Cosmic microwave background experiments have experienced an exponential increase in complexity, data size and sensitivity. One of the goals of current and future experiments is to characterize the B-mode power spectrum, which would be considered a strong evidence supporting inflation. The signal associated with inflationary B-modes is very weak, and so a successful detection requires exquisite control over systematic effects, several of which might arise due to the interaction between the electromagnetic properties of the telescope beam, the scanning strategy and the sky model. In this work, we present the Pixel Space COnvolver (PISCO), a new software tool capable of producing mock data streams for a general CMB experiment. PISCO uses a fully polarized representation of the electromagnetic properties of the telescope. PISCO also exploits the massively parallel architecture of Graphic Processing Units to accelerate the main calculation. This work shows the results of applying PISCO in several scenarios, included a realistic simulation of an ongoing experiment, the Cosmology Large Angular Scale Surveyor.
I present a method for the fast convolution of a model galaxy profile by a point-spread function (PSF) model represented as a pixel grid. The method relies upon three observations: First, most simple radial galaxy profiles of common interest (deVaucouleurs, exponential, Sersic) can be approximated as mixtures of Gaussians. Second, the Fourier transform of a Gaussian is a Gaussian, thus the Fourier transform of a mixture-of-Gausssian approximation of a galaxy can be directly evaluated as a mixture of Gaussians in Fourier space. Third, if a mixture component would result in Fourier-space aliasin…
Microwave Kinetic Inductance Detectors (MKIDs) are a class of superconducting cryogenic detectors that simultaneously exhibit energy resolution, time resolution and spatial resolution. The pixel yield of MKID arrays is usually a critical figure of merit in the characterisation of an MKIDs array. Currently, for MKIDs intended for the detection of optical and near-infrared photons, only the best arrays exhibit a pixel yield as high as 75-80%. The uniformity of the superconducting film used for the fabrication of MKIDs arrays is often regarded as the main limiting factor to the pixel yield of an …
University of California, Santa Barbara（加州大学圣塔芭芭拉分校）；Jet Propulsion Laboratory（喷气推进实验室）；California Institute of Technology（加州理工学院）
We present the development of a machine learning based pipeline to fully automate the calibration of the frequency comb used to read out optical/IR Microwave Kinetic Inductance Detector (MKID) arrays. This process involves determining the resonant frequency and optimal drive power of every pixel (i.e. resonator) in the array, which is typically done manually. Modern optical/IR MKID arrays, such as DARKNESS (DARK-speckle Near-infrared Energy-resolving Superconducting Spectrophotometer) and MEC (MKID Exoplanet Camera), contain 10-20,000 pixels, making the calibration process extremely time consu…
We describe a novel method for the application of Convolutional Neural Networks (CNNs) to fields defined on the sphere, using the HEALPix tessellation scheme. Specifically, We have developed a pixel-based approach to implement convolutional layers on the spherical surface, similarly to what is commonly done for CNNs in Euclidian space. The algorithm is fully integrable with existing libraries for NNs (e.g., PyTorch or TensorFlow). We present two applications: (i) recognition of handwritten digits projected on the sphere; (ii) estimation of cosmological parameter from Cosmic Microwave Backgroun…
In this work, we propose two convolutional neural network classifiers for detecting contaminants in astronomical images. Once trained, our classifiers are able to identify various contaminants, such as cosmic rays, hot and bad pixels, persistence effects, satellite or plane trails, residual fringe patterns, nebulous features, saturated pixels, diffraction spikes, and tracking errors in images. They encompass a broad range of ambient conditions, such as seeing, image sampling, detector type, optics, and stellar density. The first classifier, MaxiM ask , performs semantic segmentation and genera…
Building upon existing signal processing techniques and open-source software, this paper presents a baseline design for an RF System-on-Chip Frequency Division Multiplexed readout for a spatio-spectral focal plane instrument based on low temperature detectors. A trade-off analysis of different FPGA carrier boards is presented in an attempt to find an optimum next-generation solution for reading out larger arrays of Microwave Kinetic Inductance Detectors (MKIDs). The ZCU111 RF SoC FPGA board from Xilinx was selected, and it is shown how this integrated system promises to increase the number of …
We introduce a publicly available full-sky beam convolution code library intended to inform the design of future cosmic microwave background (CMB) instruments and help current experiments probe potential systematic effects. The code can be used to assess the impact of optical systematics on all stages of data reduction for a realistic experiment, including analyses beyond power spectrum estimation, by generating signal timelines that may serve as input to full analysis pipelines. The design and mathematical framework of the Python code is discussed along with a few simple benchmarking results.…
Jet Propulsion Laboratory, California Institute of Technology（加州理工学院喷气推进实验室）；Scientific Computing and Imaging Institute, University of Utah（犹他大学科学计算与影像研究所）
The cosmic microwave background (CMB) is a significant source of knowledge about the origin and evolution of our universe. However, observations of the CMB are contaminated by foreground emissions, obscuring the CMB signal and reducing its efficacy in constraining cosmological parameters. We employ deep learning as a data-driven approach to CMB cleaning from multi-frequency full-sky maps. In particular, we develop a graph-based Bayesian convolutional neural network based on the U-Net architecture that predicts cleaned CMB with pixel-wise uncertainty estimates. We demonstrate the potential of t…
In this work, six convolutional neural networks (CNNs) have been trained based on %different feature images and arrays from the database including 15,638 superflare candidates on solar-type stars, which are collected from the three-years observations of Transiting Exoplanet Survey Satellite ({\em TESS}). These networks are used to replace the artificially visual inspection, which was a direct way to search for superflares, and exclude false positive events in recent years. Unlike other methods, which only used stellar light curves to search superflare signals, we try to identify superflares th…
We present a compact two-lens HDPE f/1.6 refractor design that is capable of supporting a 28-deg diffraction-limited field of view at 1-mm wavelengths and contrast it to a similar two-lens refractor using silicon lenses. We compare the optical properties of these two systems as predicted by both geometrical and physical optics. The presented analysis suggests that by relaxing telecentricity requirements, a plastic two-lens refractor system can perform comparably to a similar silicon system across a wide field of view and wavelengths up to 1 mm. We show that for both telescope designs, cold sto…
