You must have a javascript-enabled browser and javacript and stylesheets must be enabled to use some of the functions on this site.


Revised polarisation calibration of SCIAMACHY

J.M. Krijger(1), R. Snel(1) and Sander Slijkhuis(2)

(1) SRON, Sorbonnelaan 2, 3584 CA Utrecht, Netherlands
(2) DLR, Münchner Straße 20, 82234 Weßling, Germany


The imaging spectrometer SCIAMACHY on ENVISAT has been collecting data since launch in 2002. Over the years the exposure to space has affected the optical performance and rendered the on-ground calibration data increasingly more outdated. As the overall calibration quality is continuously being improved through better in-flight characterisation and updated algorithms, instrument aspects with previously acceptable errors are now in need of improvement. This presentation covers the polarisation characterisation as derived from on-ground calibration measurements, and its development in orbit.

The original polarisation calibration of SCIAMACHY was described in the form of so-called "greek" key data, describing various ratios of instrument responses to fully polarised light. It assumed certain polarisation properties of the instrument which were not necessarily fulfilled, and complicated the polarisation analysis and correction. As a result of unexpected polarisation behaviour due to stress birefringe in one of the optical components, and of a split calibration in thermal vacuum and ambient conditions, ambiguities arose in the definition of the polarisation frame which resulted in polarisation correction errors. The correction errors canceled for nadir conditions, but could amount to non-negligible values in limb geometry.

By describing the instrument with a Mueller matrix for the scanner unit and a Stokes vector sensitivity for the optical bench module (OBM) it was possible to describe the polarisation behaviour in a general sense. Ambiguities in the polarisation frame definition were removed. The new approach required re-analysis of the on-ground calibration measurements, and pivots on measurements which were initially not intended for this type of calibration. Using scans with a linear polariser of which the plane of polarisation is rotated, in both limb and nadir viewing geometry, in combination with the ambient calibration of the scanner unit, it was possible to derive the polarisation sensitivity of the OBM independently of the scanner unit. The use of the polariser scan enabled to derive even the circular polarisation sensitivity of the OBM, which turned out to be of significance to the overall polarisation sensitivity for linearly polarised light.

With the known contamination of the scanner unit in orbit it is then possible to calculate calibration key data for the contaminated mirror surfaces, thus changing the radiometric, polarisation, and scan-angle calibration of the instrument.