 0.67 K (one sigma) over
its 25+ year record. While this work rigorously investigated the historic thermal vicarious calibration process for Landsat 5 Thematic Mapper
(TM), the approach and the new study sites can be easily extended to the investigation of similar systems.