This will count as one of your downloads.
You will have access to both the presentation and article (if available).
On-orbit relative calibration of the OLI is primarily monitored and corrected by observing an on-board primary solar diffuser panel. The panel is the most uniform target available to the OLI, though as observed but the OLI, it has a slope across the field of view due to view angle effects. Just after launch, parameters were derived using the solar diffuser data, to correct for the angular effects across the 14 modules. The residual discontinuities between arrays and the detector-to-detector uniformity continue to be monitored on a weekly basis.
The observed variations in the responses to the diffuser panel since launch are thought to be due to real instrument changes. Since launch, the Coastal/Aerosol (CA) and Blue bands have shown the most variation in relative calibration of the VNIR bands, with as much as 0.14% change (3-sigma) between consecutive relative gain estimates. The other VNIR bands (Green, Red and NIR) initially had detectors showing a slow drift of about 0.2% per year, though this stopped after an instrument power cycle about seven months after launch. The SWIR bands also exhibit variability between collects (0.11% 3-sigma) but the larger changes have been where individual detectors’ responses change suddenly by as much as 1.5%.
The mechanisms behind these changes are not well understood but in order to minimize impact to the users, the OLI relative calibration is updated on a quarterly basis in order to capture changes over time.
After TIRS was successfully activated on-orbit, checks were performed on the instrument data to determine its image quality. These checkouts included an assessment of the on-board blackbody and deep space views as well as normal Earth scene collects. The calibration parameters that were determined pre-launch were updated by utilizing data from these preliminary on-orbit assessments. The TIRS on-orbit radiometric performance was then characterized using the updated calibration parameters. Although the characterization of the instrument is continually assessed over the lifetime of the mission, the preliminary results indicate that TIRS is meeting the noise and stability requirements while the pixel-to-pixel uniformity performance and the absolute radiometric performance require further study.
The design of the Landsat-8 instruments is a significant departure from earlier Landsats. The OLI is a pushbroom instrument; all previous recent Landsat instruments were electromechanical (whiskbroom) instruments. OLI also has two new spectral bands and refined bandpasses; the thermal imaging capability on Landsat-8 is in a separate instrument.
The pushbroom design provides significantly better signal to noise performance than historically available, but at the expense of circa 70,000 detectors versus the 100 or so on previous instruments. The large focal plane and large number of detectors makes detector to detector relative calibration more challenging, increasing the propensity for banding and striping in imagery.
On-board radiometric calibration devices include a shutter to measure the dark levels, a full aperture solar panel for calibration against the sun, and multiple sets of lamps for short-term stability monitoring. Early results from the on-board calibration devices indicate that the OLI is outperforming the Landsat-7 instrument in signal-to-noise ratio by an order of magnitude, consistent with pre-launch measurements. Over the first five months, the instrument is stable to within 0.7%, as measured by the lamps and solar diffuser. A relative calibration (detector-to-detector) and a linearization parameter update have been performed that reduce visible striping; with this update, the residual striping has been reduced by half in all OLI bands.
Keywords: Landsat-7, ETM÷, Noise, Radiometry
View contact details
No SPIE Account? Create one
