Week 17 of the Early Operations system optimization period
With the successful repair reported last week, and confirmation of stable vacuum pressure of the LSST Camera (LSSTCam) cryostat during a series of test movements, nighttime observations with the Simonyi Survey Telescope resumed on 16 February.
The first night back on sky included dedicated observations to refine the telescope pointing model by gathering data across a range of telescope elevation and azimuth angles, as well as Camera rotator angles, in open loop control relying only on the precomputed look-up tables for the telescope optics. Tests later in the week using the updated pointing model, which now includes explicit dependence on the Camera rotator angle, demonstrated substantial improvements, and showed minimal dependence of residual pointing offsets on Camera rotator angle. The pointing model is connected to many aspects of the system performance, including tracking during exposures, more robust selection of candidate stars for wavefront sensing for the Active Optics System (AOS) closed loop control system, and accurate placement for the regions of interest (ROIs) around bright stars with the guider sensors. With the refined pointing model in place, the team is preparing to test an updated guider ROI specification with smaller ROI sizes, and correspondingly faster integration times, that will enable more rapid temporal sampling to characterize the telescope tracking, free atmosphere, and dome seeing effects within individual exposures in survey operations.
On-sky tests during week demonstrated that improvements to summit computing infrastructure to enable faster data ingest and improved parallelization have reduced the total time to analyze the corner wavefront sensor data and deliver AOS closed loop corrections to roughly 40 seconds for the majority of exposures. With this speed improvement, the AOS closed loop corrections derived from visit index N are now regularly being applied to visit N + 2 during survey operations (as compared to the previous N + 3 performance). The upgrades provide a more responsive optical control loop.
This week, the team partially installed a ring of baffle extension plates around the mid-level baffle of the telescope to block the unwanted light path responsible for the most prominent and prevalent stray light feature found during on-sky commissioning. The team conducted dedicated on-sky tests targeting a bright star before and after the partial installation to confirm blockage of the unwanted light path.
The team also continued pre-LSST observations driven by the Feature Based Scheduler (FBS).
The currently installed filter set is grizy. No filter swaps were conducted in February on account of the downtime and needs to prioritize other daytime engineering tasks.