Predicting the movement and development of Cloud for Dynamically scheduling Astronomical Observations
The purpose of this project is to develop software and procedures to assist with this dynamic observation scheduling, based on the current time series of cloud images.
Astronomical observatories employ a variety of techniques for detecting night-time cloud, including infra-red and visible cameras. The infra-red and visible wavebands are complementary, in that clouds radiate infra-red radiation but also block visible light from background stars. We use imaging infra-red cameras since techniques which search for sky without stars can be computationally slow and depend on a suitable spatial density of stars over the whole sky.
The Cherenkov Telescope Array (CTA) will be a major world observatory, based at two sites, to study gamma-rays from many selected interesting astronomical sources, and observing schedules will be optimised for a selection of sources each night.
However, if there is some cloud, those schedules will require optimising to still observe as much as possible in directions away from the cloud, as the cloud develops and moves. This requires not only a knowledge of the present cloud locations but also a prediction of where cloud is likely to move over the following period of observations (minutes to hours).
Cloud images are presently being recorded over Adelaide which will provide data for this investigation.