During commissioning of the spacecraft it was found that the sky background observed by Gaia is significantly higher than expected. In the past months, this issue has been investigated by an independent team involving experts from ESA, industry and the Data Processing and Analysis Consortium (DPAC). A multi-disciplinary approach has been adopted, including different kinds of analysis: experimental (background estimation as function of CCD, time, and Solar aspect angle), theoretical (diffraction assessment), laboratory (determination of scattering properties of the Sun shield material), and modelling (straylight ray-tracing simulations). Following these investigations, it is now clear that the straylight is caused by two sources: the Sun and bright objects in the sky, most notably the Milky Way.

The sensitivity of the payload to straylight from out-of-field sources was carefully analysed during the pre-launch development phase and single stellar sources were found to have a negligible impact. However, the collective effect caused by regions on the sky with high stellar densities - such as the Milky Way - was not considered and turned out to be a relevant contributor to the straylight observed in orbit.

The majority of the straylight is caused by the Sun and can be explained as follows. The sun shield of Gaia is composed of two multi-layer blankets. To ensure a high tear strength of these layers during the deployment of the sun shield after launch, these blankets have been reinforced and laminated with a fibre fabric. Some of these fibres have detached and protrude at the edges of the blankets, which has been confirmed by inspection of flight-spare blankets. The fixed panels of the sun shield have been taped, but this was not possible for the blankets which were rolled in the launch configuration. Cutting the fibres was considered too risky in the clean room as some small particles could have found their way into the spacecraft. Laboratory measurements now have confirmed that these fibres substantially scatter sunlight into the shadow region behind the sun shield. Since the fibres are protruding outside the nominal radius of the Sun-facing layer of the sun shield, the shadow-side layer of the sun shield is less efficient in preventing the scattered light entering the telescope apertures. Propagation of the sunlight entering the apertures occurs along straylight paths nominally present in the payload. A quantitative analysis using a detailed opto-mechanical model of the payload, the scattering properties of the fibres, and a realistic Solar illumination reproduces the major features observed. It is not expected that the straylight levels caused by the fibres will increase due to ageing effects.

The straylight primarily affects the performance of Gaia for the faintest objects, in particular in the Radial Velocity Spectrometer (RVS). Mitigation schemes are being implemented to optimise the mission performance, namely:

a) The RVS has been operating in full high-resolution mode since summer 2014. This reduces the total noise, improves the spectral resolution and removes the presence of faint-star spectra which are partially collected in high- and partially collected in low-resolution mode.

b) The on-board software is currently being changed to allow adaptively reducing the across-scan window size for faint objects in RVS as function of time, magnitude, and position in the focal plane. This maximises the signal-to-noise ratio of each individual CCD observation.

c) Improved background-subtraction algorithms have been developed, tested and implemented.

Taking these changes into account, the end-of-mission science-performance predictions have been updated. They can be found here.

Image credits: ESA