Eccentricities of Close Stellar Binaries
Extracted source text · arXiv 2411.09905
\documentclass[twocolumn,twocolappendix]{aastex631}
\usepackage[utf8]{inputenc}
\usepackage{amsmath, mathtools}
\usepackage{graphicx}
%\usepackage{caption}
%\usepackage{lineno}
%\usepackage{subcaption}
\usepackage{wrapfig}
\usepackage{float}
\usepackage{natbib,color,ulem,hyperref}
\makeatletter
\renewcommand\paragraph{\@startsection{paragraph}{4}{\z@}%
{-2.5ex\@plus -1ex \@minus -.25ex}%
{1.25ex \@plus .25ex}%
{\normalfont\normalsize}}
\makeatother
\setcounter{secnumdepth}{4}
\setcounter{tocdepth}{4}
%\usepackage{graphicx}%
%\usepackage{multirow}%
%\usepackage{amsmath,amssymb,amsfonts}%
%\usepackage{amsthm}%
%\usepackage{mathrsfs}%
%\usepackage[title]{appendix}%
%\usepackage{xcolor}%
%\usepackage{textcomp}%
%\usepackage{manyfoot}%
%\usepackage{booktabs}%
%\usepackage{algorithm}%
%\usepackage{algorithmicx}%
%\usepackage{algpseudocode}%
%\usepackage{listings}%
%%%%
\newcommand{\y}{\color{red}}
\newcommand{\z}{\color{orange}}
%\newcommand{\w}{\color{blue}}
%\newcommand{\w}{\color{blue}}
\newcommand{\w}[1]{#1} % To toggle off \uu
%\newcommand{\w}{#1}
\newcommand{\md}{\color{brown}}
\newcommand{\cdm}{\color{magenta}}
\newcommand{\jwd}{\color{teal}}
\def\pomega{\varpi}
\def\ii{{\rm i}}
\def\cc{{\rm c.c.}}
\newcommand{\be}{\begin{eqnarray}}
\newcommand{\ee}{\end{eqnarray}}
\newcommand{\beqn}{\begin{eqnarray}}
\newcommand{\eeqn}{\end{eqnarray}}
\newcommand{\bi}{\begin{itemize}}
\newcommand{\ei}{\end{itemize}}
\newcommand{\dt}{{\partial_t}}
\newcommand{\dd}{{\partial}}
\newcommand{\arr}[2]{\left(\begin{array}{c} #1 \\ #2 \end{array}\right)}
\def\refnew#1{(\ref{#1})}
\def\erg{\, \rm erg}
\def\g{\, \rm g}
\def\K{\, \rm K}
\def\yr{\, \rm yr}
\def\s{\, \rm s}
\def\km{\, \rm km}
\def\cm{\, \rm cm}
\def\m{\, \rm m}
\def\dyne{\, \rm dyne}
%these should be removed
\def\apj{\, ApJ}
\def\araa{\, ARA\&A}
\def\apjs{\, ApJS}
\def\aap{\, A\&Ap}
\def\mnras{\, MNRAS}
\def\aj{\, AJ}
\def\apjl{\, ApJL}
\def\nat{\, Nature}
%
\def\pomega{\varpi}
\def\ii{{\rm i}}
\def\cc{{\rm c.c.}}
\def\rth{R_{\rm tran}}
\def\rtran{{R_{\rm tran}}}
\def\rst{R_{\rm stir}}
\def\icarus{{Icarus}}
\usepackage{ulem}
\newcommand{\x}{\sout}
%\newcommand{\x}{\phantom}
%\newcommand{\x}{\color{purple}}
\shorttitle{Binary Eccentricities}
\shortauthors{Wu et al.}
\begin{document}
\title[Why are Stellar Binaries Eccentric?]{Eccentricities of Close Stellar Binaries}
\author[0000-0003-0511-0893]{Yanqin Wu}
\affiliation{Department of Astronomy \& Astrophysics, University of Toronto, 50 St George Street, Toronto, ON M5S\,3H4, Canada}
\author[0000-0002-1032-0783]{Sam Hadden}
\affiliation{Canadian Institute of Theoretical Astrophysics, University of Toronto, 50 St George Street, Toronto, On M5S\,3H4, Canada}
\author[0000-0001-9420-5194]{Janosz Dewberry}
\affiliation{Canadian Institute of Theoretical Astrophysics, University of Toronto, 50 St George Street, Toronto, On M5S\,3H4, Canada}
\author[0000-0002-6871-1752]{Kareem El-Badry}
\affiliation{Department of Astronomy, California Institute of Technology, 1200 E. California Blvd., Pasadena, CA 91125, USA}
\author[0000-0001-9732-2281]{Christopher D. Matzner}
\affiliation{Department of Astronomy \& Astrophysics, University of Toronto, 50 St George Street, Toronto, ON M5S\,3H4, Canada}
%\linenumbers
%\author*[1]{\fnm{Yanqin} \sur{Wu}}\email{wu@astro.utoronto.ca}
%\author[2]{\fnm{Sam} \sur{Hadden}}\email{hadden@cita.utoronto.ca}
%\equalcont{These authors contributed equally to this work.}
%\author[2]{\fnm{Janosz} \sur{Dewberry}}\email{jdewberry@cita.utoronto.ca}
%\equalcont{These authors contributed equally to this work.}
%\author[3]{\fnm{Kareem} \sur{El-Badry}}\email{kelbadry@caltech.edu}
%\equalcont{These authors contributed equally to this work.}
%\author[1]{\fnm{Christopher D.}\sur{Matzner}}\email{matzner@astro.utoronto.ca}
%\affil*[1]{\orgdiv{Department of Astronomy \& Astrophysics}, \orgname{University of Toronto}, \orgaddress{\street{50 St. George Street}, \city{Toronto}, \postcode{M5S 3H4}, \state{ON}, \country{Canada}}}
%\affil[2]{\orgdiv{Canadian Institute of Theoretical Astrophysics}, \orgname{University of Toronto}, \orgaddress{\street{60 St. George Street}, \city{Toronto}, \postcode{M5S 3H4}, \state{ON}, \country{Canada}}}
%\affil[3]{\orgdiv{Department of Astronomy}, \orgname{California Institute of Technology}, \orgaddress{\street{1200 E. California Blvd.}, \city{Pasadena}, \postcode{91125}, \state{CA}, \country{USA}}}
\begin{abstract}
Orbits of stellar binaries are in general eccentric. {\w These eccentricities encode information about their early lives.}
%Past studies using dozens of spectroscopic binaries have concluded that the eccentricities are uniformly distributed, outside the realm of tidal circularization.
Here, we use thousands of
%$\sim 100,000$
main-sequence binaries from the Gaia DR3 catalog to reveal that, binaries inwards of a few AU exhibit a simple Rayleigh distribution with a mode $\sigma_e \simeq 0.3$. We find the same distribution for binaries from M to A spectral types, and from tens of days to $10^3$days (possibly extending to tens of AU).
%likely encompassing all binaries that will interact
%, through tides or mass transfer or common envelope, during their lifetimes.
This observed distribution is most likely primordial {\w and its invariance suggests a single universal process. One possibility is eccentricity excitation by circumbinary disks. Another, as is suggested by the Rayleigh form, is weak scattering and ejection of brown-dwarf objects. We explore this latter scenario and find that the binary eccentricities reach an equi-partition value of $\sigma_e \simeq \sqrt{M_{\rm bd}/M_*}$. %independent of orbital period.
%largely independent of orbital periods or binary masses.
So to explain the observed mode, the brown dwarfs will have to be of order one tenth the stellar masses, and be at least as abundant in the Galaxy as the close binaries. The veracity of {\w both} proposals remains to be tested. }
%We hypothesize that when these binaries are formed, likely in gravitationally unstable massive disks, they are accompanied by a large number of brown-dwarf massed objects. As the binary eject the brown-dwarfs, their orbits shrink,
%Such a scenario probes an epoch of binary formation that is as yet un-reachable by theory and direct observations. Ejection by stellar binaries may explain the so-called 'brown-dwarf desert', and may account for the bulk of the free-floating sub-stellar objects.
%Such a process explains a few observed properties of the Rayleigh distribution: invariant with respect to primary or secondary mass; invariant within the observed range of periods ($\sim 10^2 - 10^3$ days);
\end{abstract}
\keywords{stars, binaries, orbits, dynamics, formation, proto-stellar disks, brown dwarfs}
%%\pacs[JEL Classification]{D8, H51}
%%\pacs[MSC Classification]{35A01, 65L10, 65L12, 65L20, 65L70}
%\maketitle
\section{Introduction}\label{sec1}
Binary stars are common in the Galaxy \citep{DM91,Fischer1992}.
%DM91 is solar, Fischer is M-dwarf, both find 50%
Their orbits are invariably eccentric. The distribution of these eccentricities encodes accessible information about their formation. {\w It is also used in} a wide range of studies.
Surprisingly, for binaries closer than tens of AU (`close binaries'), such a fundamental property %\x{is not well known}
{\w has not been definitely determined}. %Previous best guess, based on a dozen nearby binary systems, was ambiguous and, as we show in this work, incorrect.
%Goodwin46 (a is the semimajor axis in AU):Close binaries: a < 50 Intermediate binaries: 50 < a < 1000 Wide binaries: 1000 < a < 10000 Very wide binaries: a > 10000
Multiple lines of evidence suggest two main modes of binary formation {\w exist} \citep[see][for reviews]{Duchene2013,offner2023}. Wide binaries \citep[$> 50$ AU,\footnote{Interestingly, $50$AU, or $P\sim 10^5$days, is roughly the peak of the binary distribution for Sun-like stars \citep{DM91,Raghavan2010}.} see, e.g.,][]{Parker2009}
%, adopting the definition in][]{Parker2009},
likely form following direct collapse of individual components. These are either weakly bound at birth or are captured after birth. Their occurrence is insensitive to or rises with stellar metallicity \citep{elbadry_rix_19,Hwang2021},
%Hwang2021 inconsistent w/ Elbadry, 10^3-10^4AU
% below 100AU, the binary fraction is anti-correlated with the metallicity \cite{Raghavan2010,Moe2019}
%Badenes et al. 2018; Mazzola et al. 2020). Such anti-correlation with metallicity disappear at $\sim 200$AU (El-Badry \& Rix 2019), and the wide binary fraction at $10^3-10^4$ AU has a non-monotonic relation with the metallicity (Hwang et al. 2021)."
and the two components appear to be randomly paired in mass \citep{Moe2017}.
%Mass ratio distribution differs significantly between short- andlong-period binaries: the short period has a peak at twin binaries with a decaying $f(q)$; long periods peak at $q=0.3$
%M67/NGC188/field (out to $10^4$ day): 'consistent' with a uniform q;
Close binaries, on the other hand, are now thought to have formed through gravitational fragmentation in massive circum-stellar disks \citep[see review by][]{kratterlodato16}. Their occurrence anti-correlates with stellar metallicity \citep{Moe2019,Mazzola2020}, and the two masses are correlated \citep{Raghavan2010,Duchene2013,Moe2017} with an excess of equal-mass `twin-binaries' \citep[][]{Moe2017,Elbadrytwin}.
%\item twin binaries: \cite{Moe2017} reviewed twin-binaries:more common around solar-type binaries, and its occurrence reaches $30\%$ at $10$d, but drops to $10\%$ by $10^5$d. Twin binaries do not exhibit astrometric signal and may be selected against in the GAIA catalogue.
%\cite{Tokovinin2000} reported only excess is shorter than 40 days?-- selection bias
%\item period distribution: also a break around 50au? Period distribution: The maximum of the orbital period distribution occurs at $P \sim 250$ yr ($a \sim 45$ AU), with a dispersion of$\sigma_{\log P} \sim 2.3$ \cite{Raghavan2010}.
%
The different formation pathways also manifest in binary eccentricities -- a property that is observationally accessible and dynamically informative. Wide binaries are observationally determined to have a thermal distribution ($dN/de \propto e$), as inferred {\w using} spectroscopic
%binaries
\citep{DM91,Raghavan2010} and visual binaries \citep{Tokovinin2020,Hwang2022}.\footnote{Very wide binaries ($>10^3$ AU) also appear to be super-thermal \citep{Tokovinin2020,Hwang2022}, suggesting another mechanism at play.} This reflects a dynamic past in the birth clusters \citep{Parker2009}, where
%This is as expected. Wide binaries, however they formed, may have been processed by star scatterings in their birth clusters \cite{Parker2009}.
plentiful scatterings with other stars have relaxed the binaries towards a `thermal' equilibrium \citep{Jeans1919,Heggie1975}.
%(Jeans 1919, Ambartsumian 1937, Heggie 1975, Kroupa 2008?)
%For binaries at $> 1000$ AU, their eccentricity distribution is close to thermal (fe(e)de = 2ede), a theoretical distribution when a population of binaries reaches an equilibrium state Ambartsumian 1937; Heggie 1975; Kroupa 2008).
%Jeans (1919) a population of binaries reaches a thermal equilibrium where the energy distribution follows a Boltzmann distribution, their eccentricity distribution is $f_e(e)de = 2ede$, independent of binary separations. Ambartsumian (1937) proved that the eccentricity distribution is $f_e(e)de = 2ede$ when the distribution function only depends on the energy,
%
%Tokovinin (2020) reports that wide binaries at $10^3-10^4$ AU have a nearly thermal eccentricity distribution, and those at $< 200$AU have less eccentric orbits.
% Figure 15 of Raghavan et al. (2010) gives f(e) for visual binaries that is approximately flat be- tween 0.1 and 0.6 and drops at large and small e.
%
%Geller+2019 gives a nice intro to the thermal.vs.uniform debate
%{\cite{Geller2019}: clusters can't turn uniform to thermal.}
%\cite{Kouwenhoven2010} explicitly finds thermal e-dist, $>10^3$au binaries;
%\cite{Mathew2024} explicitly find thermal to super-thermal, depending on turbulent properties (with sink radius 250au, mostly binaries > 100au)
%Geller+2019: for nearly all binaries, dynamical encounters fail to convert an initially uniform eccentricity distribution to thermal within a star cluster’s lifetime.
%(e.g. Bate and Bonnell, 1997; Bate, 2009; Marks and Kroupa, 2012; Bate, 2019; Rozner and Perets, 2023; Guszejnov et al., 2023).
%\cite{Mathew2024} explained the thermal to super-thermal in turbulent star formation
%\cite{Reipurth2012}:
%in turbulent medium (Bate 2009, 2014; Offner et al. 2010), from capture during cluster dissolution \cite{Kouwenhoven2010,Moeckel2011}, from dynamical unfolding of unstable multiples (Reipurth \& Mikkola 2012, Bate et al 2002), and random pairings of pre-stellar cores (Tokovinin 2017).
%We expect disc fragmentation (e.g. Adams, Ruden \& Shu 1989 ; Shu et al. 1990 ; Bonnell \& Bate 1994 ; Stamatellos \& Whitworth 2009 ; Stamatellos et al. 2011 ) to be an important binary formation mechanism (see also Hennebelle, Lee \& Chabrier 2019 ) on these scales, Kratter \& Matzner 2006; \cite{Moe2019} Tokovinin \& Moe 2020).
%
%\cite{Tokovinin2020} said:" pairs that remain after dynamical interaction of a binary or a triple system with another star (scattering) have a nearly thermal f(e) (e.g. Fig. 16 in Antognini & Thompson 2016).
% wrong -- those were triples
%
%When two binaries interact dynamically in a gas cloud and survive as a 2+2 quadruple, the eccentricity of the wider remaining pair has a super-thermal distribution, while for the closer pair the distribution is thermal (Fig. 9 of Ryu et al. 2017). Disintegration of unstable triple systems leaves binaries with a mildly super-thermal f(e) (Stone \& Leigh 2019). However, wide binaries formed by ejection from unstable triples or “unfolding” (Reipurth \& Mikkola 2012) must have only very eccentric orbits. The unfolding mechanism does not match certain properties of real wide binaries and appears to be exceptional rather than typical (Tokovinin 2017)."
Such a concordance between theory and observation does not, however, extend to close binaries. There are no theoretical predictions for their e-distribution, owing to the uncertain fragmentation process. On the observational side, their e-distribution remains murky, though it clearly differs from that of the wide binaries.
The current wisdom is that they are consistent with being ``uniform'',\footnote{Binaries with massive primaries (O/B stars) appear `thermal', down to periods as short as tens of days, possibly related to their higher triple fraction
\citep{Moe2017}.
} $dN/de \propto e^0$ \citep{Tokovinin2000,Raghavan2010,Duchene2013,Moe2017,Geller2021,Hwang2022}.
%But this is only a best-guess estimate that is based on small samples\footnote{Some studies have suggested a more 'bell-shaped' e-distribution \citep{DM91,Geller2012}, but \citet{Moe2017} concluded that the data are too sparse to tell. } and has no ready theoretical explanation.
{\w This conclusion, besides from having no ready theoretical explanation, is drawn from limited samples.} For instance, if we restrict to the period range $10^2-10^3$d (see below), %so as to exclude binaries that may have experienced tidal circularization or may have been disturbed by passing-by stars,
we are left with only $18$ binaries in the volume-complete (to 25\,pc), Sun-like sample from \citet{Raghavan2010}.
%%,Moe2017}
{\w Larger samples do exist. The survey of nearby G-dwarfs by \citet{Udry1998} returns $N=110$ such systems; the massive compendium of all spectroscopic binaries, SB9 \citep{SB9}, contains nearly a thousand similar systems;
%\footnote{\w A fraction of these have unspecified eccentricities.}
and the APOGEE survey constrained eccentricities for $\sim 8600$ FGK binaries beyond $100$ days \citep{PriceWhelan2020}. However, although results from these surveys hint at an eccentricity distribution that is not ``uniform'', there have been no firm conclusions, largely because the selection bias on eccentricity has not been properly modeled. }
% The review by \cite{Duchene2013} further concludes that for binaries with periods of $10^2-10^4$ days (roughly 0.5-10AU) and different masses, the eccentricity distributions are more consistent with a uniform distribution compared to the thermal distribution. Moe \& Di Stefano (2017) report a similar finding of uniform eccentricity distributions for primary masses of $0.8-5 M_\odot$ and periods $< 10^5$ day (50 AU), except that binaries with primary masses $> 5M_\odot$ may have a thermal eccentricity distribution at periods $> 10$ day. Therefore, these studies all show that the eccentricity distribution of binaries at $< 100$ AU is consistent with a uniform distribution.
%
%Geller2021 is in M67, looks gaussian but they completeness-corrected it to uniform
%For instance, selecting from the volume-complete (to 25pc) Solar-type sample \citep{Raghavan2010,Moe2017}, for
%contains only $\sim 90$ close binaries. If we focus only on the
%binaries with periods $10^2-10^3$d, so as to exclude binaries that may have experienced tidal circularization or may have been disturbed by passing-by stars, we are left with only $18$ binaries.
% < 10^5 days, Fig. 32 of Moe
%Moe: the corrected solar-type sample contains Nprim = 404 primaries and Ncomp = 193 with 0.1<q<1, and 0 < p < 8
%encompass a broad range of primary spectral types, and
{\w It is clear that a binary sample, with a well calibrated completeness in eccentricity,} is sorely needed.
%Using a sparse sample of 164 binaries, \cite{DM91} concluded that the eccentricity distribution is 'bell-shaped' inward of $10^3$ days. The conclusion was changed by a slightly larger sample, where binaries inward of $100$au are measured to be uniform \cite{Raghavan2010}. Some hints of departure from uniform or thermal distribution is seen in the compiled Fig. 3 of \cite{Duchene2013} for $10^2-10^4$ d binaries.
The Gaia mission \citep{gaia2016,gaia2023}, especially with its most recent non-single-star catalogue published in DR3 \citep{Arenou2023}, provides just this sample. The Gaia binaries, as we show here, sharpen our vision dramatically. The e-distribution for those inward of a few AU follows, distinctly, a Rayleigh distribution. %Many questions and interesting implications follow.
%By using a volume-limited sample within 25pc, Raghavan et al. (2010) show that the eccentricity distribution of solar-type stars is close to uniform for orbital periods between $12$ and $10^6$ day ($\sim 300$ AU), and most binaries have circular orbits below $12$ days due to tidal circularization (Mathieu 1994; Zahn 2008; Price-Whelan \& Goodman 2018).
%We adopt a semi-major axis distribution of $dN/da \propto a^{-1.6} $ (El-Badry \& Rix 2018) and a range from $10$ to $10^5$ au.
\section{GAIA Binaries }\label{sec:observation}
Here we will study Gaia binaries with eccentricities explicitly determined by astrometry and/or radial velocities. %This contrasts with the resolved binaries analyzed by \cite{Tokovinin2020,Hwang2022}.
%We use GAIA's eccentricity for individual un-resolved binaries, obtained using a combination of astrometric and spectroscopic methods.
The details of our binary selection are in Appendix \ref{sec:A1}, and a short form is presented in Table \ref{tab:sample}. The `full' main-sequence binary sample includes some $150,000$ systems. We pare down this large set by different cuts (Table \ref{tab:sample}) and study their respective properties. Among these, we highlight results from the so-called `gold' sample, $\sim 3000$ Sun-like binaries that have periods from $10^2-10^3$ days,
and that lie within $150$\,pc. Binaries with too short a period can be affected by tidal circularization, while DR3 extends to just beyond $10^3$ days; binaries beyond $150$\,pc are {\w increasingly} incomplete {\w in the eccentricity space} %to various degrees
(see Appendix \ref{sec:selection}).
%\footnote{\w The `gold' sample suffers incompleteness in other dimensions. See Appendix \ref{sec:selection}.}
\begin{table*}%[h!]
\centering
\begin{tabular}{|c| c c c c |c|}
\hline
sample name & criterion & sample size & {\w mean distance} & {\w mean e-error ($\bar\epsilon_e$)} & best-fit $\sigma_e$ \\
\hline
\hline
DR3 binaries & - & $362,065$ & $1240$\,pc & $0.068$ & $0.238$ \\
`Full' & main-sequence, significance $>10$ & $147,634$ & $560$\,pc& $0.058$ & $0.240$\\
%main-sequence2 & no `EclipsingBinary' & & \\
`Primordial' & period $P \in [10^2, 10^3]$ d & $103,590$ & $607$\,pc & $0.065$ & $0.256$\\
- & `Orbital' or `AstroSpectroSB1' & $97,458$ & $618$\,pc & $0.065$ & $0.258$\\
- & goodness-of-fit cut (see text) & $90,376$ & $629$\,pc & $0.065$ & $0.257$\\
`Sun-like' & primary $R_* \in [0.7,2.0]R_\odot$ & $81,107$ & $595$\,pc & $0.064$ & $0.258$ \\
`gold' & distance $< 150$\,pc & $3,071$ & $111$pc & $0.028$ & $0.303$\\
\hline
\end{tabular}
\caption{How we pare down the Gaia binary catalog to obtain various samples in this study. The right-most column lists the best-fit Rayleigh modes. {\w These values are reduced from the true one by increasing distance and therefore increasing detection bias. They are, to a lesser degree, inflated by measurement errors (see Appendix \ref{sec:error}).}}
\label{tab:sample}
\end{table*}
\begin{figure*}
\centering
\includegraphics[width=1.0\textwidth]{data1.pdf}
%\includegraphics[width=0.45\textwidth]{data1.eps}
\caption{Eccentricity distribution
for the Gaia `Sun-like' sample (Table \ref{tab:sample}), split by distances, shown in differential (left panel) and cumulative (right panel) forms. The `gold' sample ($<150$\,pc, thick black lines) is least affected by selection %\x{effects}
{\w bias in eccentricity} and is well described by a Rayleigh distribution with a mode $\sigma_e=0.30$ (gray-shaded areas). E-distributions for binaries at larger distances gradually shift to the left, reflecting the increasing severity of selection bias {\w (not corrected here). They also suffer larger measurement uncertainties ($\bar\epsilon_e$, also not corrected here).}
The right panel also presents the %$N=18$
{\w $N=110$}
spectroscopic binaries within the same period range {\w from \citet{Udry1998}.} %\citep{Raghavan2010}.}
The Gaia gold sample is statistically consistent with this set; {\w both}
firmly
reject the `uniform' hypothesis ($dN/de \propto e^0$).
}
\label{fig:data1}
\end{figure*}
\begin{figure*}
\centering
\includegraphics[width=1\textwidth]{ecc-period-mainsequence.pdf}
\caption{Properties of the `full' sample from Table \ref{tab:sample}, split by primary radii (rows) and orbital periods (x-axis). The left panels show the period distributions, indicating different detection methods. The right panels show kernel density estimation in the $e - \log P$ plane. The effects of tidal circularization are seen for binaries short-ward of $\sim 20$\,days. But beyond that, all bins reach the same mean-eccentricity ($\langle e \rangle$, green curves), corresponding to a Rayleigh mode of
$\sigma_e = \langle e \rangle/\sqrt{\pi/2} = 0.25$.
The low-lying blue curves are the mean eccentricity errors (${\bar\epsilon}_e$), with individual uncertainties ranging typically from $0.02-0.05$. }\label{fig:mainsequence}
\end{figure*}
The eccentricity distributions of our binary samples are shown in Fig. \ref{fig:data1}.
They are all well described by a Rayleigh distribution,\footnote{Adopting a Bayesian framework (Appendix \ref{sec:model-compare}), one can formally establish that this single Rayleigh distribution is favoured over {\w a number of} other distributions, and that its mode is strongly constrained.}
\begin{equation}
\frac{dN}{de} = {e\over{\sigma_e^2}} \, \exp\left({- \frac{e^2}{2\sigma_e^2}}\right)\label{eq:rayleigh}\, .
\end{equation}
%We find that typically, $e_0 \sim {\bar\delta}_e$, the median measurement uncertainties in eccentricity. We suggest that the intrinsic distribution is likely Rayleigh (with $e_0 = 0$), but systematic measurement errors brings about a non-zero $e_0$.
For the `gold' sample, we obtain $\sigma_e = 0.303\pm 0.003$, where the uncertainty {\w only} accounts for the sample size. {\w In the mean time, $\sigma_e$ reduces for samples at increasing distances.}
%\x{Now we put our work on some statistical footing.}
{\w In Fig. \ref{fig:data1}, we also compare}
the `gold' sample % \x{statistically}
%\x{with the older, sparse sample \citep[$N=18$ spectroscopic binaries within $10^2-10^3$d in][]{Raghavan2010}, from which the 'uniform' e-distribution was deduced. }
{\w against the sample of spectroscopic G-dwarf binaries from \citet{Udry1998}. The two are highly consistent with a p-value of $0.7$ for a two-sample KS test.
This suggests that the eccentricity bias in the \citet{Udry1998} sample is light enough, one could have arrived at our conclusion using just the RV sample and a quarter century earlier}.\footnote{\w However, a comparison with the SB9 RV sample \citep{SB9} within the same period range returns a p-value of $10^{-10}$. There are more low-e binaries in the SB9 sample, which contains more red giants and likely is more strongly affected by tidal circularization.}
%\footnote{A KS 2-sample test returns $p=0.25$ between this small sample and our GAIA sample. or that this they can be drawn from the same underlying sample. }
%In this work, we have cautiously adopted a small subset of the GAIA binaries (the full main-sequence sample at $\sim 150,000$ strong, Table \ref{tab:sample}).
{\w The reduction of the Rayleigh mode at larger distances is explained in Appendix \ref{sec:selection}.}%For Gaia samples lying at larger distances, $\sigma_e$ reduces gradually
(Figs. \ref{fig:kareem} \& \ref{fig:x_extrapolate}). This results from detection bias in Gaia \citep{elbadry2024}: more eccentric binaries are harder to detect at larger distances, because they spend more time at slowly-moving apoapsis, and their photo-centre shifts are smaller at periapsis. %{\y Kareem: is this the proper reason?}
%\x{Fortunately, as we elaborate in Appendix \ref{sec:selection}, our `gold' sample is {\w ``effectively''} complete {\w in eccentricity}.% to $150$pc. }
{\w Even the 'gold' sample suffers a small degree of erosion. Correcting for incompleteness using the forward-modeling pipeline of \citet{elbadry2024}, we find that the best-fit Rayleigh mode for the 'gold' sample rises from $\sigma_e = 0.303$ to $\sigma_e = 0.336$ (Appendix~\ref{sec:selection}).}
Bearing in mind the {\w above} detection bias, we can employ the $150,000$-strong `full' {\w Gaia} sample to paint a more nuanced picture of the e-distribution. In Fig. \ref{fig:mainsequence}, we split the
`full' sample apart by primary mass (with primary radii from $0.5$ to $3 R_\odot$, corresponding to spectral types A-F-G-K-M) and by orbital periods (from $2$ days to $1200$d). We adopt the mean eccentricity $\langle e \rangle$ as a proxy for the Rayleigh mode, where
$\langle e \rangle = \sigma_e {\sqrt{\pi/2}}$ for a Rayleigh distribution.
%will assume the Rayleigh distribution applies and
We find that every bin, with the exception of those that are closer than $\sim 20$\,d and have therefore undergone various degrees of tidal circularization,
exhibits a similar mean eccentricity of $\langle e \rangle\sim 0.31$, or $\sigma_e \sim 0.25$. This latter value is close to that obtained for the `Sun-like' sample (`gold' but including all distances), $\sigma_e = 0.258$, and is a result of convolution between the true mode and detection bias. Assuming the same detection bias across all bins,\footnote{The real bias may vary from bin to bin. However, even the most biased sample in Fig. \ref{fig:data1} (those outside 1200\,pc) still returns a $\sigma_e = 0.225$, or a value lower by $\sim 25\%$ from $\sigma_e = 0.30$. Such a relative immunity from selection bias is offered by the fact that a Rayleigh distribution contains mostly low-e systems that are less vulnerable to incompleteness.} this exercise suggest that the intrinsic Raleigh mode is invariant, across a wide range of primary spectral types, and over a large span in orbital periods. Such an invariance is remarkable and points to a universal process at work.
{\w We also investigate whether the Rayleigh mode depends on the mass ratio within a binary. We fail to reach a definitive conclusion based on current Gaia data (Appendix \ref{sec:massratio}, Fig. \ref{fig:massratio}).}
%\x{Along with this general invariance, we observe a {\w tentative } hint that the Rayleigh mode may vary with the binary mass-ratio
{\w On the other hand, a piece of oblique evidence suggests that it does not: twin binaries are abundant among spectroscopic binaries \citep[see, e.g.][]{Tokovinin2000} but are largely absent in Gaia as their photo-centres hardly shift; so the agreement between RV and Gaia samples (Fig. \ref{fig:data1}) suggests that the eccentricity distribution is largely independent of the binary mass ratio.}
%The 'uniform' guess is the simplest guess based on a small sample, but it is wrong.
%\x{Along with this general invariance, we observe a hint that the Rayleigh mode may decrease with the binary mass-ratio}
{\w Such an independence may even extend to the planetary regime. Fig.\,\ref{fig:all_rv} suggests that the eccentricities of massive Jovian planets are similar to those of stellar companions. }
%Fig. \ref{fig:massratio}). However, a detailed study is needed to exclude an origin in detection bias or other pipeline issues.
\begin{figure*}
\centering
\includegraphics[width=0.98\textwidth]{all_RV.pdf}
\caption{{\w Exploring the eccentricity distribution beyond Gaia binaries. % (stellar companion, $10^2-10^3$days).
The left panel plots the mean eccentricity as a function of orbital periods, for stellar and planetary companions. RV data are combined from \citet{Udry1998,Raghavan2010,Griffin2012}; visual binary data are from \citet{Tokovinin2016}; planet data are taken from exoplanetarchive.ipac.caltech.edu and are split into light and heavy Jovian companions.
Stellar binaries beyond $10^3$ days exhibit rising mean eccentricities. The right panel compares these long period binaries against some models.}
}\label{fig:all_rv}
\end{figure*}
Lastly, we wish to gain some insights as to {\w the outer reach of} the Rayleigh distribution. \citet{Tokovinin2020,Hwang2022}
show that binaries outside 100AU ($\sim 3 \times 10^5$ d) are thermally distributed in their eccentricities ($dN/de \propto e$).
What about the binaries between a few AU ($10^3$ days) and 100AU? {\w There can be at least three possibilities: the Rayleigh distribution may persist but with a mode that increases with period, smoothly connecting up to the thermal distribution (the ``rising Rayleigh'' model);\footnote{Formally, the thermal distribution, $dN/de \propto e$, is also a Rayleigh distribution but with $\sigma_e \rightarrow \infty$.} alternatively, the Rayleigh mode may remain constant but the Rayleigh distribution is gradually replaced by a thermal one as the period increases (the ``mixture'' model); lastly, the in-between group can take the ``uniform'' form ($dN/de =$ const).}
{\w Impressed by the constant Rayleigh mode we observe in the Gaia sample (Fig. \ref{fig:mainsequence}), we hypothesize that the ``mixture'' model is correct.
%same} Rayleigh distribution may extend to tens of AU, and is gradually replaced by the thermal distribution over the above range of separation.
We are also} prejudiced by the following theoretical argument.
%\x{Part of our argument is based on theoretical prejudice.
Close binaries are thought to extend to tens of AU, likely determined by the size of massive disks \citep{Tobin2020} and by the scale at which gravitational collapse occurs \citep{rafikov05,matznerlevin05,kratter10}. {\w If all close binaries experience the same eccentricity excitation, then}
%So if all close binaries are formed in disks,
the same Rayleigh distribution should be observable out to tens of AU.
{\w We evaluate this hypothesis by examining available data. Binaries from $10^3$ to $\sim 10^5$ days are probed by RV and other techniques. We combine the RV samples from \citet{Udry1998,Griffin2012,Raghavan2010} to enhance the long-period sensitivity, and adopt the result for visual binaries from \citet{Tokovinin2016}. As Fig. \ref{fig:all_rv} shows, the mean eccentricities clearly rise beyond $10^3$ days \citep{Tokovinin2016}. Unfortunately, the rising Rayleigh and the mixture model look very similar. Both are currently permitted by data (right panel of Fig. \ref{fig:all_rv}). The uniform model, on the other hand, is disfavored.}
%We point to two samples that are consistent with this suggestion.
%One is the sample of long-period ($p= 10^3 - 10^5$d) spectroscopic or visual binaries from \citet{Raghavan2010}.
%As we argue in Appendix \ref{sec:model-compare}, a mixture model with roughly equal proportions of Rayleigh and thermal provides a description that is marginally better than a uniform distribution (Fig. \ref{fig:mix2}).
%The second sample comes from the
{\w In addition,} while \citet{Tokovinin2020,Hwang2021} showed that a `uniform' model can explain the instantaneous radial-velocity vectors for Gaia
resolved binaries \citep[$10-100$AU:][]{Elbadry2021}, Fig. \ref{fig:mixed} argues that
both the rising Rayleigh and the mixture models can also do the job.
%\x{For those from $10-100$AU, their}
%can also be interpreted by the same mixture model (Fig. \ref{fig:mixed}), in place of the original 'uniform' model in
%\citet{Tokovinin2020,Hwang2022}.
% at $10-100$au ($p \sim 10^4 - 3\times 10^5$d).
%\item age dependence: Fig. 4 of \cite{Duchene2013} shows some weak age dependence,
%To fully settle this question, the data promised by Gaia DR4{\w /DR5} are needed.
{\w In conclusion, for binaries from a few to 100AU, while the `uniform' model may be discarded, it will take a much larger sample\footnote{\w We estimate that an un-biased sample of $N\geq 2000$ is needed.} to distinguish the `rising Rayleigh' model from the `mixture' model.} This may be possible using the data promised by Gaia DR4{\w /DR5.}
% We estimate $N\sim 2000$ un-biased sample is needed.
%Price-Whelan+2020: Apogee sample, many eccentricities are estimates; do find a bell-shaped distribution
%{\w sadly, no longer the case. Looked at all catalogues, nothing of value.}\x{ Such a result can also be confirmed independently using eclipsing binaries from TESS, OGLE... In this case, the eclipses provide tight measurements on one component of the eccentricity vector, $e \cos\omega$, where $\omega$ is the periaps angle. We expect this component to be a Gaussian distribution with $\sigma_e = 0.25$. This is indeed observed in Fig. \ref{fig:data2}. As different detection methods yield the same result, we are convinced that the Rayleigh distribution is not a result of selection bias. }
%TESS 14,000 EB only have a tiny number outside 50 days, not worth it
%https://arxiv.org/pdf/2409.20540
% similarly, very few Kepler EB, see TidalQ/Eclipsing_Binary/kepler_tess.ipynb
% the 2M EB in GAIA is not usable. HRD not clearly delineated (so can't separate ms from giant); ecosomega=0 not trust-worthy? period dist very irregular, selection effects
%Fig. \ref{fig:data1} shows that, for the broadest period group ($10^3-1200$d), the distribution appears to be broadening out. However, this is likely a result of the larger measurement errors for these border-line detections (even larger than those reported by the GAIA catalogue). They may not be intrinsic.
\section{Possible Origin{\w s}}
\label{sec:theory}
The eccentricities of AU-scale binaries are likely primordial, not affected by their birth clusters \citep[Appendix \ref{sec:unimportance}, also][]{Parker2009,Spurzem2009}, even less so by passing-by stars {\w after the cluster dispersal}. Our finding of a Rayleigh distribution that is invariant points to a universal process at birth.
%\footnote{A continuum of processes is also possible. However, it requires an invariant pattern of contributions, which seems harder to achieve than a single process. }
%However, \cite{Parker2009} showed that binaries inward of $\sim 10^2$au are safe from cluster perturbations.
{\w We briefly describe two possibilities: one is excitation by the circumbinary disks, and one is by ejecting brown dwarfs. The latter is a new proposal.}
\subsection{\w Circumbinary Disks}
\label{subsec:disk}
{\w Circumbinary disks, present around young binaries, may excite their eccentricities. This is an insight gained from numerical simulations \citep[for a review, see][]{Lai2023}. These simulations also found that there exist `attractor states', where the binary eccentricity reaches an equilibrium at around $e \sim 0.3 - 0.5$ \citep{Munoz2019,Orazio2021,Zrake2021,Siwek2023,Valli2024}. The exact value depends on parameters such as binary mass ratio, disk viscosity and thermodynamics. It is argued that even a disk with only a few percent of the stellar masses suffices for the task \citep{Dittmann2024,Valli2024}.
%\citet{Dittmann2024} compare 3 codes, showing e evolution is faster than a evolution; need a few percent disk mass to reach equilibrium eccentricity; but no explanation
}
{\w While we have advocated for a Rayleigh fit for its simplicity (one parameter fit), the Gaia binary data can also be described by a two-parameter Gaussian distribution
\begin{equation}
\frac{dN}{de} = \exp\left[{- \frac{(e-e_0)^2}{2\sigma^2}}\right]\, ,
\label{eq:normal}
\end{equation}
with $e_0 = 0.38$ and $\sigma = 0.20$ for the `gold' group. In this case, one can interpret $e_0$ as the most probable value from circumbinary action, and $\sigma$ the dispersion that arises from differences in physical parameters. }
%The physics of eccentricity excitation is not yet understood ... Numerical simulations with specific set-ups (equation of state, scale height, viscosity...) suggest that binaries at the so-called equilibrium eccentricities, which range from $\sim 0.3$ to $0.5$, are at peace with the (eccentric) disk \citet[e.g.][]{Munoz??, Siwek2023,Valli2024}. \citet{Valli2024} argued that even a disk with a few percent of the binary mass may succeed in reaching this; but most of these results are isothermal, and as is argued by \citet{Wang2023}
% used beta-cooling
%may be subject to large changes when the actual thermodynamics is considered. There is as yet no prediction.
{\w Interestingly, the theory community have long predicted eccentricity excitation by circumbinary disks and have been puzzled by the lack of signature in the observed binaries. Our result seems to validate this theory.}
{\w Before a victory is declared, however, we note that such a theory is still in its infancy. The origin of the `attractor state' has not been elucidated. The simulations are still too simplified, for instance, in their treatments of disk thermodynamics \citep{Wang2023}.
It is unclear why the observed eccentricity distribution is independent of binary period and stellar mass (Fig. \ref{fig:mainsequence}).
Lastly, current models suggest that the equilibrium eccentricity rises with binary mass ratio \citep{Siwek2023,Valli2024}, at odds with what we infer above.
These disagreements motivate us to expore an alternative scenario below.}
%\citet{Munoz2020} perhaps more clear physical picture, showing a rigid eccentric mode in CBD, confined to close to the edge; how does this lead to binary e?
% good summary by Wang+2023
% all previous isothermal, they beta-cooling, yields binary inspiral
%
% goodish summary by Valli+2024
% if e0>0.1, 10% mass accretion leads to e_eq
%Siwek2023, alpha=0.1, h=0.1
% Zrake2021, alpha=0.1, h=0.1
% made a silly claim about pAGB eccentricities, small sample + tide
%Wang+22,2212.07416
%
%Dittman+2024: varying alpha and h, but didn't look at e_eq
%similar results from different codes
%{\y Previous studies have found that small eccentricities $e_b < 0.08$ are damped through interaction with circumbinary disks (Munoz et al. 2019; Zrake et al. 2021), Z21 and DD21 both determine that near-circular orbits are stable and remain near-circular with $e′ < 0$, while S23b find that such near-circular orbits are unstable. not clear if related to viscosity profile; The gas dynamics around moderate- to high-eccentricity binaries is rich and complex and can differ substantially from their circular counterparts (Muñoz & Lai 2016). Eccentric binaries accreting from VSS disks might see their eccentricity damped or excited (Muñoz et al. 2019), suggesting that there may exist an equilibrium eccentricity at eb∼0.2–0.4 (see also Roedig et al. 2011), }
\subsection{\w Scattering Brown Dwarfs}
\label{subsec:scatter}
{\w This scenario is suggested by t}he Rayleigh form itself. A Rayleigh distribution is simply a Gaussian in 2-D: the two components of the eccentricity vector are each normally distributed around zero with the same mode.\footnote{In other words, a random Gaussian distribution in the Cartesian velocities. For a more general form of a triaxial Gaussian distribution, see \citet{Greenzweig1992}.
%Greenzweig shows the 3D form,
} Such a situation naturally arises from weak random scatterings, as has been observed in numerical experiments of planetesimal scatterings \citep{Greenzweig1992,Ida1992,Tremaine2015}.
%strangely Ida1992 claim it is Gaussian, not Rayleigh, did KSt test; but Ida1992b claimed
%Greenzweigh+Lissauer: A Rayleigh distribution in e and sin i is equivalent to a triaxial Gaussian distribution of random Cartesian velocities (see, e.g., Lissauer and Stewart 1992).
In fact, gravitational scatterings have been invoked to explain the e-distribution of asteroids \citep{Malhotra2017}, and exo-planets \citep{Zhou2007,Juric2008,Ford2008}.
This line of thought stimulates our investigation of the following scenario.
\begin{figure*}
\centering
\includegraphics[width=0.48\textwidth]{scattering.pdf}
\includegraphics[width=0.48\textwidth]{separation_w_bds.png}
\caption{Results of scattering experiments. We populate binary systems (primary $0.7 M_\odot$, secondary $M_*= 0.3 M_\odot$) with brown dwarfs that have the same total mass but different individual masses. The final binary e-distributions are Rayleigh in form (left panel, 100 cases each), with modes that depend only on the individual brown dwarf mass (eq. \ref{eq:equi2}). The right panel shows an example case with the orbits of the brown dwarfs as colored curve, and the secondary as heavy black curve. The vertical axis is the distance to the binary centre of mass. All brown dwarfs are promptly ejected except for one that becomes bound to the primary.
% {\y Sam: can bd mass be in unit of $M_\odot$?}
}\label{fig:scattering}
\end{figure*}
Consider, at birth, the presence of one or more low-mass bodies (`brown dwarfs') in the binary system. Unless sheltered dynamically, they should be quickly ejected by the binary through close encounters. Such encounters tend to establish equi-partition of
%angular momentum deficit (AMD) among the bodies \cite{Ida1992b}.
%Or equi-partition of
epicyclic energy among the bodies \citep{Ida1992b},
\begin{equation}
e_*^2 M_* \simeq e_{\rm BD}^2 M_{\rm BD}\, ,
\label{eq:equipartition}
\end{equation}
where $M_*$
%$ = M_1 M_2/(M_1+M_2)$ is the reduced-mass of
is the mass of the secondary. Setting $e_{\rm BD} \sim 1$ for ejection, we obtain,
\begin{equation}
e_* \simeq \left(\frac{M_{\rm BD}}{M_*}\right)^{1/2}
\simeq 0.30
\left(\frac{M_{\rm BD}}{0.03M_\odot}\right)^{1/2}\,
\left(\frac{M_*}{0.3M_\odot}\right)^{-1/2} \, ,
\label{eq:equi2}
\end{equation}
This expression does not depend on period, because
the scattering dynamics is scale-free \citep[ejection velocity is related to the local Keplerian velocity; also see Fig. 8 of][]{Juric2008}. Moreover, if there is (even a weak) correlation between $M_{\rm BD}$ and $M_*$, this expression will also be insensitive to the stellar masses. These will then explain the {\w universal Rayleigh mode} we find in Gaia binaries.
We carry out numerical experiments (details in Appendix \ref{sec:scatterings}) and present the results in Fig. \ref{fig:scattering}. We endow each binary system with a number of low-mass siblings. As the brown dwarfs are promptly cleared away, the binary eccentricities acquire a Rayleigh distribution, with the mode described roughly by eq. (\ref{eq:equipartition}). To achieve the observed value of $\sigma_e = 0.30$, we require $M_{\rm bd} \sim M_*/10$. We find that the outcome is not sensitive to the total number of brown dwarfs -- in fact, {\w even one} brown dwarf per system suffices \cite[also see][]{Ford2008}.
Could Nature provide such a set-up consistently?
In the scenario of disk fragmentation,
%kratter10 about how accretion can lead to disk fragmentation
% kratterlodato16 is a review
it is plausible that multiple low-mass objects would form alongside the dominant binary. At gravitational fragmentation, the characteristic mass is $\sim 10^{-2}M_\odot$
\cite[e.g.,][]{rafikov05,matznerlevin05,Xu2024}. This is close to the above brown dwarf mass.
%% Matzner+Levin argued binary not formed by disk fragmentation
%Xu2024 has a very thorough comparison of the fragmentation mass.
Subsequent nonlinear evolution is currently unclear \citep{Goodman2004,Levin2007}. The fragments can accrete from the disk and/or merge with each other. They can also migrate due to disk torque. The prevalence of close binaries in nature suggests that, in many cases, some of these seeds can grow to reach the isolation mass ($\sim M_\odot$). At the same time, other small seeds may persist or may be continuously produced, as in the forming triple system observed by \citet{tobin16triple}.
If so, disk fragmentation sets the stage for later dynamical scattering.
%{\cdm The fragmentation process is characterized by several mass scales: (1) a local disk mass scale, which is of order $10^{-3}$ to $10^{-2}M_\odot$ depending on initial radius; (2) a minimum fragmentation mass, $\sim10^{-2}M_\odot$ \cite{rafikov05,matznerlevin05} which restricts fragmentation and is analogous to the classical opacity limit \cite{low76,masunaga99} in the context of accretion disks; (3) a gap-opening mass, $\sim 10^{-2}$ to $10^{-1}M_\odot$, and (4) an isolation mass, $\sim M_\odot$ (see \cite{kratterlodato16}). Simulations such as demonstrate that fragmentation can be prolific, and while a dominant duo or triple approaches the isolation mass, it is entirely plausible that other objects remain relatively small, as in the forming triple system observed by \cite{tobin16triple}. Interactions with the unstable disk can drive low-mass objects inward \cite{baruteau11} even more effectively than massive ones \cite{zhu12}. It is plausible, therefore, that disk fragmentation sets the stage for later dynamical scattering.}
Further investigations may reveal if $M_{\rm bd}/M_*$ does remain nearly constant in all disks.
{\w Previously,
\citet{Reipurth2001} and \citet{Kroupa2003} have proposed that brown dwarfs are ejected embryos from their natal systems. Our hypothesis is similar in nature to theirs.}
%numerically studied by \citet{Sterzik1998}
%The presence of brown dwarfs in open clusters places an upper limit of several kilometers per second on the ejection velocities of brown dwarfs that are retained in the cluster
If all close binaries have ejected
%at least
{\w about} one brown dwarf, this can explain all sub-stellar objects detected by imaging in young clusters and by microlensing surveys (Appendix \ref{sec:enoughbd}).
{\w Moreover, one expects these brown dwarfs to acquire velocity dispersions of order the orbital escape velocity.\footnote{\w For a 50\,au binary (peak of the binary) with $M_1 = 0.3 M_\odot$ (peak of the initial mass function), this is $\sim 2\km/\s$.} If this is larger than the dispersion velocity in a young cluster, one expects that brown dwarfs should preferentially lie at the outskirts of these clusters, and that they have larger proper motions. Currently, there are conflicting observational evidences \citep[e.g.][]{Kumar2007,Scholz2012,Downes2014,Panwar2024}, and a systematic study is required.}
%Scholz2012 said NGC1333 brown-dwarfs are co-spatial as stars,
%Kumar2007 said they are different, but not very strongly so
%Downes2014, said 25 Orionis has no difference in bd and low-mass stars
%Panwar2024, Berkeley 59, brown-dwarf are on the outside
\subsection{\w Planets and Triples}
{\w As both theories (disk and scattering) suffer from insufficiencies, we look to data for clues. Here, we examine two populations: Jovian planets, and compact hierarchical triples.}
{\w Fig. \ref{fig:all_rv} shows that heavy Jovian planets ($M\sin i > 3 M_J$) in the period range ($10^2-10^3$ days) exhibit a similar eccentricity distribution as Gaia binaries. In comparison, lighter planets are more circular. Both disk and scattering may potentially explain these facts. Heavy planets can grow in eccentricity by interacting with the disks \citep[see, e.g.][]{Papaloizou2001}; they could also be excited by destabilizing and scattering their siblings \citep[see, e.g.,][]{Ford2008,Juric2008}. The latter requires the ejected bodies to be dominated in mass by those that are $\sim 1/10$ of the Jovians. It is unclear if this is possible. Similarly, the disk theory also needs to explain why systems of such low mass ratios are excited to the same eccentricities as stellar binaries. }
{\w Compact hierarchical triples are stellar systems where the outer orbits are smaller than a few AU \citep[see review by][]{Tokovinin2021}.
%,Borkovits2022},
The eccentricity distribution of their outer orbits is also similar to that of the Gaia binaries considered here
\citep{Borkovits2016,Cza2023,Moharana2024}\footnote{Their inner orbits are commonly short and have likely experienced tidal damping.} This suggests that the presence of a tight inner binary does not affect the outer eccentricity.
%Outer components in hierarchical triples appear to exhibit similar eccentricity distribution \citet{Borkovits2016,Cza2023}
%Borkovits using Kepler
%Cza using Gaia
% Rapapport+23 using TESS
%as the Rayleigh form here.
Moreover, \citet{Borkovits2016,Tokovinin2017} demonstrated a strong degree of alignment between the inner and the outer orbits in these systems, with a dispersion of order $20\deg$.
%The alignment is strong in triples with outer projected separation less than ∼50 au, where the average Φ is about 20. I"
This is, again, naturally accommodated by both the disk and the scattering theory (see Appendix \ref{sec:scatterings}).
In conclusion, we remain in limbo.
}
\section{Conclusions}\label{sec:conclusions}
The Gaia mission greatly expands our sphere of vision. The number of AU-scale binaries, for which eccentricity information can be reliably extracted, rises
{\w from thousands} to of order $10^5$. Among these, we select a largely un-biased sample of $\sim 3000$ systems to deduce the underlying eccentricity distribution.
A simple and elegant Rayleigh distribution emerges, with a mode of $\sim 0.3$. The value of the mode appears invariant with respect to stellar mass and orbital period, but a more definitive conclusion will require careful study of the selection bias.
%The mode is observed to be invariant with stellar spectral types and with orbital periods (from $\sim 20-1200$d).
Such a distribution is almost certainly primordial in origin. Its (apparent) invariance points to a universal process. And the Rayleigh form itself suggests an origin in weak scatterings.
%We hypothesize that, during the last phase of star formation, one or more brown dwarf siblings are ejected by the stellar binary.
% Juri and ford are same issue, but Juric is earlier
%After its birth in the disk, the binary can be initially eccentric due to the m = 1 mode perturbation (Shu et al. 1990; Krumholz et al. 2007; Kratter et al. 2010; Kratter 2011), surrounded by the circumbinary disk. The interaction between the binary and the circumbinary disk induces orbital migration and excites the
%binary eccentricity (Artymowicz et al. 1991; Artymowicz \& Lubow 1994), and the effect depends on the properties of the disk and the binary (Pichardo et al. 2005; Ragusa et al. 2020; Heath \& Nixon 2020).
%Forming low-mass stars (K- and M-dwarfs) from disk fragmentation may be difficult because their disk is not sufficiently massive for gravitational instability (Kratter et al. 2008). Instead, such binaries may be formed from the dynamical interaction of unstable multiples (Bate et al. 2002), and their eccentricity distribution is subject to the interaction with disk and gas.
{\w Our numerical experiments show that, in order} to reproduce the observed Rayleigh mode, the stellar binaries need to eject brown dwarfs with masses $\sim 1/10$ as heavy.
%of the secondary masses.
It is not known why this must be so. But if true, such ejections can account for almost all free-floating
%\x{sub-stellar objects}
{\w brown dwarfs}. The kinematics {\w of these bodies should} bear imprints of the ejection process.
%These unlucky fellows must make up the bulk of the known sub-stellar objects detected by imaging and microlensing searches \cite{Lodieu2013,Sumi2023}
%And the 'brown-dwarf desert' \cite[Fig. 5 of][]{Duchene2013}, dearth of sub-stellar companions (brown-dwarf desert), can be naturally explained if such objects are only formed in massive disks and are removed by their most massive siblings.,
%
{\w An alternative theoretical explanation also exists. Binaries may spontaneously grow in eccentricity when interacting with circumbinary disks. While this behavior has been predicted by numerical simulations, theoretical understanding is lacking and current predictions do not readily agree with data. %More sophisticated treatments of thermodynamics and hydrodynamics are required.
We point out that stars with planetary companions, and stars in triple systems, may provide extra constraints on the excitation.
}
%{\y below?}
%\cite{Goodwin2007} suggested that brown-dwarfs are formed in disks $>100$au as companions to more massive stars, and then ejected from system gently by passing-by stars.
%\x{It is natural to ask how far in period such a Rayleigh distribution extends.}
{\w The radial extent of the Rayleigh distribution could also inform about its origin. Available data suggest that the mean eccentricity rises with binary period. But it is unclear if the Rayleigh mode itself rises beyond $10^3$ days.} {\w We hope that Gaia %DR4
DR5, with a baseline of 11 yrs, } will provide an answer to this question.
%The distribution of binary mass-ratios is also a fossil record of formation.
%The 2-million strong eclipsing binaries \cite{Mowlavi2023} can be queries to provide information on twin binaries, currently absent from astrometric data.
With this new discovery,
the eccentricity distribution now becomes a new marker for the process of disk fragmentation, joining rank with other measurables like period, metallicity, mass ratio, and twin-fraction. It can be used to probe many interesting questions. For instance, do binaries in extreme environments (e.g., globular clusters, nuclear star clusters) form in disks? Does the Rayleigh mode vary with environmental factors? %\x{How do close-binaries pair and how does this pairing correlate with their eccentricities (Fig. \ref{fig:massratio})?}
AU-scale binaries are important drivers for binary stellar evolution -- given the AU-sizes of giant stars, these include most of the binaries that are destined to interact during their lifetimes, through tides, mass transfer and common envelope.
%Geller+: 'For example the stellar merger rate may be overpredicted by a factor of about two (see Section 6), if an initially thermal eccentricity distribution is chosen when a uni- form distribution is more appropriate.'
The Rayleigh distribution, as opposed to the older uniform distribution, leads to fewer binary mergers. This new distribution should be adopted in synthetic studies of binary evolution.
%{\w can be tested in binary inclinations, same mode? also mass ratio is murkey, can now be attemped...\cite{Moe2017}}
%For the sample selected by astrometry (including RV+astrometry), twin binaries are excluded because they do not produce measurable photo-centre shift. What do they do? We take a look at the SB2? {\w detail?}
%primary mass: It is impossible, therefore, to tease out eccentricity differences between different stars, which may have different proclivity for disk fragmentation (Krater+2008?)
%It is also unclear if the same distribution applies to all stellar masses, e.g., M-dwarfs have massive enough disks? (Kratter+2008);
%eclipses provide tight measurements on one component of the eccentricity vector, $e \cos\omega$, where $\omega$ is the periaps angle. We expect this component to be a Gaussian distribution with the same mode as our Rayleigh distribution. This is most likely using GAIA EBs, but the current EB catalogue from GAIA DR3 \cite{Mowlavi2023} are too contaminated to be useful.
%
%TESS 14,000 EB only have a tiny number outside 50 days, not worth it
%https://arxiv.org/pdf/2409.20540
% similarly, very few Kepler EB, see TidalQ/Eclipsing_Binary/kepler_tess.ipynb
% the 2M EB in GAIA is not usable. HRD not clearly delineated (so can't separate ms from giant); ecosomega=0 not trust-worthy? period dist very irregular, selection effects
%\backmatter
%\bmhead{Supplementary information} %Please see the appendixes.
\bigskip
{\w We thank an anonymous reviewer for penetrating comments that lead to a much expanded discussion, and Ayush Moharana for bringing the compact hierarchical triples to our attention.} We thank NSERC for research funding, and the Gaia collaboration for a marvelous gold mine.
%{\y Kareem's funding boss here}
%{\w points after submission:One thing we may want to emphasize more during revision is the fact that there is no need to scatter 10 brown-dwarfs, one or two may suffice. That may help alleviate the formation problem and hiding the brown-dwarfs till they are formed, likely out of gravitational instability in the same massive disk. Disk self-gravity gives a formation timescale of order dynamical time, but the clump still needs to cool enough to become ‘brown-dwarf’ like.
%Ayush Moharana: The eccentricity distribution you present in your work was also found in compact hierarchical triples (CHTs; tertiary period<1000 days) by Czavalinga et al. 2023. Also, in Moharana et al. 2024, we found that this distribution slightly becomes thermal (for e~0.2) when considering metal-poor CHTs. I wonder if this can be recreated by external excitation (BD or other). If not, this could probably hint that the eccentricity distribution could be arising from a different physical process.
%Dear Ayush, your email is the most interesting among all I have received this morning. I will think about the CHT sample and modify our draft to include these earlier findings – didn’t find it while literature searching, our mistakes. And the CHT really do set constraint on the brown-dwarf scenario. Let me think about it more and let’s keep contact.
\bibliography{binary}
\begin{appendix}
\section{GAIA binary extraction}\label{sec:A1}
We construct a binary sample that is little affected by either detection bias or evolutionary changes.
Starting from the Gaia non-single-star catalog \citep{Arenou2023}, we apply a number of cuts consecutively. These steps and their resulting sample sizes are listed in Table \ref{tab:sample}. Our final sample (named the `gold' sample) is
homogeneous in primary properties (Sun-like dwarfs), and avoid any significant selection bias.
Here are some explanations. A fraction of the Gaia catalogue stars have determined astrophysical parameters. Among these, we retain only systems with main-sequence primaries, defined as
\begin{equation}
M_G > {\rm Max}\{4.0, \, \left[(B_p-R_p)_{\rm de-red} - 1\right]^7\}\,
\label{eq:BPRP}
\end{equation}
where $M_G$ is the absolute g-band magnitude and $(B_p-R_p)_{\rm de-red}$ is the de-reddened color.
We also remove those with `significance' $<10$. Aside from producing a cleaner sample, this automatically rejects any binaries detected as `Eclipsing Binaries'. We exclude these binaries because many of them have eccentricities artificially set to zero.
We also select binaries with periods from $10^2-10^3$days. The former is set to avoid pollution from tidal circularization, while the latter is due to the finite coverage of Gaia DR3. To meaningfully compare against the detection completeness from mock pipelines \citep{elbadry2024}, we proceed to retain only binaries characterized by astrometry (including `Orbital' and `AstroSpectroSB1' types). Among these, we require a further quality cut: `goodness-of-fit' $< 5$ if `phot-g-mean-mag $> 13$, or $< 10$ if otherwise.\footnote{\w This quantity indicates how compatible the Gaia data is with a single star solution.} This filters out bad solutions, the threshold for which depends on brightness. A system is considered `Sun-like' if the primary radius is within $[0.7,2.0]R_\odot$.
And lastly, for our `gold' sample, we set a maximum distance of $150$\,pc. This last cut leaves us with $\sim 3000$ systems, a tiny fraction of the original data($\leq 1\%$). About $\sim 45\%$ of the `gold' sample have only astrometric orbits (`Orbital'), while the rest are additionally characterized by radial velocity (`AstroSpectroSB1'). The latter group tend to be brighter.
Ideally, we would also like to remove systems with white-dwarf secondaries. Their eccentricities have likely been strongly suppressed during the giant phase. However, doing this thoroughly is difficult at the moment \citep{Shahaf2023}. Fortunately, such binaries mostly contribute at the longest periods. Results in Fig. \ref{fig:mainsequence} show that their impact is likely minor.
Regardless of the cut, all samples exhibit e-distributions that have a Rayleigh shape. Their Rayleigh modes, however, differ (Table \ref{tab:sample}).
Most cuts return a lower mode ($\sigma_e \sim 0.25$), except for the `gold' sample ($\sigma_e \sim 0.30$).\footnote{Within the `gold' group, the Rayleigh mode does not vary with brightness, nor with the detection method.}
%suggesting again there is little detection bias in our 'gold' sample.
Such a difference, we argue (Fig. \ref{fig:mainsequence}), reflects not intrinsic variation, but impact of the selection bias (Appendix \ref{sec:selection}).
%To avoid pollution by white-dwarf-main-sequence binaries, we only retain those with 'fluxratio' $F_2/F_1 >10^{-3}$ \cite{Arenou2023}.
%At short periods, the main-sequence binaries show the well-known signature of tidal circularization: a circular island that extends to $P \sim 15$days, but at the same time a profusion of eccentric binaries at shorter periods.
%%We split the ms-ms binaries by their primary radii and present the orbital properties in Fig. \ref{fig:mainsequence}. It appears that regardless of primary radii, there is a prominent circular island that extends to $P \sim 15-20$d. This is contrary to what \citet{Bashi2023} reported, that the circularization period rises for lower-mass primaries. \citet{Bashi2023} arrived at this conclusion by rejecting systems in the circular island and only focus on the upper envelope in eccentricity-period space.\footnote{The use of SB1 binaries may also introduce differences biases in periods, for different stellar types.} This procedure, we believe, is not justified. The circular island does not result from noise \citep[as is implied by][]{Bashi2023}. It is a physical feature of the data.
\section{Characterizing the Detection Bias}
\label{sec:selection}
\begin{figure*}
\centering
%\includegraphics[width=0.45\textwidth]{x_extrapolate.eps}
\includegraphics[width=0.95\textwidth]{gaia_binaries_ecc_histogram.pdf}
\caption{Correction for detection bias. The left panel repeats that in Fig. \ref{fig:data1}. The middle panel shows the completeness as a function of binary eccentricity, obtained from the mock astrometry pipeline \citep{elbadry2024}. The right panel shows the corrected values. Correction brings the different distributions into better agreements with each other, though some differences still exist. {\w Measurement errors, not included here, may help resolve some of these differences.}
}
\label{fig:kareem}
\end{figure*}
{\it Gaia} is, generically, less sensitive to high-eccentricity orbits. High-eccentricity binaries spend more time near apoapse and are thus less likely to have their orbits well sampled by {\it Gaia} observations, which occur at quasi-random times. This results in orbits that are on average less well constrained at high eccentricity, and less likely to pass the stringent quality cuts imposed on the orbital solutions published in {\it Gaia} DR3. Such a bias very likely affects both astrometric and RV orbits, but it has thus far been quantitatively modeled only for astrometric orbits, and for this reason we focus on astrometric orbits in this work.
We use the forward-model of {\it Gaia} astrometric orbit catalogs described by \cite{elbadry2024} to quantify the eccentricity bias in our sample. The model produces realistic epoch astrometry for each simulated binary and fits it using the same cascade of astrometric models used in producing the DR3 binary catalogs \cite[see][]{Halbwachs2023}. Following \cite{elbadry2024}, we generate a realistic population of binaries within 2 kpc of the Sun, mock-observe them, and produce a mock catalog. Unlike \cite{elbadry2024}, who assumed an eccentricity distribution following \cite{Moe2017}, we adopt a uniform eccentricity distribution,
%such that the eccentricity distribution of the mock catalog
which allows us to trace {\it Gaia's} eccentricity bias. Other properties of the simulated binaries (e.g. masses, ages, evolutionary states, orbital periods, 3D locations in the Galaxy) are chosen as described by \cite{elbadry2024} and are reasonably good approximations of reality. To represent the bias specific to the observational sample analyzed here, we exclude binaries containing red giants and require astrometric significance $> 10$ in addition to the quality cuts imposed on the solutions published in DR3.
The results of these simulations are shown in the middle panel of Figure \ref{fig:kareem}. As with the observed data, we show individual distance bins separately. We normalize the distributions in each eccentricity bin such that the average relative completeness at $0 < e < 0.3$ is unity. A bias against high eccentricities exists in all distance bins, but it is more severe at large distances, where orbits have smaller angular size at fixed period and thus lower astrometric SNR. This trend mirrors what is found in the observed sample. The simulations suggest that {\it Gaia} sample with $d < 150$\,pc is largely unbiased at $e < 0.5$, but a bias against high eccentricities is present at higher eccentricities. At $e>0.9$, the completeness is $\approx 3$ times lower than at $e < 0.5$.
The right panel of Figure~\ref{fig:kareem} shows the normalized eccentricity distribution of the observed samples after correcting for incompleteness. The corrected eccentricity distributions in different distance bins are in better agreement with one another, although some trend of lower eccentricity at larger distances is still present. This could reflect imperfections in the forward-model, although it could also result from a mass-dependent eccentricity distribution, since more distant binaries are on average brighter and more massive {\w (but see Appendix \ref{sec:massratio}).} {\w For the 'gold' sample, we find that these corrections bring the best-fit Rayleigh mode from $\sigma_e = 0.303$ to $\sigma_e = 0.336$.}
\begin{figure*}
\centering
%\includegraphics[width=0.45\textwidth]{x_extrapolate.eps}
\includegraphics[width=0.95\textwidth]{x_extrapolate.pdf}
\caption{We use the `Sun-like' sample in Table \ref{tab:sample} to explore the effects of detection bias. The left panel shows that the mean eccentricity ($\langle e \rangle$) gradually declines with distance. The brown line is $\langle e \rangle/\sqrt{\pi/2}$, a proxy for the Rayleigh mode. Using equation \ref{eq:decline} to describe how distance affects the mode, we find an an intrinsic Rayleigh mode of $\sigma_{e0} \approx 0.30$, and that systems inward of $d_0 \sim 150$\,pc should be free of detection bias. {\w The blue curve on the left ($\bar\epsilon_e$) denotes the mean measurement error in eccentricity.}}
\label{fig:x_extrapolate}
\end{figure*}
We also use the observed sample itself to gauge how much this bias affects the measurement for the Rayleigh mode. This is a self-calibration method and is independent of the above mock pipeline.
In Fig. \ref{fig:x_extrapolate}, we present the eccentricities and their mean values ($\langle e \rangle$) as functions of distance from Earth. We use $\langle e \rangle/\sqrt{\pi/2}$ as a proxy for the Rayleigh mode $\sigma_e$ and fit the following data-inspired form,
\begin{equation}
\sigma_e (d) = \sigma_{e0} \, {\rm Min}\left[1, 1- \log_{\rm 10}\left(\frac{d}{d_0}\right)^a\right]\,.
\label{eq:decline}
\end{equation}
where $\sigma_{e0}$ is the intrinsic Rayleigh mode. Such a form asserts that,
for systems closer than $d_0$, $\sigma_e = \sigma_{e0}$, i.e., there is little bias for the bulk of the Rayleigh distribution (but there can still be bias at high eccentricities).
Such a form is reasonable for a Rayleigh distribution (which concentrates at low-e), but is less so for a, e.g., power-law distribution.
The observed data yields the following best-fit parameters: $\sigma_{e0} = 0.301\pm 0.006$, $d_0 = 147 \pm 27$\,pc, and $a=0.29\pm 0.01$. This validates our main results, that the intrinsic Rayleigh scale is $\sigma_e \sim 0.3$, as measured from the `gold' sample. This does not mean, however, that the `gold' sample is complete to $d_0 \sim 150$\,pc -- it is not. It is missing binaries of high eccentricities (see Fig. \ref{fig:kareem}), as well as binaries of comparable brightnesses, or those with very low-mass secondaries. {\w But these deficiencies do not strongly affect the Rayleigh mode.}
{\w We notice a strange population of very circular binaries in Fig. \ref{fig:x_extrapolate}. They account for $\sim 3\%$ of the total system and have reported eccentricities that fall much below the measurement errors. This is also seen in the excess in low-eccentricity bins in Fig. \ref{fig:data1}. They likely arise from issues in the Gaia pipeline. Removing this population entirely does not change the Rayleigh mode we inferred.
}
\section{\w Eccentricity Errors}
\label{sec:error}
{\w We now turn to consider the impact of eccentricity errors. Fig. \ref{fig:x_extrapolate} presents the mean measurement errors ($\bar \epsilon_e$) as reported by {\it Gaia}. They rise with distance, with $\bar\epsilon_e \sim 0.02$ in the nearest systems to $\sim 0.10$ at $1.5$\,kpc. Eccentricity errors enlarge the Rayleigh mode and we model it as the following.
Consider one component of the eccentricity vector. Let $e_x$ be the original value with a standard derivation of $\sigma_e$, noise $\delta _e$ be normally distributed around zero with a standard deviation $\epsilon$ (with $\epsilon = \bar\epsilon_e/\sqrt{\pi/2}$). The measured value $e'_x$ should be distributed as
\begin{eqnarray}
& & P(e'_x) = \int_{-\infty}^{+\infty} P(e_x) P(\delta_e) d\delta_e\,
\nonumber \\
& & \propto \int_{-\infty}^{+\infty} \exp\left[- \frac{{(e'_x - \delta e)}^2}{2\sigma_e^2}\right]\, \exp\left[- \frac{\delta_e^2}{2\epsilon^2}\right]\, d \delta_e\nonumber \\
& & \propto
\exp\left[ - {{ab}\over{a+b}} {e'_x}^2\right]\,
\int_{-\infty}^{+\infty}
\exp\left[ - (a+b)\left(\delta_e - {{a e_x'}\over{a+b}}\right)^2
\right] \, d\delta_e\nonumber \\
& & \propto
\exp\left[ - \frac{{e'_x}^2}{2 (\sigma_e^2 + \epsilon^2)}\right]\, ,
\label{eq:convolution}
\end{eqnarray}
where $1/a = 2 \sigma_e^2$ and $1/b = 2 \epsilon^2$. So the measured Rayleigh mode for $e'$ is now larger and is $\sqrt{\sigma_e^2 + \epsilon^2}$, as is dictated by the Central limit theorem. For the above reported measurement errors, this change is relatively minor, in comparison to those caused by selection effects.
}
\section{Dependence on Mass Ratio}
\label{sec:massratio}
{\w The Gaia NSS catalogue reports masses for the primary stars, based on isochrone fitting.\footnote{\w We adopt the IsocLum mass from \citet{Arenou2023}. There is a strange cluster at mass of $1.07214M_\odot$. We replace these with their Flame masses.}
For a subset of the main-sequence binaries, the Gaia pipeline also returns mass estimates for the secondaries. It does so by assuming that the components satisfy the main-sequence mass-luminosity relation. }
%%% I actually change m2 to be
%%% compatible w/ L = M^4
{ Using %a subset of the Gaia sample where both stellar masses are reported,
{\w this subset,}
\footnote{We further restrict ourselves to those with `fluxratio' $> 10^{-3}$ in order to remove any systems with white dwarf companions. This is not important for the `gold' sample but does affect the samples at larger distances.} we analyze how the eccentricities depend on mass ratio (Fig. \ref{fig:massratio}).
%However, a detailed study is needed to exclude an origin in detection bias (Appendix \ref{sec:massratio}).
%
%\section{Dependence on Binary Mass Ratio?}
%\label{sec:massratio}
%Whether the Rayleigh mode depends on the mass-ratio in a binary is of high relevance for its origin.
We split the `Sun-like' sample by the same distance division as before, and present results for the three closest group (with enough cases).
We find that the binary counts drop off steeply towards equal-mass, in contrast to that found in the spectroscopic sample by \citet{Moe2017}. This is explained, at least partly, by the fact that an equal-brightness binary exhibits zero astrometric signal.
%and almost all binaries here are selected by astrometry.
We use mean eccentricity as a proxy for the Rayleigh mode and find that it %\x{also drops off towards equal masses. While this seems statistically significant}
{\w appears to drop off towards equal mass in the `gold' sample, but are largely flat for the other two samples we tested.}
%Moreover,
{\w This drop-off} occurs at the same mass ratio as the number drop-off, suggesting a common origin in detection bias.
{\w This leads us to conclude that we cannot at the moment extract a definitive answer from the Gaia data. }
%If we expand our horizon to include heavy Jupiters (Fig. \ref{fig:all_rv}), it appears that they have a similar eccentricity distribution as the stellar binaries studied here. Considering them as the limit of extremely small mass-ratio, this would argue that the eccentricities are not sensitive to binary mass-ratio.}
%\x{A more detailed analysis is required to establish the authenticity of this result. }
\begin{figure*}
\centering
%\includegraphics[width=0.45\textwidth]{x_extrapolate.eps}
\includegraphics[width=0.95\textwidth]{mratio.pdf}
\caption{A possible dependence of the Rayleigh mode on binary mass ratios is seen, in the Gaia `Sun-like' sample. The left panel shows the number of systems with reported mass ratios, divided by distances. The drop-off %towards equal masses
{\w beyond mass ratio $0.6$}
may be explained by astrometric bias {\w against detecting equal brightness pairs}.
The right panel shows the proxy for the Rayleigh mode, $\langle e \rangle/\sqrt{\pi/2}$. A possible reduction towards equal-mass is seen, but we remain suspicious.
}
\label{fig:massratio}
\end{figure*}
%We estimate the fraction Fsolar+WD;logP<4.5 = (3.4 ± 1.0)% of solar-type stars in the solar-neighborhood that contain close WD companions with log P (days) < 4.5 (a . 30 AU) using three independent methods (§8.3). Because ≈10% of solar-type stars contain close companions with log P (days) < 4.5 in which the nature of the secondary cannot be determined, we find that (30 ± 10)% of solar-type SB1s actually contain WD companions.
%% if required, the content of .bbl file can be included here once bbl is generated
%%\input binary.bbl
%\subsection{Rayleigh and thermal}
\section{Binaries at Longer Periods?}
\label{sec:longperiod}
%In the following, we ask whether the same Rayleigh distribution extends to binaries outside $10^3$d.
%Knowing that wide binaries have thermal distributions \citep{Tokovinin2020,Hwang2021}, we can imagine at least two possibilities: the Rayleigh may persist but with an increasing mode further out, smoothly connecting up to the thermal distribution;\footnote{Formally, the thermal distribution, $dN/de \propto e$, is a Rayleigh distribution with $\sigma_e \rightarrow \infty$.} or both distributions co-exist, with the fraction of Rayleigh (with the same mode) gradually going down to zero. Given the invariance of the Rayleigh mode in our `gold' sample, we choose to focus on the second possibility. We draw some partial conclusions based on two different samples.
%The first is the previous ground-based sample from \cite{Raghavan2010}, which includes $58$ binaries in the period range $10^3- 10^5$d (or $\sim 2 - 40$AU). {\y need to refer to other appendixes} We consider a model $\mathcal{M}_3$ in which the eccentricities are distributed according to a mixture of Rayleigh and thermal:
%\begin{equation} \mathrm{Pr}(D|\sigma_e,f,\mathcal{M}_3) =\prod_{i=1}^N \left[ f\dfrac{e_i}{\sigma_e^2} \exp\left(-\dfrac{e_i^2}{2\sigma_e^2}\right) +2(1-f)e_i \right]\, ,
%\end{equation}
%with $f$ being the Rayleigh fraction. We adopt a flat prior for $f$ and a prior $\mathrm{Pr}(\sigma_e|\mathcal{M}_3)=\delta(\sigma_e-0.303),$ which equates to conditioning based on the knowledge gained from our Gaia sample\footnote{A uniform prior in $\sigma_e$ gives similar results, as long as $\sigma_\textrm{min}<0.303<\sigma_\textrm{max}$.}.
%We then compute the evidence $\mathcal{E}_\textrm{mix}=\mathcal{E}(D_\textrm{LP}|\mathcal{M}_3)$ for the long-period sample. Our results indicate that a mixture fraction of $f=0.52$ best describes the data. Comparing against evidence values computed for a purely thermal, a purely uniform (%in the absence of any observational uncertainty $\mathcal{E}_\textrm{therm}=2^N\mathcal{P}$ and $\mathcal{E}_\textrm{uni}=1$, respectively), and a purely Rayleigh distributions, we find Bayes factors $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{therm}\approx6.5\times10^3$, $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{uni}\approx12.3$, and $\mathcal{E}_\textrm{mix}/\mathcal{E}(D_\textrm{LP}|\mathcal{M}_1)\approx20.2.$ These factors indicate that the mixture model $\mathcal{M}_3$ is much better supported by the long-period data than a thermal eccentricity distribution, and marginally better supported than a uniform or a Rayleigh distribution.
%These results are graphically displayed in Fig. \ref{fig:mix2}.
%We find \autoref{fig:mix2} compares cumulative distributions for the Gaia and spectroscopic binary samples with predictions for model $\mathcal{M}_1$ (black), model $\mathcal{M}_3$ (orange), and a uniform distribution.
%{\y figure removed, need revision}
%\begin{figure}
% \centering
%\includegraphics[width=0.45\textwidth]{moe2017_CDF_fit.pdf}
% \caption{The cumulative histogram of eccentricity distribution for the long-period ($10^3-10^5$d) binary sample from \citet{Raghavan2010,Moe2017}. It best supports a model that is a mixture of a Rayleigh and a thermal distribution (smooth orange curve). The other curves show the variety of models we compare against (thermal, uniform, and pure Rayleigh). %with Bayes factors that $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{therm}\approx6.5\times10^3$, $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{uni}\approx12.3$, $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{Rayleigh}\approx 20.2$,
% }
% \label{fig:mix2}
%\end{figure}
{\w Here, we re-examine the resolved binaries from Gaia \citep{Elbadry2021}}. \citet{Tokovinin2020,Hwang2022} established, statistically, the eccentricity distribution by measuring the instantaneous velocity-position ($v-r$) angles that are projected on the sky %, where $v$ and $r$ are the projected vectors on the sky plane
Results from \citet{Hwang2022} are shown in Fig. \ref{fig:mixed} for binaries with projected separations that fall within $10-1000$AU.
%\cite{Tokovinin2020} used a sub-sample ($\sim 3600$) of the \cite{Elbadry2018} binary catalogue, median mass $0.6 M_\odot$, median separation $450$au, goes down to $100$;
%i.e., proper-motion, not RV
%Tokovinin2020 actually used both $v-r$ angle and orbital speed
In the absence of a better model, they parameterized the eccentricity distribution as a power-law, $dN/de \propto e^\eta$ and found that the value of $\eta$ rises from 0 (`uniform') at $\sim 50$ au, to $1$ (`thermal') at $\sim 500$au, and $>1$ (`super-thermal') beyond. Alternatively, as we show in Fig. \ref{fig:mixed}, their so-called `uniform' distribution can easily be swapped for a mixture model with roughly comparable fractions in the thermal distribution and the Rayleigh distribution. %The data quality is not sufficient to tell the two apart.
%These two pieces of evidence, taken together, suggest that the Rayleigh distribution may persist to at least tens of AU, and
In fact, the apparent change in $\eta$ with period \citep{Hwang2022} may simply reflect the decreasing fraction of the Rayleigh with increasing period.
\begin{figure}
\centering
%\includegraphics[width=0.45\textwidth]{data_comparison.eps}
\includegraphics[width=0.45\textwidth]{gamma_dist.pdf}
\caption{Distribution in the sky-projected displacement-velocity angle $\gamma$ for resolved binaries. The observed values (thick histograms for different values of projected separation $s$) are adopted from Fig. 6 of \citet{Hwang2022}, and different models are in thin lines, with our best-fit mixture model (\S \ref{sec:model-compare}) in red.
Wide binaries ($10^2 < s < 10^3$au) are best described by a thermal distribution, but close binaries ($10^1 < s < 10^2$au)
can be described equally well by a uniform distribution ($dN/de = {\rm const}$) or by our mixture model.
% In other words, an apparent 'uniform' distribution results when the Rayleigh distribution is gradually replaced by a thermal distribution around 10s' of AU.
}
\label{fig:mixed}
\end{figure}
\section{The unimportance of stellar encounters}
\label{sec:unimportance}
Could the AU-scale binaries have obtained their eccentricities by scattering passing stars, especially while they are still within their birth clusters?
We hold that this is unlikely. We present multiple arguments.
For our AU-binaries, we are in the `hard' binary case \citep{Heggie1975}, where the binary orbital velocity ($V_{\rm orb}$) exceeds the mean dispersion velocity of the cluster ($ V_\infty$, e.g, $V_\infty \sim 12$\,km/s in the dense core of 47 Tuc).
%from Davis+Sigurdsson: 47 Tucanae is a massive and quite dense cluster. The central density is about 1:5 105 pc23 with a central velocity dispersion of ,12 km s21. The core radius is about 0.5 pc and the half-mass radius about 3.9pc (Djorgovski 1993; Howell, Guhathakurta & Gilliland 2000). At the half-mass radius the number density of stars is about a factor of ten lower than in the core.
If we consider impacts with the closet approach distance $R_{\rm min} \geq a$ (binary separation), the encounter is in the adiabatic limit
% adiabatic: angular speed
% of binary >> 3rd star
where the binary have time to revolve multiple times during one close-approach of the third star. Moreover, the third star orbit is close to being parabolic. An initially circular binary will receive a kick in eccentricity that is of order \citep{Heggie1996,Spurzem2009},
%Spurzem eq. (A5), nice compilation
\begin{eqnarray}
\delta e & \approx &
3\sqrt{2\pi} \frac{M_3 M_{12}^{1/4}}{M_{123}^{5/4}} \left(\frac{2 R_{\rm min}}{a}\right)^{3/4}\,\nonumber \\
& & \times \exp\left[- {2\over 3} \left(\frac{2 M_{12}}{M_{123}}\right)^{1/2}
\left(\frac{R_{\rm min}}{a}\right)^{3/2}\right]\, ,
\label{eq:deltae}
\end{eqnarray}
where the total binary mass $M_{12}=M_1+M_2$, and $M_{123}=M_1+M_2+M_3$ with $M_3$ being the perturber mass. For $V_{\rm min} = \sqrt{GM_{12}/R_{\rm min}} \gg V_\infty$, the impact parameter is of order $b \approx R_{\rm min} V_{\rm min}/V_\infty$ (or else $b \sim R_{\rm min}$). We adopt the set of masses, $M_i = [1.0,0.7,0.3]$ and find $R_{\rm min} \sim 2 a$, if one requires a kick magnitude $\delta e = 0.3$. The adiabatic limit brings about an exponential suppression of the kick. This means we need only to account for the one encounter that has the closest impact parameter. All other encounters do not add substantially.
Consider a cluster with total mass $M_{\rm c}$ and size $R_{\rm c}$, $V_\infty \sim \sqrt{G M_c/R_c}$, number density of stars $N_c = M_{\rm c}/M_*/\pi R_c^3 $. The mean-free-time to have an impact such that $R_{\rm min} \sim 2a$ is
\begin{eqnarray}
t & \sim & \frac{1}{n (\pi b^2) V_\infty}
\sim
4\times 10^8 {\rm yrs}
\left(\frac{N_c}{10^5/{\rm pc}^3}\right)^{-1}
\left(\frac{V_\infty}{10\km/\s}\right)\,\nonumber \\
& & \times
\left(\frac{M_{12}}{1.7M_\odot}\right)^{-1}
\left(\frac{a}{1{\rm AU}}\right)^{-1} \, .
\label{eq:meanfreetime}
\end{eqnarray}
where we have scaled the cluster density and velocity by values appropriate for a very dense region, the core of 47 Tuc
% the very dense core of globular cluster 47 Tuc has $V_\infty = 12\km/\s$, stellar number density of $1.5\times 10^5/{\rm pc}^3$
%47 Tucanae is a massive and quite dense cluster. The central density is about 1:5 x 10^5 pc^-3 with a central velocity dispersion of 12 km/s. The core radius is about 0.5 pc and the half-mass radius about 3.9pc At the half-mass radius the number density of stars is about a factor of ten lower than in the core.
\citep{McLaughlin2006}. Most stars are formed in much less dense clusters that dissolves in tens to hundreds of million years. Outside these birth clusters, the density is much lower and the impact is even smaller.
So overall, AU-scale binaries are unlikely to have been affected by passing-by stars \citep{Parker2009}.
%Galactic fields have $N_c \sim 1$, $V \sim 20\km/s$; Taurus has $N\sim 100$ and $V < 1 $; while high-density regions are more like the archetypal ‘cluster’, with densities of 103–105 stars pc−3 (e.g., from Orion to Westerlund 1).
Even if the above estimate is wrong, we argue that an origin in stellar scattering can be excluded. First, scattering tends to affect the wider binaries more, while the observed Rayleigh mode is largely invariant with the orbital period. Second, if stellar scattering is so strong as to produce $e\sim 0.3$ for AU-scale binaries, it would also have dissolved all binaries that are a few times wider \citep{Heggie1975,Spurzem2009}.
%\section{How does a Rayleigh arise from many Gaussian Processes}
%\label{sec:rayleigh}
\section{Numerical Simulations of Brown Dwarf Ejections}\label{sec:scatterings}
We simulate the gravitational interactions between stellar binary and multiple brown dwarfs, with a focus on the effects on binary eccentricity.
The binary is composed of a $0.7 M_\odot$ primary and a $0.3 M_\odot$ secondary. They are initialized with a circular orbit with $a=a_{\rm bin,init} = 1$AU. A crowd ($N_{\rm bd}$) of brown dwarfs each with mass $M_{\rm bd}$ is uniformly sprinkled in logarithmic distance (from $0.03 a_{\rm bin,init}$ to $0.9 a_{\rm bin,init}$) between the binary stars. All initial orbital angles are drawn randomly. The initial eccentricities are zero, and the mutual inclinations are drawn from a Rayleigh distribution with $\sigma_i = 10^{-3}$rad. {\w For our choice of parameters, ejections typically occur before physical collisions do. So} we ignore collisions and set all physical sizes of the particles to zero.
%{\w The stars have $0.7$ and $0.3R_\odot$ and perfect mergers are assumed at collisions.}
We integrate the system using the IAS15 integrator in REBOUND \citep{Rein2012,Rein2015}.
We carry out 100 simulations for each of the following two sets of parameters: $N_{\rm bd}=6$, $M_{\rm bd}=0.05 M_\odot$; $N_{\rm bd}=30$, $M_{\rm bd}=0.01 M_\odot$. The dynamics are similar. {\w The integrations are carried out for tens of thousands of orbits.} Within a short time (typically tens of orbits), most of the brown dwarfs have been ejected. The remaining few may become bound to one or the other stars. During these ejections, the initial conditions are quickly erased, and we observe that the binaries are hardened and become eccentric. The binary e-distribution is Rayleigh in form, with best-fit modes $\sigma_e = 0.19$ and $0.35$, for the two cases respectively. {\w We find that the binary plane remains largely unchanged, because the initial orbits are largely aligned and most scatterings occur far from the stars and are 2-D in nature.}
%\x{The binary inclinations are also disturbed from the original plane.} %{\y Sam: is this true? any idea how much? I am adding some discussion about it in main text, can check by spin-orbit angle...}
%{\w We integrated long enough (tens of thousands of orbit) till there is nothing much left in the system (one brown-dwarf left sometimes, bound to one of the stars); and the initial inclination is taken to be small but non-zero so that the encounters are not perfectly co-planar. The initial condition is, however, quickly forgotten. So it’s exact choice is not important.
The Rayleigh mode is sensitive only to the individual brown dwarf mass, as indicated by eq.\ (\ref{eq:equipartition}). The number of brown dwarfs is not relevant. In fact, even scattering as few as one brown dwarf may be sufficient to achieve the same e-distribution, as is observed in the 2-planet scattering experiments of \citet{Ford2008}.
%Table 1 of that paper,
% both \langle e \rangle rises, and
% only scattering one
%{\y will be good to have proof, but not essential.}
%Our experiments do not account for mergers. This is reasonable, as dynamical ejection occurs too fast to allow very close encounters.
%\section{Brown Dwarf Desert}\label{sec:A6}
\section{Enough free-floating Brown Dwarfs?}\label{sec:enoughbd}
If close binaries acquire their eccentricities by ejecting brown dwarfs, one expects to see those ejected bodies.
We crudely estimate their contributions to the Galactic mass function, for two limiting cases. In the first case (`excitation') we allow one brown dwarf per system, enough to excite the observed eccentricity. In the second case (`hardening') we assume that all close binaries were originally wide, but scatter enough brown dwarfs to wind up at their current separations.
Solar-type binaries in the field follow a log-normal distribution in orbital periods \citep{DM91,Raghavan2010,Moe2019}.
\begin{equation}
{{df_{\rm binary}}\over{d\log P}}\approx {{f_{\rm binary}}\over{\sqrt{2\pi} \sigma_{\log P}}}\,
\exp\left[ -{1\over 2} \left(\frac{\log P - \log P_0}{\sigma_{\log P}}\right)^2\right]
\label{eq:Pdist}
\end{equation}
with the total binary fraction $f_{\rm binary} \sim 0.50$, $\log P_0\sim 5$ (or $\sim 40$au), and $\sigma_{\log P} \sim 2.3$. Here, all logarithms are 10-based, and all periods are in unit of days.
{\w Counting binaries inwards of $50$au, we find a total number fraction of $f \approx 0.25$. So if one ejected brown dwarf is responsible for exciting the eccentricity in one binary, we expect a brown dwarf to star ratio of $\sim 1/4$.}
%\footnote{Appendix \ref{sec:model-compare} shows that about half of these binaries are actually thermally distributed. So this is an over-estimate.}
%We adopt $M_1 \sim 1 M_\odot$ and a companion mass of $M_2 \sim 0.5 M_\odot$.
%For `excitation', the total mass of ejected bodies relative to that in all Sun-like stars (binary and single) is
%\begin{eqnarray} f_{\rm excitation} & \equiv & {{\displaystyle\sum_{\rm binary} M_{\rm bd}}\over{\displaystyle\sum_{\rm single} M_1 + \displaystyle\sum_{\rm binary} (M_1+M_2)}}\nonumber \\ & \approx & {1\over 10} {{M_2}\over{M_1}} \int^{P_0}_{P=10d} {{df_{\rm binary}}\over{d\log P}} d\log P \,\nonumber\\ &\approx\, & 0.01\, ,
%\label{eq:excitation}
%\end{eqnarray}
%where we have adopted $M_{\rm bd}\sim M_2/10$ (eq. \ref{eq:equi2}) and included all binaries from $10$d to $P_0 = 10^5$d.\footnote{Appendix \ref{sec:model-compare} shows that about half of these binaries are actually thermally distributed. So this is an over-estimate.}
To harden binaries from an initial separation $a_0$ to a new separation $a \ll a_0$, the amount of mass ejected (with parabolic orbits) is $M_{\rm ejected} \sim \ln (a_0/a)\, M_2 \sim 2/3 \ln (P_0/P) \, M_2$. Taking $M_2/M_{\rm bd} \sim 10$, the number ratio of brown dwarfs to stars is
%(Integrating over the same period range yields
\begin{eqnarray}
%f_{\rm hardening} & \equiv & {{\displaystyle\sum_{\rm binary} M_{\rm bd}}\over{\displaystyle\sum_{\rm single} M_1 +\displaystyle\sum_{\rm binary}(M_1+M_2)}} \nonumber \\ & \approx & {2\over 3} {{M_2}\over{M_1}} \int^{P_0}_{P=10d} {{df_{\rm binary}}\over{d\log P}} \times (\ln P_0 - \ln P)\, d\log P \nonumber \\ & \approx & 0.30\, .
% the integral itself is 0.36,
% while dNdlogP integrate to 0.23 of all sun-like stars; 0.23*2/3*0.7/1=0.17
\frac{N_{\rm bd}}{N_*} %& \equiv & {{\displaystyle\sum_{\rm binary} M_{\rm bd}}\over{\displaystyle\sum_{\rm single} M_1 +\displaystyle\sum_{\rm binary}(M_1+M_2)}} \nonumber \\
& \approx & {2\over 3} \times 10 \times\int^{P_0}_{P=10d} {{df_{\rm binary}}\over{d\log P}} \times (\ln P_0 - \ln P)\, d\log P \nonumber \\ & \approx & 5.5\, .
\label{eq:hardening}
\end{eqnarray}
We compile observational constraints using two different types of studies: those that count objects in young stellar clusters, and those that employ microlensing. Using data from {\w Pleiades} (120 Myrs old), \citet{Moraux2003} showed that the mass function at the low end can be well fit by a log-normal distribution (their eq. 3). In even younger clusters ($\sigma$ Ori at 3 Myrs, Upper Sco at 5-10 Myrs), there appears to be more free-floating objects than described by this log-normal form: for masses below $\sim 0.05 M_\odot$ \citep{Lodieu2013,Bejar2011} found a mass-function that roughly goes as $dN/d\log M \propto M^0$. So we append a flat tail to the above log-normal distribution. For the microlensing data, we adopt solution CR1 in \citet{Sumi2023}, inferred using lensing events from the Galactic bulge. These two mass functions differ in form (left panel of Fig. \ref{fig:nomass}), but they give roughly the same {\w results:
%values in cumulative mass (right panel).
the number of brown dwarfs to stars is $\sim 1/3$. This is similar to that inferred by \citet{Kirkpatrick2024} from a population census in the local 20\,pc volume, $1/4$, possibly rising to $1/3$. This means there are enough brown dwarfs to excite the eccentricities, but likely not enough to harden all close-binaries.}
%is also shown in Fig. \ref{fig:nomass}. It falls much below the other two in the sub-stellar range, but may be plagued by completeness and luminosity evolution. Even for this case, the cumulative number in the brown dwarf range is $\sim 1/4$ of the star. They may suffer incompleteness below.... Fig. now use cumulative number, not cumulative mass, makes life easier.}
%We find that in the sub-stellar regime that is of interest to us, the mass function is top-heavy. Moreover, the mass in the relevant brown dwarf range ($M_{\rm bd} \sim 0.05 M_\odot$) is of order $ 2\%$ of the stellar mass. This means there are enough brown dwarfs to excite binary eccentricity, but far too few to account for binary hardening.
This comparison also suggests that, if the ejection scenario is correct, almost all brown dwarfs should be formed in massive disks, as companions to stellar binaries.
An interesting prediction for such a scenario is that the ejection event may leave evidence in the kinematics of these bodies. They should be leaving their birth systems with a velocity dispersion that is of order the orbital escape velocity. {\w
For a 50\,au binary (peak of the binary) with $M_1 = 0.3M_\odot$ (peak of the initial mass function), this is $\sim 2$ km/ s, comparable to the typical dispersion in many young clusters (e.g., Orion). More studies are warranted.}
\begin{figure*}
\centering
%\includegraphics[width=0.45\textwidth]{data_comparison.eps}
\includegraphics[width=0.95\textwidth]{nomass.pdf}
\caption{Mass functions in the low-mass end, in differential (left panel) and cumulative forms (right panel{\w , cumulative number}). The two determinations are from young clusters and from microlensing of bulge stars. For both cases,
%the cumulative mass below $0.05 M_\odot$ is about $2\%$,
{\w the number ratio between brown dwarfs (taken to be below $0.05 M_\odot$) and stars is $\sim 1/4$,}
enough to excite Solar-type close binaries
%($f_{\rm excite}$),
but not enough to harden them %($f_{\rm harden}$).
}
\label{fig:nomass}
\end{figure*}
\section{Is there only a single Rayleigh?}\label{sec:model-compare}
We have obtained a best-fit Rayleigh of $\sigma_e = 0.30$ for our `gold' sample. But given the large sample size ($N\sim 3000$), it is possible to extract more information. We query the data to determine if it prefers other solutions, in particular, two separate Rayleigh distributions that may arise when there are two different physical processes at play.
In the following, we employ a Bayesian framework to answer this question.
The marginal likelihood, or evidence, that data $D$ are generated by a model $\mathcal{M}$, is given by the integral
\begin{equation}
\mathcal{E}(D|\mathcal{M})
=\int \mathrm{Pr}(\pmb{\theta}|\mathcal{M})\mathrm{Pr}(D|\pmb{\theta},\mathcal{M}) d\pmb{\theta}\, ,
\end{equation}
{\w where $\pmb{\theta}$ is the random variable to be marginalized over.}
The ratio between marginal likelihoods for two different models then gives a Bayes factor that provides one metric for how much more (or less) strongly a given model is supported by the data. In our case the data $D$ are binary eccentricities in the `gold' sample. For a given data set $D=\{e_1,...,e_i,...,e_N\}$ we incorporate observational uncertainty by writing $\mathrm{Pr}(D|\pmb{\theta},\mathcal{M})
=\prod_{i=1}^N\int \mathrm{Pr}(e_i|e_\textrm{true})
\mathrm{Pr}(e_\textrm{true}|\pmb{\theta},\mathcal{M})
\text{d}e_\textrm{true},$ where we adopt $\mathrm{Pr}(e_i|e_\textrm{true})$ to be a truncated Gaussian distribution with a standard deviation of $0.025$.
\subsection{Single Rayleigh}
Model $\mathcal{M}_1$ assumes that the eccentricities of the entire population are distributed according to a single Rayleigh distribution,
\begin{eqnarray}
\mathrm{Pr}(D|\pmb{\theta}_e,\mathcal{M}_1) & = & \prod_{i=1}^N \frac{e_i}{\sigma_e^2}\exp\left(-\frac{e_i^2}{2\sigma_e^2}\right)
\nonumber \\
% =\sigma_e^{-2N}\left(\prod_{i=1}^{N}e_i\right)\exp\left(-\frac{\sum_{i=1}^N e_i^2}{2\sigma_e^2}\right)
& = &\sigma_e^{-2N}\mathcal{P}\exp\left(-\frac{S}{2\sigma_e^2}\right)\, ,
\end{eqnarray}
where $\mathcal{P}=\prod_i^Ne_i,$ and $S=\sum_i^Ne_i^2.$ For a uniform probability distribution $\mathrm{Pr}(\sigma_e|\mathcal{M}_1)=(\Delta\sigma)^{-1}$, where $\Delta\sigma=\sigma_\textrm{max}-\sigma_\textrm{min}$
for $\sigma_\mathrm{min} \leq \sigma_e \leq \sigma_\mathrm{max}$,
the evidence is then given (in the absence of observational uncertainty) as
\begin{align}
& & \mathcal{E}(D|\mathcal{M}_1)
=(\Delta\sigma)^{-1}
\int_{\sigma_\mathrm{min}}^{\sigma_\mathrm{max}}
\mathcal{P}
\sigma_e^{-2N}\exp\left(
-\frac{S}{2\sigma_e^2}
\right)d\sigma_e
\\\notag
& &=(\Delta\sigma)^{-1}2^{N-3/2} S^{1/2-N}\mathcal{P}
\Gamma \left(N-1/2,\frac{S}{2 \sigma_e^2}\right)
\Bigg|^{\sigma_\mathrm{max}}_{\sigma_\mathrm{min}}\, ,
\end{align}
where $\Gamma$ is the incomplete Gamma function. We find that, for our `gold' sample, the likelihood $\textrm{Pr}(D_\textrm{Gaia}|\sigma_e,\mathcal{M})$
%\approx\delta(\sigma_e-0.30)$
approaches a singularly peaked function at $\sigma_e = 0.30$. So $\mathcal{E}(D_\textrm{Gaia}|\mathcal{M}_1)\gg1$ for any choices of $\sigma_\textrm{max}$ and $\sigma_\textrm{min}$, regardless of the observational uncertainties.
%Compared to a 'uniform' distribution,
%Since the evidence for a uniform eccentricity distribution is $\mathcal{E}_\text{uni}=1$ by construction, this implies Bayes factors $\mathcal{E}(D_\textrm{Gaia}|\mathcal{M}_1)/\mathcal{E}_\text{uni}$ that are similarly $\gg1$.
\subsection{Double Rayleigh?}
%While our Gaia sample strongly prefers a model with a single Rayleigh distribution.
Can the Gaia sample be instead drawn from two separate Rayleigh distributions? %differentiate from a model in which members of the population are drawn from multiple Rayleigh distributions?
We consider a mixture model $\mathcal{M}_2$ with
\begin{eqnarray}
\mathrm{Pr}(D|\sigma_1,\sigma_2,f,\mathcal{M}_2)
& = & \prod_{i=1}^N
\left[
f\dfrac{e_i}{\sigma_1^2}\exp\left(-\dfrac{e_i^2}{2\sigma_1^2}\right)
\right.\nonumber \\
& & \left. +(1-f)\dfrac{e_i}{\sigma_2^2}\exp\left(-\dfrac{e_i^2}{2\sigma_2^2}\right)
\right],
\end{eqnarray}
where $0\leq f\leq1.$ Assuming
$\textrm{Pr}(\sigma_1,\sigma_2,f|\mathcal{M}_2)
=1/(\Delta \sigma)^2$, we compute the evidence $\mathcal{E}(D_\textrm{Gaia}|\mathcal{M}_2)$ numerically (using Clenshaw-Curtis quadratures). For like values of $\sigma_\textrm{max}$ and $\sigma_\textrm{min},$ we find $\mathcal{E}(D_\textrm{Gaia}|\mathcal{M}_1)/\mathcal{E}(D_\textrm{Gaia}|\mathcal{M}_2)\sim2-5$ when $\sigma_\textrm{min}\gtrsim0.1$ and $\sigma_\textrm{max}>0.303.$ This does not provide particularly strong support for a single Rayleigh distribution over a double one. However, for any intermediate mixing fraction $0<f<1$, we find that the likelihood $\mathcal{L}=\textrm{Pr}(D_\textrm{Gaia}|\sigma_1,\sigma_2,f,\mathcal{M}_2)$ peaks at $\sigma_1\approx\sigma_2\approx0.303$, which simply corresponds to a single Rayleigh distribution with the same mode; the flexibility of the mixture model provides no added utility. %Accordingly, we find Bayesian information criteria (which penalize free parameters) that are smaller by about ten for $\mathcal{M}_1$ than for $\mathcal{M}_2.$
When $\sigma_\textrm{min}\lesssim0.05$, the Gaia data provide a stronger evidence for a double Rayleigh model with $\sigma_1=0.303,$ $\sigma_2\ll1$, and $1-f\ll1$ (i.e., mostly the primary Rayleigh mode $\sigma_1=0.303$). However, we discount this model for two reasons. First, it amounts to an over-fit of the small eccentricity data. Second, there is a strange (but small) excess of $e=0$ points in the data (as seen in the left panel of Fig. \ref{fig:x_extrapolate}) that likely reflects issues in the Gaia pipeline.
}
\end{appendix}
\end{document}
\newpage
\begin{eqnarray}
{{dN}\over{d x}} & \propto & \,\, \exp\left(- \frac{x^2}{2\sigma^2}\right)\, \nonumber \\
{{dN}\over{d y}} & \propto & \,\, \exp\left(-\frac{y^2}{2 \sigma^2}\right)\, .\nonumber
\end{eqnarray}
\begin{equation}
e =\sqrt{x^2 + y^2}\, , \hskip0.5in
{{dN}\over{de}} \propto e \, \exp\left(- \frac{e^2}{2\sigma^2}\right)\, .\nonumber
\end{equation}
\begin{eqnarray}
{{dN}\over{d x}} & \propto & \,\, \exp\left(- \frac{x^2}{2\sigma^2}\right)\, \nonumber \\
{{dN}\over{d y}} & \propto & \,\, \exp\left(-\frac{y^2}{2\sigma^2}\right)\nonumber \\
{{dN}\over{d z}} & \propto & \,\, \exp\left(-\frac{z^2}{2 \sigma^2}\right)\, .\nonumber
\end{eqnarray}
\begin{equation}
v =\sqrt{x^2 + y^2 + z^2}\,
\nonumber
\end{equation}
\vskip0.5in
\begin{equation}
{{dN}\over{dv}} \propto v^2 \, \exp\left(- \frac{v^2}{2\sigma^2}\right)\, .\nonumber
\end{equation}
\begin{equation}
\vec{\bf e}
\nonumber
\end{equation}
%\subsection{Rayleigh and thermal}
\section{Binaries at Longer Periods?}
\label{sec:longperiod}
In the following, we ask whether the same Rayleigh distribution extends to binaries outside $10^3$d.
Knowing that wide binaries have thermal distributions \citep{Tokovinin2020,Hwang2021}, we can imagine at least two possibilities: the Rayleigh may persist but with an increasing mode further out, smoothly connecting up to the thermal distribution;\footnote{Formally, the thermal distribution, $dN/de \propto e$, is a Rayleigh distribution with $\sigma_e \rightarrow \infty$.} or both distributions co-exist, with the fraction of Rayleigh (with the same mode) gradually going down to zero. Given the invariance of the Rayleigh mode in our `gold' sample, we choose to focus on the second possibility. We draw some partial conclusions based on two different samples.
The first is the previous ground-based sample from \cite{Raghavan2010}, which includes $58$ binaries in the period range $10^3- 10^5$d (or $\sim 2 - 40$AU). {\y need to refer to other appendixes} We consider a model $\mathcal{M}_3$ in which the eccentricities are distributed according to a mixture of Rayleigh and thermal:
\begin{equation}
\mathrm{Pr}(D|\sigma_e,f,\mathcal{M}_3)
=\prod_{i=1}^N \left[
f\dfrac{e_i}{\sigma_e^2}
\exp\left(-\dfrac{e_i^2}{2\sigma_e^2}\right)
+2(1-f)e_i
\right]\, ,
\end{equation}
with $f$ being the Rayleigh fraction.
We adopt a flat prior for $f$ and a prior $\mathrm{Pr}(\sigma_e|\mathcal{M}_3)=\delta(\sigma_e-0.303),$ which equates to conditioning based on the knowledge gained from our Gaia sample\footnote{A uniform prior in $\sigma_e$ gives similar results, as long as $\sigma_\textrm{min}<0.303<\sigma_\textrm{max}$.}.
We then compute the evidence $\mathcal{E}_\textrm{mix}=\mathcal{E}(D_\textrm{LP}|\mathcal{M}_3)$ for the long-period sample. Our results indicate that a mixture fraction of $f=0.52$ best describes the data. Comparing against evidence values computed for a purely thermal, a purely uniform (%in the absence of any observational uncertainty
$\mathcal{E}_\textrm{therm}=2^N\mathcal{P}$ and $\mathcal{E}_\textrm{uni}=1$, respectively), and a purely Rayleigh distributions, we find Bayes factors $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{therm}\approx6.5\times10^3$, $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{uni}\approx12.3$, and $\mathcal{E}_\textrm{mix}/\mathcal{E}(D_\textrm{LP}|\mathcal{M}_1)\approx20.2.$ These factors indicate that the mixture model $\mathcal{M}_3$ is much better supported by the long-period data than a thermal eccentricity distribution, and marginally better supported than a uniform or a Rayleigh distribution.
%These results are graphically displayed in Fig. \ref{fig:mix2}.
%We find \autoref{fig:mix2} compares cumulative distributions for the Gaia and spectroscopic binary samples with predictions for model $\mathcal{M}_1$ (black), model $\mathcal{M}_3$ (orange), and a uniform distribution.
%{\y figure removed, need revision}
%\begin{figure}
% \centering
%\includegraphics[width=0.45\textwidth]{moe2017_CDF_fit.pdf}
% \caption{The cumulative histogram of eccentricity distribution for the long-period ($10^3-10^5$d) binary sample from \citet{Raghavan2010,Moe2017}. It best supports a model that is a mixture of a Rayleigh and a thermal distribution (smooth orange curve). The other curves show the variety of models we compare against (thermal, uniform, and pure Rayleigh). %with Bayes factors that $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{therm}\approx6.5\times10^3$, $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{uni}\approx12.3$, $\mathcal{E}_\textrm{mix}/\mathcal{E}_\textrm{Rayleigh}\approx 20.2$,
% }
% \label{fig:mix2}
%\end{figure}
The second sample we consider comes from resolved Gaia binaries \citep{Elbadry2021}. \citet{Hwang2022} established, statistically, an eccentricity distribution by measuring the instantaneous velocity-position ($v-r$) angles that are projected on the sky %, where $v$ and $r$ are the projected vectors on the sky plane
\citep[see also][]{Tokovinin2020}. Their results are shown in Fig. \ref{fig:mixed} for binaries with projected separations that fall within $10-1000$AU.
%\cite{Tokovinin2020} used a sub-sample ($\sim 3600$) of the \cite{Elbadry2018} binary catalogue, median mass $0.6 M_\odot$, median separation $450$au, goes down to $100$;
%i.e., proper-motion, not RV
%Tokovinin2020 actually used both $v-r$ angle and orbital speed
In the absence of a better model, they parameterized the eccentricity distribution as a power-law, $dN/de \propto e^\eta$ and found that the value of $\eta$ rises from 0 (`uniform') at $\sim 50$ au, to $1$ (`thermal') at $\sim 500$au, and $>1$ (`super-thermal') beyond. Alternatively, as we show in Fig. \ref{fig:mixed}, the so-called `uniform' distribution can easily be swapped for our above mixture model. The data quality is not sufficient to tell the two apart.
These two pieces of evidence, taken together, suggest that the Rayleigh distribution may persist to at least tens of AU, and the apparent change in $\eta$ with period \citep{Hwang2022} may simply reflect the decreasing fraction of the Rayleigh with increasing period.
\begin{figure}
\centering
%\includegraphics[width=0.45\textwidth]{data_comparison.eps}
\includegraphics[width=0.45\textwidth]{gamma_dist.pdf}
\caption{Distribution in the sky-projected displacement-velocity angle $\gamma$ for resolved binaries. The observed values (thick histograms for different values of projected separation $s$) are adopted from Fig. 6 of \citet{Hwang2022}, and different models are in thin lines, with our best-fit mixture model (\S \ref{sec:model-compare}) in red.
Wide binaries ($10^2 < s < 10^3$au) are best described by a thermal distribution, but close binaries ($10^1 < s < 10^2$au)
can be described equally well by a uniform distribution ($dN/de = {\rm const}$) or by our mixture model.
% In other words, an apparent 'uniform' distribution results when the Rayleigh distribution is gradually replaced by a thermal distribution around 10s' of AU.
}
\label{fig:mixed}
\end{figure}
\section{Equations}\label{sec4}
Equations in \LaTeX\ can either be inline or on-a-line by itself (``display equations''). For
inline equations use the \verb+$...$+ commands. E.g.: The equation
$H\psi = E \psi$ is written via the command \verb+$H \psi = E \psi$+.
For display equations (with auto generated equation numbers)
one can use the equation or align environments:
\begin{equation}
\|\tilde{X}(k)\|^2 \leq\frac{\sum\limits_{i=1}^{p}\left\|\tilde{Y}_i(k)\right\|^2+\sum\limits_{j=1}^{q}\left\|\tilde{Z}_j(k)\right\|^2 }{p+q}.\label{eq1}
\end{equation}
where,
\begin{align}
D_\mu &= \partial_\mu - ig \frac{\lambda^a}{2} A^a_\mu \nonumber \\
F^a_{\mu\nu} &= \partial_\mu A^a_\nu - \partial_\nu A^a_\mu + g f^{abc} A^b_\mu A^a_\nu \label{eq2}
\end{align}
Notice the use of \verb+\nonumber+ in the align environment at the end
of each line, except the last, so as not to produce equation numbers on
lines where no equation numbers are required. The \verb+\label{}+ command
should only be used at the last line of an align environment where
\verb+\nonumber+ is not used.
\begin{equation}
Y_\infty = \left( \frac{m}{\textrm{GeV}} \right)^{-3}
\left[ 1 + \frac{3 \ln(m/\textrm{GeV})}{15}
+ \frac{\ln(c_2/5)}{15} \right]
\end{equation}
The class file also supports the use of \verb+\mathbb{}+, \verb+\mathscr{}+ and
\verb+\mathcal{}+ commands. As such \verb+\mathbb{R}+, \verb+\mathscr{R}+
and \verb+\mathcal{R}+ produces $\mathbb{R}$, $\mathscr{R}$ and $\mathcal{R}$
respectively (refer Subsubsection~\ref{subsubsec2}).
\section{Tables}\label{sec5}
Tables can be inserted via the normal table and tabular environment. To put
footnotes inside tables you should use \verb+\footnotetext[]{...}+ tag.
The footnote appears just below the table itself (refer Tables~\ref{tab1} and \ref{tab2}).
For the corresponding footnotemark use \verb+\footnotemark[...]+
\begin{table}[h]
\caption{Caption text}\label{tab1}%
\begin{tabular}{@{}llll@{}}
\toprule
Column 1 & Column 2 & Column 3 & Column 4\\
\midrule
row 1 & data 1 & data 2 & data 3 \\
row 2 & data 4 & data 5\footnotemark[1] & data 6 \\
row 3 & data 7 & data 8 & data 9\footnotemark[2] \\
\botrule
\end{tabular}
\footnotetext{Source: This is an example of table footnote. This is an example of table footnote.}
\footnotetext[1]{Example for a first table footnote. This is an example of table footnote.}
\footnotetext[2]{Example for a second table footnote. This is an example of table footnote.}
\end{table}
\noindent
The input format for the above table is as follows:
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\bigskip
\begin{verbatim}
\begin{table}[<placement-specifier>]
\caption{<table-caption>}\label{<table-label>}%
\begin{tabular}{@{}llll@{}}
\toprule
Column 1 & Column 2 & Column 3 & Column 4\\
\midrule
row 1 & data 1 & data 2 & data 3 \\
row 2 & data 4 & data 5\footnotemark[1] & data 6 \\
row 3 & data 7 & data 8 & data 9\footnotemark[2]\\
\botrule
\end{tabular}
\footnotetext{Source: This is an example of table footnote.
This is an example of table footnote.}
\footnotetext[1]{Example for a first table footnote.
This is an example of table footnote.}
\footnotetext[2]{Example for a second table footnote.
This is an example of table footnote.}
\end{table}
\end{verbatim}
\bigskip
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\begin{table}[h]
\caption{Example of a lengthy table which is set to full textwidth}\label{tab2}
\begin{tabular*}{\textwidth}{@{\extracolsep\fill}lcccccc}
\toprule%
& \multicolumn{3}{@{}c@{}}{Element 1\footnotemark[1]} & \multicolumn{3}{@{}c@{}}{Element 2\footnotemark[2]} \\\cmidrule{2-4}\cmidrule{5-7}%
Project & Energy & $\sigma_{calc}$ & $\sigma_{expt}$ & Energy & $\sigma_{calc}$ & $\sigma_{expt}$ \\
\midrule
Element 3 & 990 A & 1168 & $1547\pm12$ & 780 A & 1166 & $1239\pm100$\\
Element 4 & 500 A & 961 & $922\pm10$ & 900 A & 1268 & $1092\pm40$\\
\botrule
\end{tabular*}
\footnotetext{Note: This is an example of table footnote. This is an example of table footnote this is an example of table footnote this is an example of~table footnote this is an example of table footnote.}
\footnotetext[1]{Example for a first table footnote.}
\footnotetext[2]{Example for a second table footnote.}
\end{table}
In case of double column layout, tables which do not fit in single column width should be set to full text width. For this, you need to use \verb+\begin{table*}+ \verb+...+ \verb+\end{table*}+ instead of \verb+\begin{table}+ \verb+...+ \verb+\end{table}+ environment. Lengthy tables which do not fit in textwidth should be set as rotated table. For this, you need to use \verb+\begin{sidewaystable}+ \verb+...+ \verb+\end{sidewaystable}+ instead of \verb+\begin{table*}+ \verb+...+ \verb+\end{table*}+ environment. This environment puts tables rotated to single column width. For tables rotated to double column width, use \verb+\begin{sidewaystable*}+ \verb+...+ \verb+\end{sidewaystable*}+.
\begin{sidewaystable}
\caption{Tables which are too long to fit, should be written using the ``sidewaystable'' environment as shown here}\label{tab3}
\begin{tabular*}{\textheight}{@{\extracolsep\fill}lcccccc}
\toprule%
& \multicolumn{3}{@{}c@{}}{Element 1\footnotemark[1]}& \multicolumn{3}{@{}c@{}}{Element\footnotemark[2]} \\\cmidrule{2-4}\cmidrule{5-7}%
Projectile & Energy & $\sigma_{calc}$ & $\sigma_{expt}$ & Energy & $\sigma_{calc}$ & $\sigma_{expt}$ \\
\midrule
Element 3 & 990 A & 1168 & $1547\pm12$ & 780 A & 1166 & $1239\pm100$ \\
Element 4 & 500 A & 961 & $922\pm10$ & 900 A & 1268 & $1092\pm40$ \\
Element 5 & 990 A & 1168 & $1547\pm12$ & 780 A & 1166 & $1239\pm100$ \\
Element 6 & 500 A & 961 & $922\pm10$ & 900 A & 1268 & $1092\pm40$ \\
\botrule
\end{tabular*}
\footnotetext{Note: This is an example of table footnote this is an example of table footnote this is an example of table footnote this is an example of~table footnote this is an example of table footnote.}
\footnotetext[1]{This is an example of table footnote.}
\end{sidewaystable}
\section{Figures}\label{sec6}
As per the \LaTeX\ standards you need to use eps images for \LaTeX\ compilation and \verb+pdf/jpg/png+ images for \verb+PDFLaTeX+ compilation. This is one of the major difference between \LaTeX\ and \verb+PDFLaTeX+. Each image should be from a single input .eps/vector image file. Avoid using subfigures. The command for inserting images for \LaTeX\ and \verb+PDFLaTeX+ can be generalized. The package used to insert images in \verb+LaTeX/PDFLaTeX+ is the graphicx package. Figures can be inserted via the normal figure environment as shown in the below example:
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\bigskip
\begin{verbatim}
\begin{figure}[<placement-specifier>]
\centering
\includegraphics{<eps-file>}
\caption{<figure-caption>}\label{<figure-label>}
\end{figure}
\end{verbatim}
\bigskip
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\begin{figure}[h]
\centering
\includegraphics[width=0.9\textwidth]{fig.eps}
\caption{This is a widefig. This is an example of long caption this is an example of long caption this is an example of long caption this is an example of long caption}\label{fig1}
\end{figure}
In case of double column layout, the above format puts figure captions/images to single column width. To get spanned images, we need to provide \verb+\begin{figure*}+ \verb+...+ \verb+\end{figure*}+.
For sample purpose, we have included the width of images in the optional argument of \verb+\includegraphics+ tag. Please ignore this.
\section{Algorithms, Program codes and Listings}\label{sec7}
Packages \verb+algorithm+, \verb+algorithmicx+ and \verb+algpseudocode+ are used for setting algorithms in \LaTeX\ using the format:
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\bigskip
\begin{verbatim}
\begin{algorithm}
\caption{<alg-caption>}\label{<alg-label>}
\begin{algorithmic}[1]
. . .
\end{algorithmic}
\end{algorithm}
\end{verbatim}
\bigskip
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
You may refer above listed package documentations for more details before setting \verb+algorithm+ environment. For program codes, the ``verbatim'' package is required and the command to be used is \verb+\begin{verbatim}+ \verb+...+ \verb+\end{verbatim}+.
Similarly, for \verb+listings+, use the \verb+listings+ package. \verb+\begin{lstlisting}+ \verb+...+ \verb+\end{lstlisting}+ is used to set environments similar to \verb+verbatim+ environment. Refer to the \verb+lstlisting+ package documentation for more details.
A fast exponentiation procedure:
\lstset{texcl=true,basicstyle=\small\sf,commentstyle=\small\rm,mathescape=true,escapeinside={(*}{*)}}
\begin{lstlisting}
begin
for $i:=1$ to $10$ step $1$ do
expt($2,i$);
newline() od (*\textrm{Comments will be set flush to the right margin}*)
where
proc expt($x,n$) $\equiv$
$z:=1$;
do if $n=0$ then exit fi;
do if odd($n$) then exit fi;
comment: (*\textrm{This is a comment statement;}*)
$n:=n/2$; $x:=x*x$ od;
{ $n>0$ };
$n:=n-1$; $z:=z*x$ od;
print($z$).
end
\end{lstlisting}
\begin{algorithm}
\caption{Calculate $y = x^n$}\label{algo1}
\begin{algorithmic}[1]
\Require $n \geq 0 \vee x \neq 0$
\Ensure $y = x^n$
\State $y \Leftarrow 1$
\If{$n < 0$}\label{algln2}
\State $X \Leftarrow 1 / x$
\State $N \Leftarrow -n$
\Else
\State $X \Leftarrow x$
\State $N \Leftarrow n$
\EndIf
\While{$N \neq 0$}
\If{$N$ is even}
\State $X \Leftarrow X \times X$
\State $N \Leftarrow N / 2$
\Else[$N$ is odd]
\State $y \Leftarrow y \times X$
\State $N \Leftarrow N - 1$
\EndIf
\EndWhile
\end{algorithmic}
\end{algorithm}
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\bigskip
\begin{minipage}{\hsize}%
\lstset{frame=single,framexleftmargin=-1pt,framexrightmargin=-17pt,framesep=12pt,linewidth=0.98\textwidth,language=pascal}% Set your language (you can change the language for each code-block optionally)
%%% Start your code-block
\begin{lstlisting}
for i:=maxint to 0 do
begin
{ do nothing }
end;
Write('Case insensitive ');
Write('Pascal keywords.');
\end{lstlisting}
\end{minipage}
\section{Cross referencing}\label{sec8}
Environments such as figure, table, equation and align can have a label
declared via the \verb+\label{#label}+ command. For figures and table
environments use the \verb+\label{}+ command inside or just
below the \verb+\caption{}+ command. You can then use the
\verb+\ref{#label}+ command to cross-reference them. As an example, consider
the label declared for Figure~\ref{fig1} which is
\verb+\label{fig1}+. To cross-reference it, use the command
\verb+Figure \ref{fig1}+, for which it comes up as
``Figure~\ref{fig1}''.
To reference line numbers in an algorithm, consider the label declared for the line number 2 of Algorithm~\ref{algo1} is \verb+\label{algln2}+. To cross-reference it, use the command \verb+\ref{algln2}+ for which it comes up as line~\ref{algln2} of Algorithm~\ref{algo1}.
\subsection{Details on reference citations}\label{subsec7}
Standard \LaTeX\ permits only numerical citations. To support both numerical and author-year citations this template uses \verb+natbib+ \LaTeX\ package. For style guidance please refer to the template user manual.
Here is an example for \verb+\cite{...}+: \cite{bib1}. Another example for \verb+\citep{...}+: \citep{bib2}. For author-year citation mode, \verb+\cite{...}+ prints Jones et al. (1990) and \verb+\citep{...}+ prints (Jones et al., 1990).
All cited bib entries are printed at the end of this article: \cite{bib3}, \cite{bib4}, \cite{bib5}, \cite{bib6}, \cite{bib7}, \cite{bib8}, \cite{bib9}, \cite{bib10}, \cite{bib11}, \cite{bib12} and \cite{bib13}.
\section{Examples for theorem like environments}\label{sec10}
For theorem like environments, we require \verb+amsthm+ package. There are three types of predefined theorem styles exists---\verb+thmstyleone+, \verb+thmstyletwo+ and \verb+thmstylethree+
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
\bigskip
\begin{tabular}{|l|p{19pc}|}
\hline
\verb+thmstyleone+ & Numbered, theorem head in bold font and theorem text in italic style \\\hline
\verb+thmstyletwo+ & Numbered, theorem head in roman font and theorem text in italic style \\\hline
\verb+thmstylethree+ & Numbered, theorem head in bold font and theorem text in roman style \\\hline
\end{tabular}
\bigskip
%%=============================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%=============================================%%
For mathematics journals, theorem styles can be included as shown in the following examples:
\begin{theorem}[Theorem subhead]\label{thm1}
Example theorem text. Example theorem text. Example theorem text. Example theorem text. Example theorem text.
Example theorem text. Example theorem text. Example theorem text. Example theorem text. Example theorem text.
Example theorem text.
\end{theorem}
Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text.
\begin{proposition}
Example proposition text. Example proposition text. Example proposition text. Example proposition text. Example proposition text.
Example proposition text. Example proposition text. Example proposition text. Example proposition text. Example proposition text.
\end{proposition}
Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text.
\begin{example}
Phasellus adipiscing semper elit. Proin fermentum massa
ac quam. Sed diam turpis, molestie vitae, placerat a, molestie nec, leo. Maecenas lacinia. Nam ipsum ligula, eleifend
at, accumsan nec, suscipit a, ipsum. Morbi blandit ligula feugiat magna. Nunc eleifend consequat lorem.
\end{example}
Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text.
\begin{remark}
Phasellus adipiscing semper elit. Proin fermentum massa
ac quam. Sed diam turpis, molestie vitae, placerat a, molestie nec, leo. Maecenas lacinia. Nam ipsum ligula, eleifend
at, accumsan nec, suscipit a, ipsum. Morbi blandit ligula feugiat magna. Nunc eleifend consequat lorem.
\end{remark}
Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text.
\begin{definition}[Definition sub head]
Example definition text. Example definition text. Example definition text. Example definition text. Example definition text. Example definition text. Example definition text. Example definition text.
\end{definition}
Additionally a predefined ``proof'' environment is available: \verb+\begin{proof}+ \verb+...+ \verb+\end{proof}+. This prints a ``Proof'' head in italic font style and the ``body text'' in roman font style with an open square at the end of each proof environment.
\begin{proof}
Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text.
\end{proof}
Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text. Sample body text.
\begin{proof}[Proof of Theorem~{\upshape\ref{thm1}}]
Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text. Example for proof text.
\end{proof}
\noindent
For a quote environment, use \verb+\begin{quote}...\end{quote}+
\begin{quote}
Quoted text example. Aliquam porttitor quam a lacus. Praesent vel arcu ut tortor cursus volutpat. In vitae pede quis diam bibendum placerat. Fusce elementum
convallis neque. Sed dolor orci, scelerisque ac, dapibus nec, ultricies ut, mi. Duis nec dui quis leo sagittis commodo.
\end{quote}
Sample body text. Sample body text. Sample body text. Sample body text. Sample body text (refer Figure~\ref{fig1}). Sample body text. Sample body text. Sample body text (refer Table~\ref{tab3}).
\section{Methods}\label{sec11}
Topical subheadings are allowed. Authors must ensure that their Methods section includes adequate experimental and characterization data necessary for others in the field to reproduce their work. Authors are encouraged to include RIIDs where appropriate.
\textbf{Ethical approval declarations} (only required where applicable) Any article reporting experiment/s carried out on (i)~live vertebrate (or higher invertebrates), (ii)~humans or (iii)~human samples must include an unambiguous statement within the methods section that meets the following requirements:
\begin{enumerate}[1.]
\item Approval: a statement which confirms that all experimental protocols were approved by a named institutional and/or licensing committee. Please identify the approving body in the methods section
\item Accordance: a statement explicitly saying that the methods were carried out in accordance with the relevant guidelines and regulations
\item Informed consent (for experiments involving humans or human tissue samples): include a statement confirming that informed consent was obtained from all participants and/or their legal guardian/s
\end{enumerate}
If your manuscript includes potentially identifying patient/participant information, or if it describes human transplantation research, or if it reports results of a clinical trial then additional information will be required. Please visit (\url{https://www.nature.com/nature-research/editorial-policies}) for Nature Portfolio journals, (\url{https://www.springer.com/gp/authors-editors/journal-author/journal-author-helpdesk/publishing-ethics/14214}) for Springer Nature journals, or (\url{https://www.biomedcentral.com/getpublished/editorial-policies\#ethics+and+consent}) for BMC.
\section{Discussion}\label{sec12}
Discussions should be brief and focused. In some disciplines use of Discussion or `Conclusion' is interchangeable. It is not mandatory to use both. Some journals prefer a section `Results and Discussion' followed by a section `Conclusion'. Please refer to Journal-level guidance for any specific requirements.
\section{Conclusion}\label{sec13}
Conclusions may be used to restate your hypothesis or research question, restate your major findings, explain the relevance and the added value of your work, highlight any limitations of your study, describe future directions for research and recommendations.
In some disciplines use of Discussion or 'Conclusion' is interchangeable. It is not mandatory to use both. Please refer to Journal-level guidance for any specific requirements.
\noindent
If any of the sections are not relevant to your manuscript, please include the heading and write `Not applicable' for that section.
%%===================================================%%
%% For presentation purpose, we have included %%
%% \bigskip command. Please ignore this. %%
%%===================================================%%
\bigskip
\begin{flushleft}%
Editorial Policies for:
\bigskip\noindent
Springer journals and proceedings: \url{https://www.springer.com/gp/editorial-policies}
\bigskip\noindent
Nature Portfolio journals: \url{https://www.nature.com/nature-research/editorial-policies}
\bigskip\noindent
\textit{Scientific Reports}: \url{https://www.nature.com/srep/journal-policies/editorial-policies}
\bigskip\noindent
BMC journals: \url{https://www.biomedcentral.com/getpublished/editorial-policies}
\end{flushleft}
%%=================================================================== ========================%%
%% If you are submitting to one of the Nature Portfolio journals, using the eJP submission %%
%% system, please include the references within the manuscript file itself. You may do this %%
%% by copying the reference list from your .bbl file, paste it into the main manuscript .tex %%
%% file, and delete the associated \verb+\bibliography+ commands. %%
%%===========================================================================================%%
\end{document}
\begin{thebibliography}{}
\expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
\providecommand{\url}[1]{\href{#1}{#1}}
\providecommand{\dodoi}[1]{doi:~\href{http://doi.org/#1}{\nolinkurl{#1}}}
\providecommand{\doeprint}[1]{\href{http://ascl.net/#1}{\nolinkurl{http://ascl.net/#1}}}
\providecommand{\doarXiv}[1]{\href{https://arxiv.org/abs/#1}{\nolinkurl{https://arxiv.org/abs/#1}}}
\bibitem[{{B{\'e}jar} {et~al.}(2011){B{\'e}jar}, {Zapatero Osorio}, {Rebolo}, {Caballero}, {Barrado}, {Mart{\'\i}n}, {Mundt}, \& {Bailer-Jones}}]{Bejar2011}
{B{\'e}jar}, V.~J.~S., {Zapatero Osorio}, M.~R., {Rebolo}, R., {et~al.} 2011, \apj, 743, 64, \dodoi{10.1088/0004-637X/743/1/64}
\bibitem[{{Borkovits} {et~al.}(2016){Borkovits}, {Hajdu}, {Sztakovics}, {Rappaport}, {Levine}, {B{\'\i}r{\'o}}, \& {Klagyivik}}]{Borkovits2016}
{Borkovits}, T., {Hajdu}, T., {Sztakovics}, J., {et~al.} 2016, \mnras, 455, 4136, \dodoi{10.1093/mnras/stv2530}
\bibitem[{{Czavalinga} {et~al.}(2023){Czavalinga}, {Mitnyan}, {Rappaport}, {Borkovits}, {Gagliano}, {Omohundro}, {Kristiansen}, \& {P{\'a}l}}]{Cza2023}
{Czavalinga}, D.~R., {Mitnyan}, T., {Rappaport}, S.~A., {et~al.} 2023, \aap, 670, A75, \dodoi{10.1051/0004-6361/202245300}
\bibitem[{{Dittmann} \& {Ryan}(2024)}]{Dittmann2024}
{Dittmann}, A.~J., \& {Ryan}, G. 2024, \apj, 967, 12, \dodoi{10.3847/1538-4357/ad2f1e}
\bibitem[{{D'Orazio} \& {Duffell}(2021)}]{Orazio2021}
{D'Orazio}, D.~J., \& {Duffell}, P.~C. 2021, \apjl, 914, L21, \dodoi{10.3847/2041-8213/ac0621}
\bibitem[{{Downes} {et~al.}(2014){Downes}, {Brice{\~n}o}, {Mateu}, {Hern{\'a}ndez}, {Vivas}, {Calvet}, {Hartmann}, {Petr-Gotzens}, \& {Allen}}]{Downes2014}
{Downes}, J.~J., {Brice{\~n}o}, C., {Mateu}, C., {et~al.} 2014, \mnras, 444, 1793, \dodoi{10.1093/mnras/stu1553}
\bibitem[{{Duch{\^e}ne} \& {Kraus}(2013)}]{Duchene2013}
{Duch{\^e}ne}, G., \& {Kraus}, A. 2013, \araa, 51, 269, \dodoi{10.1146/annurev-astro-081710-102602}
\bibitem[{{Duquennoy} \& {Mayor}(1991)}]{DM91}
{Duquennoy}, A., \& {Mayor}, M. 1991, \aap, 248, 485
\bibitem[{{El-Badry} {et~al.}(2024){El-Badry}, {Lam}, {Holl}, {Halbwachs}, {Rix}, {Mazeh}, \& {Shahaf}}]{elbadry2024}
{El-Badry}, K., {Lam}, C., {Holl}, B., {et~al.} 2024, arXiv e-prints, arXiv:2411.00088.
\newblock \doarXiv{2411.00088}
\bibitem[{El-Badry \& Rix(2019)}]{elbadry_rix_19}
El-Badry, K., \& Rix, H.-W. 2019, Monthly Notices of the Royal Astronomical Society: Letters, 482, L139
\bibitem[{{El-Badry} {et~al.}(2021){El-Badry}, {Rix}, \& {Heintz}}]{Elbadry2021}
{El-Badry}, K., {Rix}, H.-W., \& {Heintz}, T.~M. 2021, \mnras, 506, 2269, \dodoi{10.1093/mnras/stab323}
\bibitem[{{El-Badry} {et~al.}(2019){El-Badry}, {Rix}, {Tian}, {Duch{\^e}ne}, \& {Moe}}]{Elbadrytwin}
{El-Badry}, K., {Rix}, H.-W., {Tian}, H., {Duch{\^e}ne}, G., \& {Moe}, M. 2019, \mnras, 489, 5822, \dodoi{10.1093/mnras/stz2480}
\bibitem[{{Fischer} \& {Marcy}(1992)}]{Fischer1992}
{Fischer}, D.~A., \& {Marcy}, G.~W. 1992, \apj, 396, 178, \dodoi{10.1086/171708}
\bibitem[{{Ford} \& {Rasio}(2008)}]{Ford2008}
{Ford}, E.~B., \& {Rasio}, F.~A. 2008, \apj, 686, 621, \dodoi{10.1086/590926}
\bibitem[{{Gaia Collaboration} {et~al.}(2016){Gaia Collaboration}, {Prusti}, {de Bruijne}, {Brown}, {Vallenari}, {Babusiaux}, {Bailer-Jones}, {Bastian}, {Biermann}, {Evans}, {Eyer}, {Jansen}, {Jordi}, {Klioner}, {Lammers}, {Lindegren}, {Luri}, {Mignard}, {Milligan}, {Panem}, {Poinsignon}, {Pourbaix}, {Randich}, {Sarri}, {Sartoretti}, {Siddiqui}, {Soubiran}, {Valette}, {van Leeuwen}, {Walton}, {Aerts}, {Arenou}, {Cropper}, {Drimmel}, {H{\o}g}, {Katz}, {Lattanzi}, {O'Mullane}, {Grebel}, {Holland}, {Huc}, {Passot}, {Bramante}, {Cacciari}, {Casta{\~n}eda}, {Chaoul}, {Cheek}, {De Angeli}, {Fabricius}, {Guerra}, {Hern{\'a}ndez}, {Jean-Antoine-Piccolo}, {Masana}, {Messineo}, {Mowlavi}, {Nienartowicz}, {Ord{\'o}{\~n}ez-Blanco}, {Panuzzo}, {Portell}, {Richards}, {Riello}, {Seabroke}, {Tanga}, {Th{\'e}venin}, {Torra}, {Els}, {Gracia-Abril}, {Comoretto}, {Garcia-Reinaldos}, {Lock}, {Mercier}, {Altmann}, {Andrae}, {Astraatmadja}, {Bellas-Velidis}, {Benson}, {Berthier}, {Blomme}, {Busso}, {Carry}, {Cellino}, {Clementini},
{Cowell}, {Creevey}, {Cuypers}, {Davidson}, {De Ridder}, {de Torres}, {Delchambre}, {Dell'Oro}, {Ducourant}, {Fr{\'e}mat}, {Garc{\'\i}a-Torres}, {Gosset}, {Halbwachs}, {Hambly}, {Harrison}, {Hauser}, {Hestroffer}, {Hodgkin}, {Huckle}, {Hutton}, {Jasniewicz}, {Jordan}, {Kontizas}, {Korn}, {Lanzafame}, {Manteiga}, {Moitinho}, {Muinonen}, {Osinde}, {Pancino}, {Pauwels}, {Petit}, {Recio-Blanco}, {Robin}, {Sarro}, {Siopis}, {Smith}, {Smith}, {Sozzetti}, {Thuillot}, {van Reeven}, {Viala}, {Abbas}, {Abreu Aramburu}, {Accart}, {Aguado}, {Allan}, {Allasia}, {Altavilla}, {{\'A}lvarez}, {Alves}, {Anderson}, {Andrei}, {Anglada Varela}, {Antiche}, {Antoja}, {Ant{\'o}n}, {Arcay}, {Atzei}, {Ayache}, {Bach}, {Baker}, {Balaguer-N{\'u}{\~n}ez}, {Barache}, {Barata}, {Barbier}, {Barblan}, {Baroni}, {Barrado y Navascu{\'e}s}, {Barros}, {Barstow}, {Becciani}, {Bellazzini}, {Bellei}, {Bello Garc{\'\i}a}, {Belokurov}, {Bendjoya}, {Berihuete}, {Bianchi}, {Bienaym{\'e}}, {Billebaud}, {Blagorodnova}, {Blanco-Cuaresma}, {Boch},
{Bombrun}, {Borrachero}, {Bouquillon}, {Bourda}, {Bouy}, {Bragaglia}, {Breddels}, {Brouillet}, {Br{\"u}semeister}, {Bucciarelli}, {Budnik}, {Burgess}, {Burgon}, {Burlacu}, {Busonero}, {Buzzi}, {Caffau}, {Cambras}, {Campbell}, {Cancelliere}, {Cantat-Gaudin}, {Carlucci}, {Carrasco}, {Castellani}, {Charlot}, {Charnas}, {Charvet}, {Chassat}, {Chiavassa}, {Clotet}, {Cocozza}, {Collins}, {Collins}, {Costigan}, {Crifo}, {Cross}, {Crosta}, {Crowley}, {Dafonte}, {Damerdji}, {Dapergolas}, {David}, {David}, {De Cat}, {de Felice}, {de Laverny}, {De Luise}, {De March}, {de Martino}, {de Souza}, {Debosscher}, {del Pozo}, {Delbo}, {Delgado}, {Delgado}, {di Marco}, {Di Matteo}, {Diakite}, {Distefano}, {Dolding}, {Dos Anjos}, {Drazinos}, {Dur{\'a}n}, {Dzigan}, {Ecale}, {Edvardsson}, {Enke}, {Erdmann}, {Escolar}, {Espina}, {Evans}, {Eynard Bontemps}, {Fabre}, {Fabrizio}, {Faigler}, {Falc{\~a}o}, {Farr{\`a}s Casas}, {Faye}, {Federici}, {Fedorets}, {Fern{\'a}ndez-Hern{\'a}ndez}, {Fernique}, {Fienga}, {Figueras}, {Filippi},
{Findeisen}, {Fonti}, {Fouesneau}, {Fraile}, {Fraser}, {Fuchs}, {Furnell}, {Gai}, {Galleti}, {Galluccio}, {Garabato}, {Garc{\'\i}a-Sedano}, {Gar{\'e}}, {Garofalo}, {Garralda}, {Gavras}, {Gerssen}, {Geyer}, {Gilmore}, {Girona}, {Giuffrida}, {Gomes}, {Gonz{\'a}lez-Marcos}, {Gonz{\'a}lez-N{\'u}{\~n}ez}, {Gonz{\'a}lez-Vidal}, {Granvik}, {Guerrier}, {Guillout}, {Guiraud}, {G{\'u}rpide}, {Guti{\'e}rrez-S{\'a}nchez}, {Guy}, {Haigron}, {Hatzidimitriou}, {Haywood}, {Heiter}, {Helmi}, {Hobbs}, {Hofmann}, {Holl}, {Holland}, {Hunt}, {Hypki}, {Icardi}, {Irwin}, {Jevardat de Fombelle}, {Jofr{\'e}}, {Jonker}, {Jorissen}, {Julbe}, {Karampelas}, {Kochoska}, {Kohley}, {Kolenberg}, {Kontizas}, {Koposov}, {Kordopatis}, {Koubsky}, {Kowalczyk}, {Krone-Martins}, {Kudryashova}, {Kull}, {Bachchan}, {Lacoste-Seris}, {Lanza}, {Lavigne}, {Le Poncin-Lafitte}, {Lebreton}, {Lebzelter}, {Leccia}, {Leclerc}, {Lecoeur-Taibi}, {Lemaitre}, {Lenhardt}, {Leroux}, {Liao}, {Licata}, {Lindstr{\o}m}, {Lister}, {Livanou}, {Lobel}, {L{\"o}ffler},
{L{\'o}pez}, {Lopez-Lozano}, {Lorenz}, {Loureiro}, {MacDonald}, {Magalh{\~a}es Fernandes}, {Managau}, {Mann}, {Mantelet}, {Marchal}, {Marchant}, {Marconi}, {Marie}, {Marinoni}, {Marrese}, {Marschalk{\'o}}, {Marshall}, {Mart{\'\i}n-Fleitas}, {Martino}, {Mary}, {Matijevi{\v{c}}}, {Mazeh}, {McMillan}, {Messina}, {Mestre}, {Michalik}, {Millar}, {Miranda}, {Molina}, {Molinaro}, {Molinaro}, {Moln{\'a}r}, {Moniez}, {Montegriffo}, {Monteiro}, {Mor}, {Mora}, {Morbidelli}, {Morel}, {Morgenthaler}, {Morley}, {Morris}, {Mulone}, {Muraveva}, {Musella}, {Narbonne}, {Nelemans}, {Nicastro}, {Noval}, {Ord{\'e}novic}, {Ordieres-Mer{\'e}}, {Osborne}, {Pagani}, {Pagano}, {Pailler}, {Palacin}, {Palaversa}, {Parsons}, {Paulsen}, {Pecoraro}, {Pedrosa}, {Pentik{\"a}inen}, {Pereira}, {Pichon}, {Piersimoni}, {Pineau}, {Plachy}, {Plum}, {Poujoulet}, {Pr{\v{s}}a}, {Pulone}, {Ragaini}, {Rago}, {Rambaux}, {Ramos-Lerate}, {Ranalli}, {Rauw}, {Read}, {Regibo}, {Renk}, {Reyl{\'e}}, {Ribeiro}, {Rimoldini}, {Ripepi}, {Riva}, {Rixon},
{Roelens}, {Romero-G{\'o}mez}, {Rowell}, {Royer}, {Rudolph}, {Ruiz-Dern}, {Sadowski}, {Sagrist{\`a} Sell{\'e}s}, {Sahlmann}, {Salgado}, {Salguero}, {Sarasso}, {Savietto}, {Schnorhk}, {Schultheis}, {Sciacca}, {Segol}, {Segovia}, {Segransan}, {Serpell}, {Shih}, {Smareglia}, {Smart}, {Smith}, {Solano}, {Solitro}, {Sordo}, {Soria Nieto}, {Souchay}, {Spagna}, {Spoto}, {Stampa}, {Steele}, {Steidelm{\"u}ller}, {Stephenson}, {Stoev}, {Suess}, {S{\"u}veges}, {Surdej}, {Szabados}, {Szegedi-Elek}, {Tapiador}, {Taris}, {Tauran}, {Taylor}, {Teixeira}, {Terrett}, {Tingley}, {Trager}, {Turon}, {Ulla}, {Utrilla}, {Valentini}, {van Elteren}, {Van Hemelryck}, {van Leeuwen}, {Varadi}, {Vecchiato}, {Veljanoski}, {Via}, {Vicente}, {Vogt}, {Voss}, {Votruba}, {Voutsinas}, {Walmsley}, {Weiler}, {Weingrill}, {Werner}, {Wevers}, {Whitehead}, {Wyrzykowski}, {Yoldas}, {{\v{Z}}erjal}, {Zucker}, {Zurbach}, {Zwitter}, {Alecu}, {Allen}, {Allende Prieto}, {Amorim}, {Anglada-Escud{\'e}}, {Arsenijevic}, {Azaz}, {Balm}, {Beck}, {Bernstein},
{Bigot}, {Bijaoui}, {Blasco}, {Bonfigli}, {Bono}, {Boudreault}, {Bressan}, {Brown}, {Brunet}, {Bunclark}, {Buonanno}, {Butkevich}, {Carret}, {Carrion}, {Chemin}, {Ch{\'e}reau}, {Corcione}, {Darmigny}, {de Boer}, {de Teodoro}, {de Zeeuw}, {Delle Luche}, {Domingues}, {Dubath}, {Fodor}, {Fr{\'e}zouls}, {Fries}, {Fustes}, {Fyfe}, {Gallardo}, {Gallegos}, {Gardiol}, {Gebran}, {Gomboc}, {G{\'o}mez}, {Grux}, {Gueguen}, {Heyrovsky}, {Hoar}, {Iannicola}, {Isasi Parache}, {Janotto}, {Joliet}, {Jonckheere}, {Keil}, {Kim}, {Klagyivik}, {Klar}, {Knude}, {Kochukhov}, {Kolka}, {Kos}, {Kutka}, {Lainey}, {LeBouquin}, {Liu}, {Loreggia}, {Makarov}, {Marseille}, {Martayan}, {Martinez-Rubi}, {Massart}, {Meynadier}, {Mignot}, {Munari}, {Nguyen}, {Nordlander}, {Ocvirk}, {O'Flaherty}, {Olias Sanz}, {Ortiz}, {Osorio}, {Oszkiewicz}, {Ouzounis}, {Palmer}, {Park}, {Pasquato}, {Peltzer}, {Peralta}, {P{\'e}turaud}, {Pieniluoma}, {Pigozzi}, {Poels}, {Prat}, {Prod'homme}, {Raison}, {Rebordao}, {Risquez}, {Rocca-Volmerange}, {Rosen},
{Ruiz-Fuertes}, {Russo}, {Sembay}, {Serraller Vizcaino}, {Short}, {Siebert}, {Silva}, {Sinachopoulos}, {Slezak}, {Soffel}, {Sosnowska}, {Strai{\v{z}}ys}, {ter Linden}, {Terrell}, {Theil}, {Tiede}, {Troisi}, {Tsalmantza}, {Tur}, {Vaccari}, {Vachier}, {Valles}, {Van Hamme}, {Veltz}, {Virtanen}, {Wallut}, {Wichmann}, {Wilkinson}, {Ziaeepour}, \& {Zschocke}}]{gaia2016}
{Gaia Collaboration}, {Prusti}, T., {de Bruijne}, J.~H.~J., {et~al.} 2016, \aap, 595, A1, \dodoi{10.1051/0004-6361/201629272}
\bibitem[{{Gaia Collaboration} {et~al.}(2023{\natexlab{a}}){Gaia Collaboration}, {Vallenari}, {Brown}, {Prusti}, {de Bruijne}, {Arenou}, {Babusiaux}, {Biermann}, {Creevey}, {Ducourant}, {Evans}, {Eyer}, {Guerra}, {Hutton}, {Jordi}, {Klioner}, {Lammers}, {Lindegren}, {Luri}, {Mignard}, {Panem}, {Pourbaix}, {Randich}, {Sartoretti}, {Soubiran}, {Tanga}, {Walton}, {Bailer-Jones}, {Bastian}, {Drimmel}, {Jansen}, {Katz}, {Lattanzi}, {van Leeuwen}, {Bakker}, {Cacciari}, {Casta{\~n}eda}, {De Angeli}, {Fabricius}, {Fouesneau}, {Fr{\'e}mat}, {Galluccio}, {Guerrier}, {Heiter}, {Masana}, {Messineo}, {Mowlavi}, {Nicolas}, {Nienartowicz}, {Pailler}, {Panuzzo}, {Riclet}, {Roux}, {Seabroke}, {Sordo}, {Th{\'e}venin}, {Gracia-Abril}, {Portell}, {Teyssier}, {Altmann}, {Andrae}, {Audard}, {Bellas-Velidis}, {Benson}, {Berthier}, {Blomme}, {Burgess}, {Busonero}, {Busso}, {C{\'a}novas}, {Carry}, {Cellino}, {Cheek}, {Clementini}, {Damerdji}, {Davidson}, {de Teodoro}, {Nu{\~n}ez Campos}, {Delchambre}, {Dell'Oro}, {Esquej},
{Fern{\'a}ndez-Hern{\'a}ndez}, {Fraile}, {Garabato}, {Garc{\'\i}a-Lario}, {Gosset}, {Haigron}, {Halbwachs}, {Hambly}, {Harrison}, {Hern{\'a}ndez}, {Hestroffer}, {Hodgkin}, {Holl}, {Jan{\ss}en}, {Jevardat de Fombelle}, {Jordan}, {Krone-Martins}, {Lanzafame}, {L{\"o}ffler}, {Marchal}, {Marrese}, {Moitinho}, {Muinonen}, {Osborne}, {Pancino}, {Pauwels}, {Recio-Blanco}, {Reyl{\'e}}, {Riello}, {Rimoldini}, {Roegiers}, {Rybizki}, {Sarro}, {Siopis}, {Smith}, {Sozzetti}, {Utrilla}, {van Leeuwen}, {Abbas}, {{\'A}brah{\'a}m}, {Abreu Aramburu}, {Aerts}, {Aguado}, {Ajaj}, {Aldea-Montero}, {Altavilla}, {{\'A}lvarez}, {Alves}, {Anders}, {Anderson}, {Anglada Varela}, {Antoja}, {Baines}, {Baker}, {Balaguer-N{\'u}{\~n}ez}, {Balbinot}, {Balog}, {Barache}, {Barbato}, {Barros}, {Barstow}, {Bartolom{\'e}}, {Bassilana}, {Bauchet}, {Becciani}, {Bellazzini}, {Berihuete}, {Bernet}, {Bertone}, {Bianchi}, {Binnenfeld}, {Blanco-Cuaresma}, {Blazere}, {Boch}, {Bombrun}, {Bossini}, {Bouquillon}, {Bragaglia}, {Bramante}, {Breedt},
{Bressan}, {Brouillet}, {Brugaletta}, {Bucciarelli}, {Burlacu}, {Butkevich}, {Buzzi}, {Caffau}, {Cancelliere}, {Cantat-Gaudin}, {Carballo}, {Carlucci}, {Carnerero}, {Carrasco}, {Casamiquela}, {Castellani}, {Castro-Ginard}, {Chaoul}, {Charlot}, {Chemin}, {Chiaramida}, {Chiavassa}, {Chornay}, {Comoretto}, {Contursi}, {Cooper}, {Cornez}, {Cowell}, {Crifo}, {Cropper}, {Crosta}, {Crowley}, {Dafonte}, {Dapergolas}, {David}, {David}, {de Laverny}, {De Luise}, {De March}, {De Ridder}, {de Souza}, {de Torres}, {del Peloso}, {del Pozo}, {Delbo}, {Delgado}, {Delisle}, {Demouchy}, {Dharmawardena}, {Di Matteo}, {Diakite}, {Diener}, {Distefano}, {Dolding}, {Edvardsson}, {Enke}, {Fabre}, {Fabrizio}, {Faigler}, {Fedorets}, {Fernique}, {Fienga}, {Figueras}, {Fournier}, {Fouron}, {Fragkoudi}, {Gai}, {Garcia-Gutierrez}, {Garcia-Reinaldos}, {Garc{\'\i}a-Torres}, {Garofalo}, {Gavel}, {Gavras}, {Gerlach}, {Geyer}, {Giacobbe}, {Gilmore}, {Girona}, {Giuffrida}, {Gomel}, {Gomez}, {Gonz{\'a}lez-N{\'u}{\~n}ez},
{Gonz{\'a}lez-Santamar{\'\i}a}, {Gonz{\'a}lez-Vidal}, {Granvik}, {Guillout}, {Guiraud}, {Guti{\'e}rrez-S{\'a}nchez}, {Guy}, {Hatzidimitriou}, {Hauser}, {Haywood}, {Helmer}, {Helmi}, {Sarmiento}, {Hidalgo}, {Hilger}, {H{\l}adczuk}, {Hobbs}, {Holland}, {Huckle}, {Jardine}, {Jasniewicz}, {Jean-Antoine Piccolo}, {Jim{\'e}nez-Arranz}, {Jorissen}, {Juaristi Campillo}, {Julbe}, {Karbevska}, {Kervella}, {Khanna}, {Kontizas}, {Kordopatis}, {Korn}, {K{\'o}sp{\'a}l}, {Kostrzewa-Rutkowska}, {Kruszy{\'n}ska}, {Kun}, {Laizeau}, {Lambert}, {Lanza}, {Lasne}, {Le Campion}, {Lebreton}, {Lebzelter}, {Leccia}, {Leclerc}, {Lecoeur-Taibi}, {Liao}, {Licata}, {Lindstr{\o}m}, {Lister}, {Livanou}, {Lobel}, {Lorca}, {Loup}, {Madrero Pardo}, {Magdaleno Romeo}, {Managau}, {Mann}, {Manteiga}, {Marchant}, {Marconi}, {Marcos}, {Marcos Santos}, {Mar{\'\i}n Pina}, {Marinoni}, {Marocco}, {Marshall}, {Martin Polo}, {Mart{\'\i}n-Fleitas}, {Marton}, {Mary}, {Masip}, {Massari}, {Mastrobuono-Battisti}, {Mazeh}, {McMillan}, {Messina}, {Michalik},
{Millar}, {Mints}, {Molina}, {Molinaro}, {Moln{\'a}r}, {Monari}, {Mongui{\'o}}, {Montegriffo}, {Montero}, {Mor}, {Mora}, {Morbidelli}, {Morel}, {Morris}, {Muraveva}, {Murphy}, {Musella}, {Nagy}, {Noval}, {Oca{\~n}a}, {Ogden}, {Ordenovic}, {Osinde}, {Pagani}, {Pagano}, {Palaversa}, {Palicio}, {Pallas-Quintela}, {Panahi}, {Payne-Wardenaar}, {Pe{\~n}alosa Esteller}, {Penttil{\"a}}, {Pichon}, {Piersimoni}, {Pineau}, {Plachy}, {Plum}, {Poggio}, {Pr{\v{s}}a}, {Pulone}, {Racero}, {Ragaini}, {Rainer}, {Raiteri}, {Rambaux}, {Ramos}, {Ramos-Lerate}, {Re Fiorentin}, {Regibo}, {Richards}, {Rios Diaz}, {Ripepi}, {Riva}, {Rix}, {Rixon}, {Robichon}, {Robin}, {Robin}, {Roelens}, {Rogues}, {Rohrbasser}, {Romero-G{\'o}mez}, {Rowell}, {Royer}, {Ruz Mieres}, {Rybicki}, {Sadowski}, {S{\'a}ez N{\'u}{\~n}ez}, {Sagrist{\`a} Sell{\'e}s}, {Sahlmann}, {Salguero}, {Samaras}, {Sanchez Gimenez}, {Sanna}, {Santove{\~n}a}, {Sarasso}, {Schultheis}, {Sciacca}, {Segol}, {Segovia}, {S{\'e}gransan}, {Semeux}, {Shahaf}, {Siddiqui}, {Siebert},
{Siltala}, {Silvelo}, {Slezak}, {Slezak}, {Smart}, {Snaith}, {Solano}, {Solitro}, {Souami}, {Souchay}, {Spagna}, {Spina}, {Spoto}, {Steele}, {Steidelm{\"u}ller}, {Stephenson}, {S{\"u}veges}, {Surdej}, {Szabados}, {Szegedi-Elek}, {Taris}, {Taylor}, {Teixeira}, {Tolomei}, {Tonello}, {Torra}, {Torra}, {Torralba Elipe}, {Trabucchi}, {Tsounis}, {Turon}, {Ulla}, {Unger}, {Vaillant}, {van Dillen}, {van Reeven}, {Vanel}, {Vecchiato}, {Viala}, {Vicente}, {Voutsinas}, {Weiler}, {Wevers}, {Wyrzykowski}, {Yoldas}, {Yvard}, {Zhao}, {Zorec}, {Zucker}, \& {Zwitter}}]{gaia2023}
{Gaia Collaboration}, {Vallenari}, A., {Brown}, A.~G.~A., {et~al.} 2023{\natexlab{a}}, \aap, 674, A1, \dodoi{10.1051/0004-6361/202243940}
\bibitem[{{Gaia Collaboration} {et~al.}(2023{\natexlab{b}}){Gaia Collaboration}, {Arenou}, {Babusiaux}, {Barstow}, {Faigler}, {Jorissen}, {Kervella}, {Mazeh}, {Mowlavi}, {Panuzzo}, {Sahlmann}, {Shahaf}, {Sozzetti}, {Bauchet}, {Damerdji}, {Gavras}, {Giacobbe}, {Gosset}, {Halbwachs}, {Holl}, {Lattanzi}, {Leclerc}, {Morel}, {Pourbaix}, {Re Fiorentin}, {Sadowski}, {S{\'e}gransan}, {Siopis}, {Teyssier}, {Zwitter}, {Planquart}, {Brown}, {Vallenari}, {Prusti}, {de Bruijne}, {Biermann}, {Creevey}, {Ducourant}, {Evans}, {Eyer}, {Guerra}, {Hutton}, {Jordi}, {Klioner}, {Lammers}, {Lindegren}, {Luri}, {Mignard}, {Panem}, {Randich}, {Sartoretti}, {Soubiran}, {Tanga}, {Walton}, {Bailer-Jones}, {Bastian}, {Drimmel}, {Jansen}, {Katz}, {van Leeuwen}, {Bakker}, {Cacciari}, {Casta{\~n}eda}, {De Angeli}, {Fabricius}, {Fouesneau}, {Fr{\'e}mat}, {Galluccio}, {Guerrier}, {Heiter}, {Masana}, {Messineo}, {Nicolas}, {Nienartowicz}, {Pailler}, {Riclet}, {Roux}, {Seabroke}, {Sordo}, {Th{\'e}venin}, {Gracia-Abril}, {Portell}, {Altmann},
{Andrae}, {Audard}, {Bellas-Velidis}, {Benson}, {Berthier}, {Blomme}, {Burgess}, {Busonero}, {Busso}, {C{\'a}novas}, {Carry}, {Cellino}, {Cheek}, {Clementini}, {Davidson}, {de Teodoro}, {Nu{\~n}ez Campos}, {Delchambre}, {Dell'Oro}, {Esquej}, {Fern{\'a}ndez-Hern{\'a}ndez}, {Fraile}, {Garabato}, {Garc{\'\i}a-Lario}, {Haigron}, {Hambly}, {Harrison}, {Hern{\'a}ndez}, {Hestroffer}, {Hodgkin}, {Jan{\ss}en}, {Jevardat de Fombelle}, {Jordan}, {Krone-Martins}, {Lanzafame}, {L{\"o}ffler}, {Marchal}, {Marrese}, {Moitinho}, {Muinonen}, {Osborne}, {Pancino}, {Pauwels}, {Recio-Blanco}, {Reyl{\'e}}, {Riello}, {Rimoldini}, {Roegiers}, {Rybizki}, {Sarro}, {Smith}, {Utrilla}, {van Leeuwen}, {Abbas}, {{\'A}brah{\'a}m}, {Abreu Aramburu}, {Aerts}, {Aguado}, {Ajaj}, {Aldea-Montero}, {Altavilla}, {{\'A}lvarez}, {Alves}, {Anders}, {Anderson}, {Anglada Varela}, {Antoja}, {Baines}, {Baker}, {Balaguer-N{\'u}{\~n}ez}, {Balbinot}, {Balog}, {Barache}, {Barbato}, {Barros}, {Bartolom{\'e}}, {Bassilana}, {Becciani}, {Bellazzini},
{Berihuete}, {Bernet}, {Bertone}, {Bianchi}, {Binnenfeld}, {Blanco-Cuaresma}, {Blazere}, {Boch}, {Bombrun}, {Bossini}, {Bouquillon}, {Bragaglia}, {Bramante}, {Breedt}, {Bressan}, {Brouillet}, {Brugaletta}, {Bucciarelli}, {Burlacu}, {Butkevich}, {Buzzi}, {Caffau}, {Cancelliere}, {Cantat-Gaudin}, {Carballo}, {Carlucci}, {Carnerero}, {Carrasco}, {Casamiquela}, {Castellani}, {Castro-Ginard}, {Chaoul}, {Charlot}, {Chemin}, {Chiaramida}, {Chiavassa}, {Chornay}, {Comoretto}, {Contursi}, {Cooper}, {Cornez}, {Cowell}, {Crifo}, {Cropper}, {Crosta}, {Crowley}, {Dafonte}, {Dapergolas}, {David}, {de Laverny}, {De Luise}, {De March}, {De Ridder}, {de Souza}, {de Torres}, {del Peloso}, {del Pozo}, {Delbo}, {Delgado}, {Delisle}, {Demouchy}, {Dharmawardena}, {Diakite}, {Diener}, {Distefano}, {Dolding}, {Enke}, {Fabre}, {Fabrizio}, {Fedorets}, {Fernique}, {Figueras}, {Fournier}, {Fouron}, {Fragkoudi}, {Gai}, {Garcia-Gutierrez}, {Garcia-Reinaldos}, {Garc{\'\i}a-Torres}, {Garofalo}, {Gavel}, {Gerlach}, {Geyer}, {Gilmore},
{Girona}, {Giuffrida}, {Gomel}, {Gomez}, {Gonz{\'a}lez-N{\'u}{\~n}ez}, {Gonz{\'a}lez-Santamar{\'\i}a}, {Gonz{\'a}lez-Vidal}, {Granvik}, {Guillout}, {Guiraud}, {Guti{\'e}rrez-S{\'a}nchez}, {Guy}, {Hatzidimitriou}, {Hauser}, {Haywood}, {Helmer}, {Helmi}, {Sarmiento}, {Hidalgo}, {Hilger}, {H{\l}adczuk}, {Hobbs}, {Holland}, {Huckle}, {Jardine}, {Jasniewicz}, {Jean-Antoine Piccolo}, {Jim{\'e}nez-Arranz}, {Juaristi Campillo}, {Julbe}, {Karbevska}, {Khanna}, {Kordopatis}, {Korn}, {K{\'o}sp{\'a}l}, {Kostrzewa-Rutkowska}, {Kruszy{\'n}ska}, {Kun}, {Laizeau}, {Lambert}, {Lanza}, {Lasne}, {Le Campion}, {Lebreton}, {Lebzelter}, {Leccia}, {Lecoeur-Taibi}, {Liao}, {Licata}, {Lindstr{\o}m}, {Lister}, {Livanou}, {Lobel}, {Lorca}, {Loup}, {Madrero Pardo}, {Magdaleno Romeo}, {Managau}, {Mann}, {Manteiga}, {Marchant}, {Marconi}, {Marcos}, {Marcos Santos}, {Mar{\'\i}n Pina}, {Marinoni}, {Marocco}, {Marshall}, {Martin Polo}, {Mart{\'\i}n-Fleitas}, {Marton}, {Mary}, {Masip}, {Massari}, {Mastrobuono-Battisti}, {McMillan},
{Messina}, {Michalik}, {Millar}, {Mints}, {Molina}, {Molinaro}, {Moln{\'a}r}, {Monari}, {Mongui{\'o}}, {Montegriffo}, {Montero}, {Mor}, {Mora}, {Morbidelli}, {Morris}, {Muraveva}, {Murphy}, {Musella}, {Nagy}, {Noval}, {Oca{\~n}a}, {Ogden}, {Ordenovic}, {Osinde}, {Pagani}, {Pagano}, {Palaversa}, {Palicio}, {Pallas-Quintela}, {Panahi}, {Payne-Wardenaar}, {Pe{\~n}alosa Esteller}, {Penttil{\"a}}, {Pichon}, {Piersimoni}, {Pineau}, {Plachy}, {Plum}, {Poggio}, {Pr{\v{s}}a}, {Pulone}, {Racero}, {Ragaini}, {Rainer}, {Raiteri}, {Ramos}, {Ramos-Lerate}, {Regibo}, {Richards}, {Rios Diaz}, {Ripepi}, {Riva}, {Rix}, {Rixon}, {Robichon}, {Robin}, {Robin}, {Roelens}, {Rogues}, {Rohrbasser}, {Romero-G{\'o}mez}, {Rowell}, {Royer}, {Ruz Mieres}, {Rybicki}, {S{\'a}ez N{\'u}{\~n}ez}, {Sagrist{\`a} Sell{\'e}s}, {Salguero}, {Samaras}, {Sanchez Gimenez}, {Sanna}, {Santove{\~n}a}, {Sarasso}, {Schultheis}, {Sciacca}, {Segol}, {Segovia}, {Semeux}, {Siddiqui}, {Siebert}, {Siltala}, {Silvelo}, {Slezak}, {Slezak}, {Smart}, {Snaith},
{Solano}, {Solitro}, {Souami}, {Souchay}, {Spagna}, {Spina}, {Spoto}, {Steele}, {Steidelm{\"u}ller}, {Stephenson}, {S{\"u}veges}, {Surdej}, {Szabados}, {Szegedi-Elek}, {Taris}, {Taylor}, {Teixeira}, {Tolomei}, {Tonello}, {Torra}, {Torra}, {Torralba Elipe}, {Trabucchi}, {Tsounis}, {Turon}, {Ulla}, {Unger}, {Vaillant}, {van Dillen}, {van Reeven}, {Vanel}, {Vecchiato}, {Viala}, {Vicente}, {Voutsinas}, {Weiler}, {Wevers}, {Wyrzykowski}, {Yoldas}, {Yvard}, {Zhao}, {Zorec}, \& {Zucker}}]{Arenou2023}
{Gaia Collaboration}, {Arenou}, F., {Babusiaux}, C., {et~al.} 2023{\natexlab{b}}, \aap, 674, A34, \dodoi{10.1051/0004-6361/202243782}
\bibitem[{{Geller} {et~al.}(2021){Geller}, {Mathieu}, {Latham}, {Pollack}, {Torres}, \& {Leiner}}]{Geller2021}
{Geller}, A.~M., {Mathieu}, R.~D., {Latham}, D.~W., {et~al.} 2021, \aj, 161, 190, \dodoi{10.3847/1538-3881/abdd23}
\bibitem[{{Goodman} \& {Tan}(2004)}]{Goodman2004}
{Goodman}, J., \& {Tan}, J.~C. 2004, \apj, 608, 108, \dodoi{10.1086/386360}
\bibitem[{{Greenzweig} \& {Lissauer}(1992)}]{Greenzweig1992}
{Greenzweig}, Y., \& {Lissauer}, J.~J. 1992, \icarus, 100, 440, \dodoi{10.1016/0019-1035(92)90110-S}
\bibitem[{{Griffin}(2012)}]{Griffin2012}
{Griffin}, R.~F. 2012, Journal of Astrophysics and Astronomy, 33, 29, \dodoi{10.1007/s12036-012-9137-5}
\bibitem[{{Halbwachs} {et~al.}(2023){Halbwachs}, {Pourbaix}, {Arenou}, {Galluccio}, {Guillout}, {Bauchet}, {Marchal}, {Sadowski}, \& {Teyssier}}]{Halbwachs2023}
{Halbwachs}, J.-L., {Pourbaix}, D., {Arenou}, F., {et~al.} 2023, \aap, 674, A9, \dodoi{10.1051/0004-6361/202243969}
\bibitem[{{Heggie}(1975)}]{Heggie1975}
{Heggie}, D.~C. 1975, \mnras, 173, 729, \dodoi{10.1093/mnras/173.3.729}
\bibitem[{{Heggie} \& {Rasio}(1996)}]{Heggie1996}
{Heggie}, D.~C., \& {Rasio}, F.~A. 1996, \mnras, 282, 1064, \dodoi{10.1093/mnras/282.3.1064}
\bibitem[{{Hwang} {et~al.}(2021){Hwang}, {Ting}, {Schlaufman}, {Zakamska}, \& {Wyse}}]{Hwang2021}
{Hwang}, H.-C., {Ting}, Y.-S., {Schlaufman}, K.~C., {Zakamska}, N.~L., \& {Wyse}, R. F.~G. 2021, \mnras, 501, 4329, \dodoi{10.1093/mnras/staa3854}
\bibitem[{{Hwang} {et~al.}(2022){Hwang}, {Ting}, \& {Zakamska}}]{Hwang2022}
{Hwang}, H.-C., {Ting}, Y.-S., \& {Zakamska}, N.~L. 2022, \mnras, 512, 3383, \dodoi{10.1093/mnras/stac675}
\bibitem[{{Ida} \& {Makino}(1992{\natexlab{a}})}]{Ida1992}
{Ida}, S., \& {Makino}, J. 1992{\natexlab{a}}, \icarus, 96, 107, \dodoi{10.1016/0019-1035(92)90008-U}
\bibitem[{{Ida} \& {Makino}(1992{\natexlab{b}})}]{Ida1992b}
---. 1992{\natexlab{b}}, \icarus, 98, 28, \dodoi{10.1016/0019-1035(92)90203-J}
\bibitem[{{Jeans}(1919)}]{Jeans1919}
{Jeans}, J.~H. 1919, \mnras, 79, 408, \dodoi{10.1093/mnras/79.6.408}
\bibitem[{{Juri{\'c}} \& {Tremaine}(2008)}]{Juric2008}
{Juri{\'c}}, M., \& {Tremaine}, S. 2008, \apj, 686, 603, \dodoi{10.1086/590047}
\bibitem[{{Kirkpatrick} {et~al.}(2024){Kirkpatrick}, {Marocco}, {Gelino}, {Raghu}, {Faherty}, {Bardalez Gagliuffi}, {Schurr}, {Apps}, {Schneider}, {Meisner}, {Kuchner}, {Caselden}, {Smart}, {Casewell}, {Raddi}, {Kesseli}, {Stevnbak Andersen}, {Antonini}, {Beaulieu}, {Bickle}, {Bilsing}, {Chieng}, {Colin}, {Deen}, {Dereveanco}, {Doll}, {Durantini Luca}, {Frazer}, {Gantier}, {Gramaize}, {Grant}, {Hamlet}, {Higashimura}, {Hyogo}, {Ja{\l}owiczor}, {Jonkeren}, {Kabatnik}, {Kiwy}, {Martin}, {Michaels}, {Pendrill}, {Pessanha Machado}, {Pumphrey}, {Rothermich}, {Russwurm}, {Sainio}, {Sanchez}, {Sapelkin-Tambling}, {Sch{\"u}mann}, {Selg-Mann}, {Singh}, {Stenner}, {Sun}, {Tanner}, {Th{\'e}venot}, {Ventura}, {Voloshin}, {Walla}, {W{\k{e}}dracki}, {Adorno}, {Aganze}, {Allers}, {Brooks}, {Burgasser}, {Calamari}, {Connor}, {Costa}, {Eisenhardt}, {Gagn{\'e}}, {Gerasimov}, {Gonzales}, {Hsu}, {Kiman}, {Li}, {Low}, {Mamajek}, {Pantoja}, {Popinchalk}, {Rees}, {Stern}, {Su{\'a}rez}, {Theissen}, {Tsai}, {Vos}, {Zurek}, \& {The
Backyard Worlds: Planet 9 Collaboration}}]{Kirkpatrick2024}
{Kirkpatrick}, J.~D., {Marocco}, F., {Gelino}, C.~R., {et~al.} 2024, \apjs, 271, 55, \dodoi{10.3847/1538-4365/ad24e2}
\bibitem[{{Kratter} \& {Lodato}(2016)}]{kratterlodato16}
{Kratter}, K., \& {Lodato}, G. 2016, \araa, 54, 271, \dodoi{10.1146/annurev-astro-081915-023307}
\bibitem[{{Kratter} {et~al.}(2010){Kratter}, {Matzner}, {Krumholz}, \& {Klein}}]{kratter10}
{Kratter}, K.~M., {Matzner}, C.~D., {Krumholz}, M.~R., \& {Klein}, R.~I. 2010, \apj, 708, 1585, \dodoi{10.1088/0004-637X/708/2/1585}
\bibitem[{{Kroupa} \& {Bouvier}(2003)}]{Kroupa2003}
{Kroupa}, P., \& {Bouvier}, J. 2003, \mnras, 346, 369, \dodoi{10.1046/j.1365-2966.2003.07224.x}
\bibitem[{{Kumar} \& {Schmeja}(2007)}]{Kumar2007}
{Kumar}, M.~S.~N., \& {Schmeja}, S. 2007, \aap, 471, L33, \dodoi{10.1051/0004-6361:20077875}
\bibitem[{{Lai} \& {Mu{\~n}oz}(2023)}]{Lai2023}
{Lai}, D., \& {Mu{\~n}oz}, D.~J. 2023, \araa, 61, 517, \dodoi{10.1146/annurev-astro-052622-022933}
\bibitem[{{Levin}(2007)}]{Levin2007}
{Levin}, Y. 2007, \mnras, 374, 515, \dodoi{10.1111/j.1365-2966.2006.11155.x}
\bibitem[{{Lodieu}(2013)}]{Lodieu2013}
{Lodieu}, N. 2013, \mnras, 431, 3222, \dodoi{10.1093/mnras/stt402}
\bibitem[{{Malhotra} \& {Wang}(2017)}]{Malhotra2017}
{Malhotra}, R., \& {Wang}, X. 2017, \mnras, 465, 4381, \dodoi{10.1093/mnras/stw3009}
\bibitem[{{Matzner} \& {Levin}(2005)}]{matznerlevin05}
{Matzner}, C.~D., \& {Levin}, Y. 2005, \apj, 628, 817, \dodoi{10.1086/430813}
\bibitem[{{Mazzola} {et~al.}(2020){Mazzola}, {Badenes}, {Moe}, {Koposov}, {Kounkel}, {Kratter}, {Covey}, {Walker}, {Thompson}, {Andrews}, {Freeman}, {Anguiano}, {Carlberg}, {De Lee}, {Frinchaboy}, {Lewis}, {Majewski}, {Nidever}, {Nitschelm}, {Price-Whelan}, {Roman-Lopes}, {Stassun}, \& {Troup}}]{Mazzola2020}
{Mazzola}, C.~N., {Badenes}, C., {Moe}, M., {et~al.} 2020, \mnras, 499, 1607, \dodoi{10.1093/mnras/staa2859}
\bibitem[{{McLaughlin} {et~al.}(2006){McLaughlin}, {Anderson}, {Meylan}, {Gebhardt}, {Pryor}, {Minniti}, \& {Phinney}}]{McLaughlin2006}
{McLaughlin}, D.~E., {Anderson}, J., {Meylan}, G., {et~al.} 2006, \apjs, 166, 249, \dodoi{10.1086/505692}
\bibitem[{{Moe} \& {Di Stefano}(2017)}]{Moe2017}
{Moe}, M., \& {Di Stefano}, R. 2017, \apjs, 230, 15, \dodoi{10.3847/1538-4365/aa6fb6}
\bibitem[{{Moe} {et~al.}(2019){Moe}, {Kratter}, \& {Badenes}}]{Moe2019}
{Moe}, M., {Kratter}, K.~M., \& {Badenes}, C. 2019, \apj, 875, 61, \dodoi{10.3847/1538-4357/ab0d88}
\bibitem[{{Moharana} {et~al.}(2024){Moharana}, {Helminiak}, {Pawar}, \& {Pawar}}]{Moharana2024}
{Moharana}, A., {Helminiak}, K.~G., {Pawar}, T., \& {Pawar}, G. 2024, arXiv e-prints, arXiv:2411.11459, \dodoi{10.48550/arXiv.2411.11459}
\bibitem[{{Moraux} {et~al.}(2003){Moraux}, {Bouvier}, {Stauffer}, \& {Cuillandre}}]{Moraux2003}
{Moraux}, E., {Bouvier}, J., {Stauffer}, J.~R., \& {Cuillandre}, J.~C. 2003, \aap, 400, 891, \dodoi{10.1051/0004-6361:20021903}
\bibitem[{{Mu{\~n}oz} {et~al.}(2019){Mu{\~n}oz}, {Miranda}, \& {Lai}}]{Munoz2019}
{Mu{\~n}oz}, D.~J., {Miranda}, R., \& {Lai}, D. 2019, \apj, 871, 84, \dodoi{10.3847/1538-4357/aaf867}
\bibitem[{{Offner} {et~al.}(2023){Offner}, {Moe}, {Kratter}, {Sadavoy}, {Jensen}, \& {Tobin}}]{offner2023}
{Offner}, S.~S.~R., {Moe}, M., {Kratter}, K.~M., {et~al.} 2023, in Astronomical Society of the Pacific Conference Series, Vol. 534, Protostars and Planets VII, ed. S.~{Inutsuka}, Y.~{Aikawa}, T.~{Muto}, K.~{Tomida}, \& M.~{Tamura}, 275, \dodoi{10.48550/arXiv.2203.10066}
\bibitem[{{Panwar} {et~al.}(2024){Panwar}, {Rishi}, {Sharma}, {Ojha}, {Samal}, {Singh}, \& {Yadav}}]{Panwar2024}
{Panwar}, N., {Rishi}, C., {Sharma}, S., {et~al.} 2024, \aj, 168, 89, \dodoi{10.3847/1538-3881/ad5552}
\bibitem[{{Papaloizou} {et~al.}(2001){Papaloizou}, {Nelson}, \& {Masset}}]{Papaloizou2001}
{Papaloizou}, J.~C.~B., {Nelson}, R.~P., \& {Masset}, F. 2001, \aap, 366, 263, \dodoi{10.1051/0004-6361:20000011}
\bibitem[{{Parker} {et~al.}(2009){Parker}, {Goodwin}, {Kroupa}, \& {Kouwenhoven}}]{Parker2009}
{Parker}, R.~J., {Goodwin}, S.~P., {Kroupa}, P., \& {Kouwenhoven}, M.~B.~N. 2009, \mnras, 397, 1577, \dodoi{10.1111/j.1365-2966.2009.15032.x}
\bibitem[{{Pourbaix} {et~al.}(2004){Pourbaix}, {Tokovinin}, {Batten}, {Fekel}, {Hartkopf}, {Levato}, {Morrell}, {Torres}, \& {Udry}}]{SB9}
{Pourbaix}, D., {Tokovinin}, A.~A., {Batten}, A.~H., {et~al.} 2004, \aap, 424, 727, \dodoi{10.1051/0004-6361:20041213}
\bibitem[{Price-Whelan {et~al.}(2020)Price-Whelan, Hogg, Rix, Beaton, Lewis, Nidever, Almeida, Badenes, Barba, Beers, Carlberg, De~Lee, Fernández-Trincado, Frinchaboy, García-Hernández, Green, Hasselquist, Longa-Peña, Majewski, Nitschelm, Sobeck, Stassun, Stringfellow, \& Troup}]{PriceWhelan2020}
Price-Whelan, A.~M., Hogg, D.~W., Rix, H.-W., {et~al.} 2020, The Astrophysical Journal, 895, 2, \dodoi{10.3847/1538-4357/ab8acc}
\bibitem[{{Rafikov}(2005)}]{rafikov05}
{Rafikov}, R.~R. 2005, \apjl, 621, L69, \dodoi{10.1086/428899}
\bibitem[{{Raghavan} {et~al.}(2010){Raghavan}, {McAlister}, {Henry}, {Latham}, {Marcy}, {Mason}, {Gies}, {White}, \& {ten Brummelaar}}]{Raghavan2010}
{Raghavan}, D., {McAlister}, H.~A., {Henry}, T.~J., {et~al.} 2010, \apjs, 190, 1, \dodoi{10.1088/0067-0049/190/1/1}
\bibitem[{{Rein} \& {Liu}(2012)}]{Rein2012}
{Rein}, H., \& {Liu}, S.~F. 2012, \aap, 537, A128, \dodoi{10.1051/0004-6361/201118085}
\bibitem[{{Rein} \& {Spiegel}(2015)}]{Rein2015}
{Rein}, H., \& {Spiegel}, D.~S. 2015, \mnras, 446, 1424, \dodoi{10.1093/mnras/stu2164}
\bibitem[{{Reipurth} \& {Clarke}(2001)}]{Reipurth2001}
{Reipurth}, B., \& {Clarke}, C. 2001, \aj, 122, 432, \dodoi{10.1086/321121}
\bibitem[{{Scholz} {et~al.}(2012){Scholz}, {Muzic}, {Geers}, {Bonavita}, {Jayawardhana}, \& {Tamura}}]{Scholz2012}
{Scholz}, A., {Muzic}, K., {Geers}, V., {et~al.} 2012, \apj, 744, 6, \dodoi{10.1088/0004-637X/744/1/6}
\bibitem[{{Shahaf} {et~al.}(2023){Shahaf}, {Bashi}, {Mazeh}, {Faigler}, {Arenou}, {El-Badry}, \& {Rix}}]{Shahaf2023}
{Shahaf}, S., {Bashi}, D., {Mazeh}, T., {et~al.} 2023, \mnras, 518, 2991, \dodoi{10.1093/mnras/stac3290}
\bibitem[{{Siwek} {et~al.}(2023){Siwek}, {Weinberger}, \& {Hernquist}}]{Siwek2023}
{Siwek}, M., {Weinberger}, R., \& {Hernquist}, L. 2023, \mnras, 522, 2707, \dodoi{10.1093/mnras/stad1131}
\bibitem[{{Spurzem} {et~al.}(2009){Spurzem}, {Giersz}, {Heggie}, \& {Lin}}]{Spurzem2009}
{Spurzem}, R., {Giersz}, M., {Heggie}, D.~C., \& {Lin}, D.~N.~C. 2009, \apj, 697, 458, \dodoi{10.1088/0004-637X/697/1/458}
\bibitem[{{Sumi} {et~al.}(2023){Sumi}, {Koshimoto}, {Bennett}, {Rattenbury}, {Abe}, {Barry}, {Bhattacharya}, {Bond}, {Fujii}, {Fukui}, {Hamada}, {Hirao}, {Silva}, {Itow}, {Kirikawa}, {Kondo}, {Matsubara}, {Miyazaki}, {Muraki}, {Olmschenk}, {Ranc}, {Satoh}, {Suzuki}, {Tomoyoshi}, {Tristram}, {Vandorou}, {Yama}, \& {Yamashita}}]{Sumi2023}
{Sumi}, T., {Koshimoto}, N., {Bennett}, D.~P., {et~al.} 2023, \aj, 166, 108, \dodoi{10.3847/1538-3881/ace688}
\bibitem[{{Tobin} {et~al.}(2016){Tobin}, {Kratter}, {Persson}, {Looney}, {Dunham}, {Segura-Cox}, {Li}, {Chandler}, {Sadavoy}, {Harris}, {Melis}, \& {P{\'e}rez}}]{tobin16triple}
{Tobin}, J.~J., {Kratter}, K.~M., {Persson}, M.~V., {et~al.} 2016, \nat, 538, 483, \dodoi{10.1038/nature20094}
\bibitem[{{Tobin} {et~al.}(2020){Tobin}, {Sheehan}, {Megeath}, {D{\'\i}az-Rodr{\'\i}guez}, {Offner}, {Murillo}, {van 't Hoff}, {van Dishoeck}, {Osorio}, {Anglada}, {Furlan}, {Stutz}, {Reynolds}, {Karnath}, {Fischer}, {Persson}, {Looney}, {Li}, {Stephens}, {Chandler}, {Cox}, {Dunham}, {Tychoniec}, {Kama}, {Kratter}, {Kounkel}, {Mazur}, {Maud}, {Patel}, {Perez}, {Sadavoy}, {Segura-Cox}, {Sharma}, {Stephenson}, {Watson}, \& {Wyrowski}}]{Tobin2020}
{Tobin}, J.~J., {Sheehan}, P.~D., {Megeath}, S.~T., {et~al.} 2020, \apj, 890, 130, \dodoi{10.3847/1538-4357/ab6f64}
\bibitem[{{Tokovinin}(2017)}]{Tokovinin2017}
{Tokovinin}, A. 2017, \apj, 844, 103, \dodoi{10.3847/1538-4357/aa7746}
\bibitem[{{Tokovinin}(2020)}]{Tokovinin2020}
---. 2020, \mnras, 496, 987, \dodoi{10.1093/mnras/staa1639}
\bibitem[{{Tokovinin}(2021)}]{Tokovinin2021}
---. 2021, Universe, 7, 352, \dodoi{10.3390/universe7090352}
\bibitem[{{Tokovinin} \& {Kiyaeva}(2016)}]{Tokovinin2016}
{Tokovinin}, A., \& {Kiyaeva}, O. 2016, \mnras, 456, 2070, \dodoi{10.1093/mnras/stv2825}
\bibitem[{{Tokovinin}(2000)}]{Tokovinin2000}
{Tokovinin}, A.~A. 2000, \aap, 360, 997
\bibitem[{{Tremaine}(2015)}]{Tremaine2015}
{Tremaine}, S. 2015, \apj, 807, 157, \dodoi{10.1088/0004-637X/807/2/157}
\bibitem[{{Udry} {et~al.}(1998){Udry}, {Mayor}, {Latham}, {Stefanik}, {Torres}, {Mazeh}, {Goldberg}, {Andersen}, \& {Nordstrom}}]{Udry1998}
{Udry}, S., {Mayor}, M., {Latham}, D.~W., {et~al.} 1998, in Astronomical Society of the Pacific Conference Series, Vol. 154, Cool Stars, Stellar Systems, and the Sun, ed. R.~A. {Donahue} \& J.~A. {Bookbinder}, 2148
\bibitem[{{Valli} {et~al.}(2024){Valli}, {Tiede}, {Vigna-G{\'o}mez}, {Cuadra}, {Siwek}, {Ma}, {D'Orazio}, {Zrake}, \& {de Mink}}]{Valli2024}
{Valli}, R., {Tiede}, C., {Vigna-G{\'o}mez}, A., {et~al.} 2024, \aap, 688, A128, \dodoi{10.1051/0004-6361/202449421}
\bibitem[{{Wang} {et~al.}(2023){Wang}, {Bai}, \& {Lai}}]{Wang2023}
{Wang}, H.-Y., {Bai}, X.-N., \& {Lai}, D. 2023, \apj, 943, 175, \dodoi{10.3847/1538-4357/acac77}
\bibitem[{{Xu} {et~al.}(2024){Xu}, {Jiang}, {Kunz}, \& {Stone}}]{Xu2024}
{Xu}, W., {Jiang}, Y.-F., {Kunz}, M.~W., \& {Stone}, J.~M. 2024, arXiv e-prints, arXiv:2410.12042, \dodoi{10.48550/arXiv.2410.12042}
\bibitem[{{Zhou} {et~al.}(2007){Zhou}, {Lin}, \& {Sun}}]{Zhou2007}
{Zhou}, J.-L., {Lin}, D. N.~C., \& {Sun}, Y.-S. 2007, \apj, 666, 423, \dodoi{10.1086/519918}
\bibitem[{{Zrake} {et~al.}(2021){Zrake}, {Tiede}, {MacFadyen}, \& {Haiman}}]{Zrake2021}
{Zrake}, J., {Tiede}, C., {MacFadyen}, A., \& {Haiman}, Z. 2021, \apjl, 909, L13, \dodoi{10.3847/2041-8213/abdd1c}
\end{thebibliography}