Uncertainty-aware learning for improvements in image quality of the Canada–France–Hawaii Telescope

Abstract

We leverage state-of-the-art machine-learning methods and archival data from the Canada-France-Hawaii Telescope (CFHT) to predict observatory image quality (IQ) from environmental conditions and observatory operating parameters. Drawing on nearly a decade’s worth of collected data we develop accurate and interpretable models of the complex dependence between data features and observed IQ for CFHT’s wide field camera, MegaCam. Using specialized loss functions and neural network architectures, we predict the probability distribution function (PDF) of the IQ associated with any given sample. One of our main results is that, based purely on environmental and observatory operating conditions, we can predict the effective MegaPrime IQ to a mean accuracy of ∼0.07″. We further explore how reconfiguration of adjustable observatory parameters can effect improvements in IQ. In particular, we consider actuation of 12 dome “vents”, installed in 2013-14, to accelerate the flushing of hot air from the dome, thereby reducing internal air turbulence and improving IQ. To enable trustworthy predictions of the effect of vent actuation, we first need to differentiate between uncertainties that arise from the randomness inherent to the input data and those which arise due to imperfect modeling. Using predictions of these uncertainties, in conjunction with probabilistic generative modeling, we identify candidate vent adjustments that are in-distribution and, for the optimal in-distribution vent configuration we calculate the predicted reduction in required observing time. The IQ prediction reduction, averaged across all samples, is about $\sim 25\%$. Finally, we use Shapley values to compute robust feature ranking. This allows us to identify the most predictive variables from the sensor data for each observation. Building from this work, our long-term goal is to construct a reliable and robust model that can forecast optimal operating conditions for real-time optimization of IQ. Such forecasts can be fed into real-time scheduling protocols to accelerate scientific productivity, and into predictive maintenance routines. We anticipate that the data-driven approaches we explore herein will become standard in automating observatory operations and in improving observatory maintenance by the time CFHT’s successor, the Maunakea Spectroscopic Explorer (MSE), is installed in the next decade.

Artificial Intelligence in Astronomical Optical Telescopes: Present Status and Future Perspectives

“…In addition, Gilda et al [17] developed a method for predicting the probability distribution function of observed image quality based on environmental conditions and observatory operating parameters using Canada-France-Hawaii Telescope data and a mixture density network. They then combined this approach with a robust variational autoencoder to forecast the optimal configuration of 12 vents to reduce the time required to reach a fixed signal-to-noise ratio (SNR) for observations.…”

“…As early as the 1990s, neural networks were applied in the observation planning of the Hubble Space Telescope (HST) [10], expert systems were used in the fault diagnosis of HST energy systems [11], and statistical machine learning algorithms were widely used in the preprocessing of database data to label quasars, stars, and galaxies [12][13][14]. With AI's evolution, its applications in telescope intelligence have broadened, encompassing the selection of excellent stations, the calibration of telescope optical systems, and the optimization of imaging quality [15][16][17]. In general, large, ground-based astronomical telescopes are integrated optical and mechatronics devices encompassing mechanical and drive systems, optical paths and optical systems, imaging observation, control systems, and environmental conditions.…”

Fighting the devil within: improving image quality by thwarting local turbulence

“…For example, we know that an excess of high spatial frequencies will flatten the phase structure function and produce stronger wings on the PSF, lowering the beta index of a Moffat fit. The correlation between IQ and turbulence and environmental sensors could take the form of a principal component analysis, but we suspect that due to the non-linear onset of turbulence this may be a task best performed by machine learning which may be better able to account for turbulence step functions and sudden onset of the turbulent regime (Gilda et al, 2022).…”

Beyond mirkwood: Enhancing SED Modeling with Conformal Predictions

“…Machine learning (ML), with its ability to handle large datasets and uncover complex patterns, has emerged as a powerful tool in SED fitting [1,2,7,8]. The traditional parametric and often linear approaches are being supplemented, and in some cases replaced, by nonparametric, highly flexible ML techniques that can model the non-linear relationships intrinsic to astronomical data more effectively [9][10][11][12][13]. This paradigm shift is not just a matter of computational convenience but represents a fundamental change in how we interpret vast and complex astronomical datasets.…”