Weak gravitational lensing is one of the most powerful tools for cosmology, while subject to challenges in quantifying subtle systematic biases. The point spread function (PSF) can cause biases in weak lensing shear inference when the PSF model does not match the true PSF that is convolved with the galaxy light profile. Although the effect of PSF size and shape errors – i.e. errors in second moments – is well studied, weak lensing systematics associated with errors in higher moments of the PSF model require further investigation. The goal of our study is to estimate their potential impact for LSST weak lensing analysis. We go beyond second moments of the PSF by using image simulations to relate multiplicative bias in shear to errors in the higher moments of the PSF model. We find that the current level of errors in higher moments of the PSF model in data from the Hyper Suprime-Cam survey can induce a ∼0.05 per cent shear bias, making this effect unimportant for ongoing surveys but relevant at the precision of upcoming surveys such as LSST.
- Award ID(s):
- 2020295
- NSF-PAR ID:
- 10293510
- Date Published:
- Journal Name:
- Monthly Notices of the Royal Astronomical Society
- Volume:
- 504
- Issue:
- 4
- ISSN:
- 0035-8711
- Page Range / eLocation ID:
- 5543 to 5555
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
ABSTRACT -
more » « less
ABSTRACT Morphology is a fundamental property of any galaxy population. It is a major indicator of the physical processes that drive galaxy evolution and in turn the evolution of the entire Universe. Historically, galaxy images were visually classified by trained experts. However, in the era of big data, more efficient techniques are required. In this work, we present a k-nearest neighbours based approach that utilizes non-parametric morphological quantities to classify galaxy morphology in Sloan Digital Sky Survey images. Most previous studies used only a handful of morphological parameters to identify galaxy types. In contrast, we explore 1023 morphological spaces (defined by up to 10 non-parametric statistics) to find the best combination of morphological parameters. Additionally, while most previous studies broadly classified galaxies into early types and late types or ellipticals, spirals, and irregular galaxies, we classify galaxies into 11 morphological types with an average accuracy of ${\sim} 80\!-\!90 \, {{\rm per\, cent}}$ per T-type. Our method is simple, easy to implement, and is robust to varying sizes and compositions of the training and test samples. Preliminary results on the performance of our technique on deeper images from the Hyper Suprime-Cam Subaru Strategic Survey reveal that an extension of our method to modern surveys with better imaging capabilities might be possible.
-
more » « less
ABSTRACT Using the TNG50 cosmological simulation and observations from the Kilo-Degree Survey (KiDS), we investigate the connection between galaxy mergers and optical morphology in the local Universe over a wide range of galaxy stellar masses (8.5 ≤ log (M*/M⊙) ≤ 11). To this end, we have generated over 16 000 synthetic images of TNG50 galaxies designed to match KiDS observations, including the effects of dust attenuation and scattering, and used the statmorph code to measure various image-based morphological diagnostics in the r-band for both data sets. Such measurements include the Gini–M20 and concentration–asymmetry–smoothness statistics. Overall, we find good agreement between the optical morphologies of TNG50 and KiDS galaxies, although the former are slightly more concentrated and asymmetric than their observational counterparts. Afterwards, we trained a random forest classifier to identify merging galaxies in the simulation (including major and minor mergers) using the morphological diagnostics as the model features, along with merger statistics from the merger trees as the ground truth. We find that the asymmetry statistic exhibits the highest feature importance of all the morphological parameters considered. Thus, the performance of our algorithm is comparable to that of the more traditional method of selecting highly asymmetric galaxies. Finally, using our trained model, we estimate the galaxy merger fraction in both our synthetic and observational galaxy samples, finding in both cases that the galaxy merger fraction increases steadily as a function of stellar mass.
-
more » « less
ABSTRACT Machine learning models can greatly improve the search for strong gravitational lenses in imaging surveys by reducing the amount of human inspection required. In this work, we test the performance of supervised, semi-supervised, and unsupervised learning algorithms trained with the ResNetV2 neural network architecture on their ability to efficiently find strong gravitational lenses in the Deep Lens Survey (DLS). We use galaxy images from the survey, combined with simulated lensed sources, as labeled data in our training data sets. We find that models using semi-supervised learning along with data augmentations (transformations applied to an image during training, e.g. rotation) and Generative Adversarial Network (GAN) generated images yield the best performance. They offer 5 – 10 times better precision across all recall values compared to supervised algorithms. Applying the best performing models to the full 20 deg2 DLS survey, we find 3 Grade-A lens candidates within the top 17 image predictions from the model. This increases to 9 Grade-A and 13 Grade-B candidates when 1 per cent (∼2500 images) of the model predictions are visually inspected. This is ≳ 10 × the sky density of lens candidates compared to current shallower wide-area surveys (such as the Dark Energy Survey), indicating a trove of lenses awaiting discovery in upcoming deeper all-sky surveys. These results suggest that pipelines tasked with finding strong lens systems can be highly efficient, minimizing human effort. We additionally report spectroscopic confirmation of the lensing nature of two Grade-A candidates identified by our model, further validating our methods.
-
Generative models learned from training using deep learning methods can be used as priors in under-determined inverse problems, including imaging from sparse set of measurements. In this paper, we present a novel hierarchical deep-generative model MrSARP for SAR imagery that can synthesize SAR images of a target at different resolutions jointly. MrSARP is trained in conjunction with a critic that scores multi resolution images jointly to decide if they are realistic images of a target at different resolutions. We show how this deep generative model can be used to retrieve the high spatial resolution image from low resolution images of the same target. The cost function of the generator is modified to improve its capability to retrieve the input parameters for a given set of resolution images. We evaluate the model's performance using three standard error metrics used for evaluating super-resolution performance on simulated data and compare it to upsampling and sparsity based image super-resolution approaches.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1093/mnras/stab1214
