Modeling Light Enlightens Telescope Design

May 5, 2015

In a nice marriage of theory and experiment, KIPAC scientists have investigated the effects of small layers of contamination on optical surfaces, which is important in building the super telescope that will probe dark energy.

The stars are the measured light change through a glass disk with a ~13 micron layer of condensed water on it. The colored curves are the modeled transmittance change through different ice configurations, which can be summed to give the red curve.

The Large Synoptic Survey Telescope (LSST) project is a big collaboration between the Department of Energy, National Science Foundation, and many institutions, and one of the top priorities for American astrophysics. Once built, it will systematically survey the sky to an unprecedented depth, likely revolutionizing many areas of astronomy. One of the most prominent of these areas is constraining dark energy, where LSST is an integral component of the DOE's mission.
LSST will get a handle on dark energy by measuring the shapes and colors of a billion or more galaxies. These measurements require precise calibration of the response of the LSST system to light - converting measured charge in a CCD pixel to a brightness on the sky. One of the main challenges that has dogged telescopes in the past is contamination on the optics, including the CCD detectors themselves, which can alter the light transmittance to the detectors in an unknown way. As the CCD surfaces are usually kept very cold an in an isolated vessel, anything within the vessel is a potential source of contaminants. This will be a special challenge for LSST, which will feature the largest telescope camera ever, to be constructed at SLAC.
KIPAC scientist Andrew Rasmussen has been studying the effects of contamination on optics, using both calculations from fundamental theory and computer software to model the propagation of light through different media and at the surfaces between them. When the potential contaminant is individual particles, of the kind that might arise through exposure to ambient rooms and people during assembly, the effects of absorption and scattering of light can be calculated using specialized theories of geometric optics. When the potential contaminant is tiny microscopic layers of a substance that has condensed onto a cold surface, a miniature version of dew on grass, computer code that computes the interactions of light with different materials is needed.
Rasmussen models the effects of contamination across all of the wavelengths of light that LSST will be sensitive to. These calculations are necessary to engineer the design and construction of the LSST camera to keep contamination to a level that will not compromise the science goals of the telescope. For instance, reconstructing the redshifts of distant galaxies to a certain percent accuracy - an essential component of the 'weak lensing' analysis that will constrain dark energy - requires a certain stability of light transmittance in different filter bands, which in turn can be translated to an allowable layer thickness of condensation on surfaces such as the CCDs, lenses, and filters with Rasmussen's calculations, which can then guide the materials and vacuum strategies used for the camera.
KIPAC scientist Rafe Schindler and postdoc Jack Singal conduct a program of testing materials and contaminant absorption plans in order to reach the outgassing specifications for the LSST camera. Circling back to Rasmussen's work, a measurement they have carried out determined the deposition rate and therefore layer thickness of outgassed water onto a glass disk maintained at a cold temperature in vacuum. They then measured the light transmittance change through the disk in wavelength bands corresponding to the LSST filters. Given the determined thickness of the condensed water layer, Rasmussen was able to accurately model the scattering processes in the water layer, and at the boundry bewteen the water layer and glass, that contributed to the transmittance decline that was measured, reproducing the measured results. This is a small example of the ways in which simulations and measurements come together in the design of the Universe's movie camera, LSST.
This work is based in part on a paper published in Review of Scientific Instruments (Rev Sci Instrum, 2009, 81, 025181).
Science Contact:
Dr. Andrew Rasmussen