arXiv subjects

David W. Hogg

Publications and source records attributed to David W. Hogg.

At least 37 records · Page 2Linked to original sources

arXiv2024

A Data-Driven Search For Mid-Infrared Excesses Among Five Million Main-Sequence FGK Stars

Stellar infrared excesses can indicate various phenomena of interest, from protoplanetary disks to debris disks, or (more speculatively) techno-signatures along the lines of Dyson spheres. In this paper, we conduct a large search for such excesses, designed as a data-driven contextual anomaly detection pipeline. We focus our search on FGK stars close to the main sequence to favour non-young host stars. We look for excess in the mid-infrared, unlocking a large sample to search in while favouring extreme IR excess akin to the ones produced by Extreme Debris Disks (EDD). We combine observations from ESA Gaia DR3, 2MASS, and the unWISE of NASA WISE, and create a catalogue of 4,898,812 stars with $G < 16$ mag. We consider a star to have an excess if it is substantially brighter in $W1$ and $W2$ bands than what is predicted from an ensemble of machine-learning models trained on the data, taking optical and near-infrared information as input features. We apply a set of additional cuts (derived from the ML models and the objects' astronomical features) to avoid false-positive and identify a set of 53 objects (a rate of $1.1\times 10^{-5}$), including one previously identified EDD candidate. Typical infrared-excess fractional luminosities we find are in the range 0.005 to 0.1, consistent with known EDDs.

astro-ph.EP↗

arXiv2024

Improving Radial Velocities by Marginalizing over Stars and Sky: Achieving 30 m/s RV Precision for APOGEE in the Plate Era

The radial velocity catalog from the Apache Point Observatory Galactic Evolution Experiment (APOGEE) is unique in its simultaneously large volume and high precision as a result of its decade-long survey duration, multiplexing (600 fibers), and spectral resolution of $R \sim 22,500$. However, previous data reductions of APOGEE have not fully realized the potential radial velocity (RV) precision of the instrument. Here we present an RV catalog based on a new reduction of all 2.6 million visits of APOGEE DR17 and validate it against improved estimates for the theoretical RV performance. The core ideas of the new reduction are the simultaneous modeling of all components in the spectra, rather than a separate subtraction of point estimates for the sky, and a marginalization over stellar types, rather than a grid search for an optimum. We show that this catalog, when restricted to RVs measured with the same fiber, achieves noise-limited precision down to 30 m/s and delivers well-calibrated uncertainties. We also introduce a general method for calibrating fiber-to-fiber constant RV offsets and demonstrate its importance for high RV precision work in multi-fiber spectrographs. After calibration, we achieve 47 m/s RV precision on the combined catalog with RVs measured with different fibers. This degradation in precision relative to measurements with only a single fiber suggests that refining line spread function models should be a focus in SDSS-V to improve the fiber-unified RV catalog.

astro-ph.IM↗

arXiv2024

Signal-preserving CMB component separation with machine learning

Analysis of microwave sky signals, such as the cosmic microwave background, often requires component separation with multi-frequency methods, where different signals are isolated by their frequency behaviors. Many so-called "blind" methods, such as the internal linear combination (ILC), make minimal assumptions about the spatial distribution of the signal or contaminants, and only assume knowledge of the frequency dependence of the signal. The ILC is a minimum-variance linear combination of the measured frequency maps. In the case of Gaussian, statistically isotropic fields, this is the optimal linear combination, as the variance is the only statistic of interest. However, in many cases the signal we wish to isolate, or the foregrounds we wish to remove, are non-Gaussian and/or statistically anisotropic (in particular for Galactic foregrounds). In such cases, it is possible that machine learning (ML) techniques can be used to exploit the non-Gaussian features of the foregrounds and thereby improve component separation. However, many ML techniques require the use of complex, difficult-to-interpret operations on the data. We propose a hybrid method whereby we train an ML model using only combinations of the data that $\textit{do not contain the signal}$, and combine the resulting ML-predicted foreground estimate with the ILC solution to reduce the error from the ILC. We demonstrate our methods on simulations of extragalactic temperature and Galactic polarization foregrounds, and show that our ML model can exploit non-Gaussian features, such as point sources and spatially-varying spectral indices, to produce lower-variance maps than ILC - eg, reducing the variance of the B-mode residual by factors of up to 5 - while preserving the signal of interest in an unbiased manner. Moreover, we often find improved performance when applying our model to foreground models on which it was not trained.

astro-ph.CO↗

arXiv2024

Iron Snails: non-equilibrium dynamics and spiral abundance patterns

Galaxies are not in a dynamical steady state. They continually undergo perturbations, e.g., from infalling dwarf galaxies and dark-matter substructure. After a dynamical perturbation, stars phase mix towards a new steady state; in so doing they generally form spiral structures, such as spiral density waves in galaxy disks and the Gaia Snail observed in the vertical phase-space density in the solar neighborhood. Structures in phase-space density can be hard to measure accurately, because spatially varying selection effects imprint their own patterns on the density. However, stellar labels such as metallicity, or other element abundances, or stellar masses and ages, can be measured even in the face of complex or unknown spatial selection functions. We show that if the equilibrium galaxy has phase-space gradients in these labels, any perturbation that could raise a spiral wave in the phase-space density will raise a spiral wave in the distribution of labels as well. We work out the relationship between the spiral patterns in the density and in the labels. As an example, we analyze the Gaia Snail and show that its amplitude and dynamical age as derived from elemental abundances (mainly [Mg/Fe]) follow similar patterns to those derived from the phase-space density. Our best model dates the Snail's perturbation to about 400 Myr ago although we find significant variations with angular momentum in the best-fit age. Conceptually, the ideas presented here are related to Orbital Torus Imaging, chemical tagging, and other methods that use stellar labels to trace dynamics.

astro-ph.GA↗

arXiv2024

Is machine learning good or bad for the natural sciences?

Machine learning (ML) methods are having a huge impact across all of the sciences. However, ML has a strong ontology - in which only the data exist - and a strong epistemology - in which a model is considered good if it performs well on held-out training data. These philosophies are in strong conflict with both standard practices and key philosophies in the natural sciences. Here we identify some locations for ML in the natural sciences at which the ontology and epistemology are valuable. For example, when an expressive machine learning model is used in a causal inference to represent the effects of confounders, such as foregrounds, backgrounds, or instrument calibration parameters, the model capacity and loose philosophy of ML can make the results more trustworthy. We also show that there are contexts in which the introduction of ML introduces strong, unwanted statistical biases. For one, when ML models are used to emulate physical (or first-principles) simulations, they amplify confirmation biases. For another, when expressive regressions are used to label datasets, those labels cannot be used in downstream joint or ensemble analyses without taking on uncontrolled biases. The question in the title is being asked of all of the natural sciences; that is, we are calling on the scientific communities to take a step back and consider the role and value of ML in their fields; the (partial) answers we give here come from the particular perspective of physics.

stat.ML↗

arXiv2024

NotPlaNET: Removing False Positives from Planet Hunters TESS with Machine Learning

Differentiating between real transit events and false positive signals in photometric time series data is a bottleneck in the identification of transiting exoplanets, particularly long-period planets. This differentiation typically requires visual inspection of a large number of transit-like signals to rule out instrumental and astrophysical false positives that mimic planetary transit signals. We build a one-dimensional convolutional neural network (CNN) to separate eclipsing binaries and other false positives from potential planet candidates, reducing the number of light curves that require human vetting. Our CNN is trained using the TESS light curves that were identified by Planet Hunters citizen scientists as likely containing a transit. We also include the background flux and centroid information. The light curves are visually inspected and labeled by project scientists and are minimally pre-processed, with only normalization and data augmentation taking place before training. The median percentage of contaminants flagged across the test sectors is 18% with a maximum of 37% and a minimum of 10%. Our model keeps 100% of the planets for 16 of the 18 test sectors, while incorrectly flagging one planet candidate (0.3%) for one sector and two (0.6%) for the remaining sector. Our method shows potential to reduce the number of light curves requiring manual vetting by up to a third with minimal misclassification of planet candidates.

astro-ph.EP↗

arXiv2024

Planet Hunters TESS V: a planetary system around a binary star, including a mini-Neptune in the habitable zone

We report on the discovery and validation of a transiting long-period mini-Neptune orbiting a bright (V = 9.0 mag) G dwarf (TOI 4633; R = 1.05 RSun, M = 1.10 MSun). The planet was identified in data from the Transiting Exoplanet Survey Satellite by citizen scientists taking part in the Planet Hunters TESS project. Modeling of the transit events yields an orbital period of 271.9445 +/- 0.0040 days and radius of 3.2 +/- 0.20 REarth. The Earth-like orbital period and an incident flux of 1.56 +/- 0.2 places it in the optimistic habitable zone around the star. Doppler spectroscopy of the system allowed us to place an upper mass limit on the transiting planet and revealed a non-transiting planet candidate in the system with a period of 34.15 +/- 0.15 days. Furthermore, the combination of archival data dating back to 1905 with new high angular resolution imaging revealed a stellar companion orbiting the primary star with an orbital period of around 230 years and an eccentricity of about 0.9. The long period of the transiting planet, combined with the high eccentricity and close approach of the companion star makes this a valuable system for testing the formation and stability of planets in binary systems.

astro-ph.EP↗

arXiv2024

Quaia, the Gaia-unWISE Quasar Catalog: An All-Sky Spectroscopic Quasar Sample

We present a new, all-sky quasar catalog, Quaia, that samples the largest comoving volume of any existing spectroscopic quasar sample. The catalog draws on the 6,649,162 quasar candidates identified by the Gaia mission that have redshift estimates from the space observatory's low-resolution BP/RP spectra. This initial sample is highly homogeneous and complete, but has low purity, and 18% of even the bright ($G<20.0$) confirmed quasars have discrepant redshift estimates ($|Δz/(1+z)|>0.2$) compared to those from the Sloan Digital Sky Survey (SDSS). In this work, we combine the Gaia candidates with unWISE infrared data (based on the Wide-field Infrared Survey Explorer survey) to construct a catalog useful for cosmological and astrophysical quasar studies. We apply cuts based on proper motions and Gaia and unWISE colors, reducing the number of contaminants by $\sim$4$\times$. We improve the redshifts by training a $k$-nearest neighbors model on SDSS redshifts, and achieve estimates on the $G<20.0$ sample with only 6% (10%) catastrophic errors with $|Δz/(1+z)|>0.2$ ($0.1$), a reduction of $\sim$3$\times$ ($\sim$2$\times$) compared to the Gaia redshifts. The final catalog has 1,295,502 quasars with $G<20.5$, and 755,850 candidates in an even cleaner $G<20.0$ sample, with accompanying rigorous selection function models. We compare Quaia to existing quasar catalogs, showing that its large effective volume makes it a highly competitive sample for cosmological large-scale structure analyses. The catalog is publicly available at https://zenodo.org/records/10403370.

astro-ph.GA↗

arXiv2024

Frizzle: Combining spectra or images by forward modeling

When there are many observations of an astronomical source - many images with different dithers, or many spectra taken at different barycentric velocities - it is standard practice to shift and stack the data, to (for example) make a high signal-to-noise average image or mean spectrum. Bound-saturating measurements are made by manipulating a likelihood function, where the data are treated as fixed, and model parameters are modified to fit the data. Traditional shifting and stacking of data can be converted into a model-fitting procedure, such that the data are not modified, and yet the output is the shift-adjusted mean. The key component of this conversion is a spectral model that is completely flexible but also a continuous function of wavelength (or position in the case of imaging) that can represent any signal being measured by the device after any reasonable translation (or rotation or field distortion). The benefits of a modeling approach are myriad: The sacred data never are modified. Noise maps, data gaps, and bad-data masks don't require interpolation. The output can take the form of an image or spectrum evaluated on a pixel grid, as is traditional. In addition to shifts, the model can account for line-spread or point-spread function variations, world-coordinate-system variations, and calibration or normalization variations. The noise in the output becomes uncorrelated across neighboring pixels as the shifts deliver good coverage in some sense. The only cost is a small increase in computational complexity over that of traditional methods. We demonstrate the method with a small data example and we provide open source sample code for re-use.

astro-ph.IM↗

arXiv2024

Data-driven Dynamics with Orbital Torus Imaging: A Flexible Model of the Vertical Phase Space of the Galaxy

The vertical kinematics of stars near the Sun can be used to measure the total mass distribution near the Galactic disk and to study out-of-equilibrium dynamics. With contemporary stellar surveys, the tracers of vertical dynamics are so numerous and so well measured that the shapes of underlying orbits are almost directly visible in the data through element abundances or even stellar density. These orbits can be used to infer a mass model for the Milky Way, enabling constraints on the dark matter distribution in the inner galaxy. Here we present a flexible model for foliating the vertical position-velocity phase space with orbits, for use in data-driven studies of dynamics. The vertical acceleration profile in the vicinity of the disk, along with the orbital actions, angles, and frequencies for individual stars, can all be derived from that orbit foliation. We show that this framework - "Orbital Torus Imaging" (OTI) - is rigorously justified in the context of dynamical theory, and does a good job of fitting orbits to simulated stellar abundance data with varying degrees of realism. OTI (1) does not require a global model for the Milky Way mass distribution, and (2) does not require detailed modeling of the selection function of the input survey data. We discuss the approximations and limitations of the OTI framework, which currently trades dynamical interpretability for flexibility in representing the data in some regimes, and which also presently separates the vertical and radial dynamics. We release an open-source tool, torusimaging, to accompany this article.

astro-ph.GA↗

arXiv2023

Orbital Torus Imaging: Acceleration, density, and dark matter in the Galactic disk measured with element abundance gradients

Under the assumption of a simple and time-invariant gravitational potential, many Galactic dynamics techniques infer the Milky Way's mass and dark matter distribution from stellar kinematic observations. These methods typically rely on parameterized potential models of the Galaxy and must take into account non-trivial survey selection effects, because they make use of the density of stars in phase space. Large-scale spectroscopic surveys now supply information beyond kinematics in the form of precise stellar label measurements (especially element abundances). These element abundances are known to correlate with orbital actions or other dynamical invariants. Here, we use the Orbital Torus Imaging (OTI) framework that uses abundance gradients in phase space to map orbits. In many cases these gradients can be measured without detailed knowledge of the selection function. We use stellar surface abundances from the APOGEE survey combined with kinematic data from the Gaia mission. Our method reveals the vertical ($z$-direction) orbit structure in the Galaxy and enables empirical measurements of the vertical acceleration field and orbital frequencies in the disk. From these measurements, we infer the total surface mass density, $Σ$, and midplane volume density, $ρ_0$, as a function of Galactocentric radius and height. Around the Sun, we find $Σ_{\odot}(z=1.1$ kpc)$=72^{+6}_{-9}$M$_{\odot}$pc$^{-2}$ and $ρ_{\odot}(z=0)=0.081^{+0.015}_{-0.009}$ M$_{\odot}$pc$^{-3}$ using the most constraining abundance ratio, [Mg/Fe]. This corresponds to a dark matter contribution in surface density of $Σ_{\odot,\mathrm{DM}}(z=1.1$ kpc)$=24\pm4$ M$_{\odot}$pc$^{-2}$, and in total volume mass density of $ρ_{\odot,\mathrm{DM}}(z=0)=0.011\pm0.002$ M$_{\odot}$pc$^{-3}$. Moreover, using these mass density values we estimate the scale length of the low-$α$ disc to be $h_R=2.24\pm0.06$kpc.

astro-ph.GA↗

arXiv2023

KPM: A Flexible and Data-Driven K-Process Model for Nucleosynthesis

The element abundance pattern found in Milky Way disk stars is close to two-dimensional, dominated by production from one prompt process and one delayed process. This simplicity is remarkable, since the elements are produced by a multitude of nucleosynthesis mechanisms operating in stars with a wide range of progenitor masses. We fit the abundances of 14 elements for 48,659 red-giant stars from APOGEE DR17 using a flexible, data-driven K-process model -- dubbed KPM. In our fiducial model, with $K=2$, each abundance in each star is described as the sum of a prompt and a delayed process contribution. We find that KPM with $K=2$ is able to explain the abundances well, recover the observed abundance bimodality, and detect the bimodality over a greater range in metallicity than previously has been possible. We compare to prior work by Weinberg et al. (2022), finding that KPM produces similar results, but that KPM better predicts stellar abundances, especially for elements C+N and Mn and for stars at super-solar metallicities. The model fixes the relative contribution of the prompt and delayed process to two elements to break degeneracies and improve interpretability; we find that some of the nucleosynthetic implications are dependent upon these detailed choices. We find that moving to four processes adds flexibility and improves the model's ability to predict the stellar abundances, but doesn't qualitatively change the story. The results of KPM will help us to interpret and constrain the formation of the Galaxy disk, the relationship between abundances and ages, and the physics of nucleosynthesis.

astro-ph.GA↗

arXiv2023

AspGap: Augmented Stellar Parameters and Abundances for 23 million RGB stars from Gaia XP low-resolution spectra

We present AspGap, a new approach to infer stellar labels from low-resolution Gaia XP spectra, including precise [$α$/M] estimates for the first time. AspGap is a neural-network based regression model trained on APOGEE spectra. In the training step, AspGap learns to use XP spectra not only to predict stellar labels but also the high-resolution APOGEE spectra that lead to the same stellar labels. The inclusion of this last model component -- dubbed the hallucinator -- creates a more physically motivated mapping and significantly improves the prediction of stellar labels in the validation, particularly of [$α$/M]. For giant stars, we find cross-validated rms accuracies for Teff, log g, [M/H], [$α$/M] of ~1%, 0.12 dex, 0.07 dex, 0.03 dex, respectively. We also validate our labels through comparison with external datasets and through a range of astrophysical tests that demonstrate that we are indeed determining [$α$/M] from the XP spectra, rather than just inferring it indirectly from correlations with other labels. We publicly release the AspGap codebase, along with our stellar parameter catalog for all giants observed by Gaia XP. AspGap enables new insights into the formation and chemo-dynamics of our Galaxy by providing precise [$α$/M] estimates for 23 million giant stars, including 12 million with radial velocities from Gaia.

astro-ph.SR↗

arXiv2023

The Eighteenth Data Release of the Sloan Digital Sky Surveys: Targeting and First Spectra from SDSS-V

The eighteenth data release of the Sloan Digital Sky Surveys (SDSS) is the first one for SDSS-V, the fifth generation of the survey. SDSS-V comprises three primary scientific programs, or "Mappers": Milky Way Mapper (MWM), Black Hole Mapper (BHM), and Local Volume Mapper (LVM). This data release contains extensive targeting information for the two multi-object spectroscopy programs (MWM and BHM), including input catalogs and selection functions for their numerous scientific objectives. We describe the production of the targeting databases and their calibration- and scientifically-focused components. DR18 also includes ~25,000 new SDSS spectra and supplemental information for X-ray sources identified by eROSITA in its eFEDS field. We present updates to some of the SDSS software pipelines and preview changes anticipated for DR19. We also describe three value-added catalogs (VACs) based on SDSS-IV data that have been published since DR17, and one VAC based on the SDSS-V data in the eFEDS field.

astro-ph.GA↗

arXiv2023

Constraining cosmology with the Gaia-unWISE Quasar Catalog and CMB lensing: structure growth

We study the angular clustering of Quaia, a Gaia- and unWISE-based catalog of over a million quasars with an exceptionally well-defined selection function. With it, we derive cosmology constraints from the amplitude and growth of structure across cosmic time. We divide the sample into two redshift bins, centered at $z=1.0$ and $z=2.1$, and measure both overdensity auto-correlations and cross-correlations with maps of the Cosmic Microwave Background convergence measured by Planck. From these data, and including a prior from measurements of the baryon acoustic oscillations scale, we place constraints on the amplitude of the matter power spectrum $σ_8=0.766\pm 0.034$, and on the matter density parameter $Ω_m=0.343^{+0.017}_{-0.019}$. These measurements are in reasonable agreement with \planck at the $\sim$ 1.4$σ$ level, and are found to be robust with respect to observational and theoretical uncertainties. We find that our slightly lower value of $σ_8$ is driven by the higher-redshift sample, which favours a low amplitude of matter fluctuations. We present plausible arguments showing that this could be driven by contamination of the CMB lensing map by high-redshift extragalactic foregrounds, which should also affect other cross-correlations with tracers of large-scale structure beyond $z\sim1.5$. Our constraints are competitive with those from state-of-the-art 3$\times$2-point analyses, but arise from a range of scales and redshifts that is highly complementary to those covered by cosmic shear data and most galaxy clustering samples. This, coupled with the unprecedented combination of volume and redshift precision achieved by Quaia allows us to break the usual degeneracy between $Ω_m$ and $σ_8$.

astro-ph.CO↗

arXiv2023

Towards fully covariant machine learning

Any representation of data involves arbitrary investigator choices. Because those choices are external to the data-generating process, each choice leads to an exact symmetry, corresponding to the group of transformations that takes one possible representation to another. These are the passive symmetries; they include coordinate freedom, gauge symmetry, and units covariance, all of which have led to important results in physics. In machine learning, the most visible passive symmetry is the relabeling or permutation symmetry of graphs. Our goal is to understand the implications for machine learning of the many passive symmetries in play. We discuss dos and don'ts for machine learning practice if passive symmetries are to be respected. We discuss links to causal modeling, and argue that the implementation of passive symmetries is particularly valuable when the goal of the learning problem is to generalize out of sample. This paper is conceptual: It translates among the languages of physics, mathematics, and machine-learning. We believe that consideration and implementation of passive symmetries might help machine learning in the same ways that it transformed physics in the twentieth century.

stat.ML↗

arXiv2023

The Panchromatic Hubble Andromeda Treasury XX: The Disk of M31 is Thick

We present a new approach to measuring the thickness of a partially face-on stellar disk, using dust geometry. In a moderately-inclined disk galaxy, the fraction of reddened stars is expected to be 50% everywhere, assuming that dust lies in a thin midplane. In a thickened disk, however, a wide range of radii project onto the line of sight. Assuming stellar density declines with radius, this geometrical projection leads to differences in the numbers of stars on the near and far sides of the thin dust layer. The fraction of reddened stars will thus differ from the 50% prediction, with a deviation that becomes larger for puffier disks. We map the fraction of reddened red giant branch (RGB) stars across M31, which shows prominent dust lanes on only one side of the major axis. The fraction of reddened stars varies systematically from 20% to 80%, which requires that these stars have an exponential scale height h_z that is 0.14+/-0.015 times the exponential scale length (h_r~5.5kpc). M31's RGB stars must therefore have h_z=770+/-80pc, which is far thicker than the Milky Way's thin disk, but comparable to its thick disk. The lack of a significant thin disk in M31 is unexpected, but consistent with its interaction history and high disk velocity dispersion. We suggest that asymmetric reddening be used as a generic criteria for identifying ``thick disk'' dominated systems, and discuss prospects for future 3-dimensional tomographic mapping of the gas and stars in M31.

astro-ph.GA↗

arXiv2023

Scalars are universal: Equivariant machine learning, structured like classical physics

There has been enormous progress in the last few years in designing neural networks that respect the fundamental symmetries and coordinate freedoms of physical law. Some of these frameworks make use of irreducible representations, some make use of high-order tensor objects, and some apply symmetry-enforcing constraints. Different physical laws obey different combinations of fundamental symmetries, but a large fraction (possibly all) of classical physics is equivariant to translation, rotation, reflection (parity), boost (relativity), and permutations. Here we show that it is simple to parameterize universally approximating polynomial functions that are equivariant under these symmetries, or under the Euclidean, Lorentz, and Poincaré groups, at any dimensionality $d$. The key observation is that nonlinear O($d$)-equivariant (and related-group-equivariant) functions can be universally expressed in terms of a lightweight collection of scalars -- scalar products and scalar contractions of the scalar, vector, and tensor inputs. We complement our theory with numerical examples that show that the scalar-based method is simple, efficient, and scalable.

cs.LG↗