The PAU Survey: Background light estimation with deep learning techniques

Abstract

In any imaging survey, measuring accurately the astronomical background light is crucial to obtain good photometry. This paper introduces BKGnet, a deep neural network to predict the background and its associated error. BKGnet has been developed for data from the Physics of the Accelerating Universe Survey (PAUS), an imaging survey using a 40 narrow-band filter camera (PAUCam). Images obtained with PAUCam are affected by scattered light: an optical effect consisting of light multiply that deposits energy in specific detector regions contaminating the science measurements. Fortunately, scattered light is not a random effect, but it can be predicted and corrected for. We have found that BKGnet background predictions are very robust to distorting effects, while still being statistically accurate. On average, the use of BKGnet improves the photometric flux measurements by 7% and up to 20% at the bright end. BKGnet also removes a systematic trend in the background error estimation with magnitude in the i-band that is present with the current PAU data management method. With BKGnet, we reduce the photometric redshift outlier rate

16 pages, 13 figures

References in corpus (30)

• Adam: A Method for Stochastic Optimization
• Scikit-learn: Machine Learning in Python
• Gaia Data Release 2. Summary of the contents and survey properties
• LSST: from Science Drivers to Reference Design and Anticipated Data Products
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• Empirical Evaluation of Rectified Activations in Convolutional Network
• Accurate photometric redshifts for the CFHT Legacy Survey calibrated using the VIMOS VLT Deep Survey
• SparsePak: A Formatted Fiber Field Unit for The WIYN Telescope Bench Spectrograph. I. Design, Construction, and Calibration
• The COSMOS2015 Catalog: Exploring the 1<z<6 Universe with half a million galaxies
• zCOSMOS: A Large VLT/VIMOS redshift survey covering 0 < z < 3 in the COSMOS field
• Euclid Definition Study Report
• Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics
• The Dark Energy Survey Data Release 1
• Visualizing and Understanding Convolutional Networks
• Deep Learning for Real-time Gravitational Wave Detection and Parameter Estimation: Results with Advanced LIGO Data
• Photometric redshifts from SDSS images using a Convolutional Neural Network
• Photometric redshift estimation via deep learning
• Cosmological constraints from noisy convergence maps through deep learning
• Photometric redshifts for the Kilo-Degree Survey. Machine-learning analysis with artificial neural networks
• The PAU Survey: Early demonstration of photometric redshift performance in the COSMOS field
• Precise photometric redshifts with a narrow-band filter set: The PAU Survey at the William Herschel Telescope
• DeepSource: Point Source Detection using Deep Learning
• Fast Point Spread Function Modeling with Deep Learning
• Deep Learning for Image Sequence Classification of Astronomical Events
• The Physics of the Accelerating Universe Camera
• The PAU Survey: star-galaxy classification with multi narrow-band data
• The PAU Survey: Spectral features and galaxy clustering using simulated narrow band photometry
• The PAU Survey: A Forward Modeling Approach for Narrow-band Imaging
• A method of complex background estimation in astronomical images
• The PAU Survey: Operation and orchestration of multi-band survey data

Cited by in corpus (8)

• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• The PAU Survey: Measurement of Narrow-band galaxy properties with Approximate Bayesian Computation
• The PAU Survey: Intrinsic alignments and clustering of narrow-band photometric galaxies
• The PAU survey: Estimating galaxy photometry with deep learning
• The PAU Survey: Narrow-band image photometry
• The PAU Survey: Close galaxy pairs identification and analysis
• Estimating Cosmological Parameters and Reconstructing Hubble Constant with Artificial Neural Networks: A Test with covariance matrix and mock H(z)
• The PAU Survey: Classifying low-z SEDs using Machine Learning clustering