Searching for the lost Unicorn: a prominent feature in the radial velocity distribution of stars in Vela from Gaia DR2 data
Extracted source text ยท arXiv 1808.09921
%
% Letter to Monthly Notices of the Royal Astronomical Society
% about a prominent feature in the velocity distribution of
% stars in Vela from Gaia DR2 data
%
% by R. de la Fuente Marcos and C. de la Fuente Marcos
%
%@arxiver{fdistVrxn.eps,fpmM4C2n.eps,m4cFULLn.eps}
\documentclass[fleqn,usenatbib,useAMS]{mnras}
\usepackage{graphicx}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage[T1]{fontenc}
\usepackage{ae,aecompl}
\usepackage{txfonts}
\usepackage{enumerate}
\title[Searching for the lost Unicorn in Vela]
{Searching for the lost Unicorn: a prominent feature in the radial
velocity distribution of stars in Vela from \textit{\textbf{Gaia}} DR2 data}
\author[R. de la Fuente Marcos and C. de la Fuente Marcos]
{R.~de~la~Fuente Marcos$^{1}$\thanks{E-mail: rauldelafuentemarcos@ucm.es}
and
C. de la Fuente Marcos$^2$ \\
$^1$AEGORA Research Group,
Facultad de Ciencias Matem\'aticas,
Universidad Complutense de Madrid,
Ciudad Universitaria, E-28040 Madrid, Spain \\
$^2$ Universidad Complutense de Madrid,
Ciudad Universitaria, E-28040 Madrid, Spain}
\date{Accepted 2018 August 29.
Received 2018 August 29;
in original form 2018 July 8}
\pubyear{2018}
\begin{document}
\label{firstpage}
\pagerange{\pageref{firstpage}--\pageref{lastpage}}
\maketitle
\begin{abstract}
Stellar streams are ubiquitous in the Galactic halo and they can be used
to improve our understanding of the formation and evolution of the Milky
Way as a whole. The so-called Monoceros Ring might have been the result
of satellite accretion. Guglielmo et al. have used $N$-body simulations
to search for the progenitor of this structure. Their analysis shows
that, if the Ring has a dwarf galaxy progenitor, it might be found in
the background of one out of eight specific areas in the sky. Here, we
use {\it Gaia} DR2 data to perform a systematic exploration aimed at
confirming or rejecting this remarkable prediction. Focusing on the
values of the radial velocity to uncover possible multimodal spreads,
we identify a bimodal Gaussian distribution towards Galactic coordinates
($l$, $b$) = (271{\degr}, +2{\degr}) in Vela, which is one of the
locations of the progenitor proposed by Guglielmo et al. This prominent
feature with central values 60$\pm$7~km~s$^{-1}$ and
97$\pm$10~km~s$^{-1}$, may signal the presence of the long sought
progenitor of the Monoceros Ring, but the data might also be compatible
with the existence of an unrelated, previously unknown, kinematically
coherent structure.
\end{abstract}
\begin{keywords}
methods: statistical -- celestial mechanics --
Galaxy: disc -- Galaxy: evolution --
Galaxy: stellar content -- Galaxy: structure.
\end{keywords}
\section{Introduction}
Beyond the nominal edge of the Milky Way disc, 15~kpc from the Galactic centre, lies a complex network of coherent stellar structures
whose origins are not yet fully understood (see e.g. \citealt{2011PhDT.......171S,2011ApJ...738...79X,Pila15,2016ASSL..420..113S}).
Some may be the result of the Milky Way cannibalizing nearby dwarf galaxies \citep{1994Natur.370..194I}, others could be just stellar
lumps induced by the combined action of the gravitational potential of the Galaxy and those of passing and/or falling neighbours
\citep{2015ApJ...801..105X,2018MNRAS.478.3809S}. The study of these structures can help in understanding how the Milky Way came into
existence and how it has evolved progressively to become what we observe now. Among all these structures, the true nature of the
so-called Monoceros Ring remains elusive. Originally identified by \citet{2002ApJ...569..245N}, the structure is considered by some as
a {\it bona fide} stellar stream (e.g. \citealt{2003MNRAS.340L..21I,2003ApJ...588..824Y,2007MNRAS.376..939C,2008ApJ...689L.117G,
2008ApJ...684..287I,2011MNRAS.414L...1M,2011ApJ...730L...6S,2018MNRAS.473.1218L}), while others put its origin in the (flared thick)
disc of the Milky Way (e.g. \citealt{2005ApJ...630L.153C,2006A&A...451..515M,2014ApJ...794...90K,2014A&A...567A.106L,2018ApJ...854...47S,
2018MNRAS.478.3367W}). Alternative scenarios put its provenance in an outer spiral arm such as the one described by
\citet{2011ApJ...734L..24D} or in undulations of the disc \citep{2017ApJ...844...74L} like those discussed by \citet{2015ApJ...801..105X}.
\citet{2018MNRAS.474.4584G} have recently explored the kinematics, proper motions, and the nature of the putative progenitor of the
Monoceros Ring. Although they could not confirm that the Ring has its origin in an accretion episode, their analysis argues that, if
the Ring has a dwarf galaxy progenitor, it might be found in the background of one out of eight well-defined areas in the sky. Here, we
use {\it Gaia} DR2 data to perform a systematic exploration of these areas aimed at confirming or rejecting their prediction, focusing
on the values of the radial velocity to uncover possible multimodal distributions. This Letter is organized as follows. Section~2
presents the input data and tools used in our analysis. The eight radial velocity distributions are explored in Section~3. Section~4
presents the analysis of a bimodal Gaussian distribution found towards Galactic coordinates ($l$, $b$) = (271{\degr}, +2{\degr}) in
Vela. In Section~5, we study the statistical significance of our findings. Results are discussed in Section~6 and conclusions are
summarized in Section~7.
\section{Input data and tools}
\citet{2018MNRAS.474.4584G} have proposed that, if the Monoceros Ring has a dwarf galaxy progenitor, it might be found towards one out
of eight patches of sky. In their fig. 2, three patches are linked to their Model 2, two to their Model 3, and three more to their
Model 4. Model 2 predicts a progenitor located in the range of Galactocentric distances, $d$, 27$\pm$12~kpc, Model 3 predictions are
for the range 40$\pm$12~kpc, and those from Model 4 span the range 31$\pm$13~kpc. Model 2 indicates three possible locations centred at
Galactic coordinates ($l$, $b$) of (14{\degr}, $-$1{\degr}) in the constellation of Sagittarius (henceforth Model 2 A), (352{\degr},
$-$2{\degr}) in Scorpius (Model 2 B), and (232{\degr}, +2{\degr}) in Puppis (Model 2 C); Model 3 gives (11{\degr}, $-$2{\degr}) in
Sagittarius (Model 3 A) and (354{\degr}, $-$1{\degr}) in Scorpius (Model 3 B); Model 4 suggests (17{\degr}, $-$2{\degr}) in Scutum
(Model 4 A), (249{\degr}, $-$1{\degr}) in Puppis (Model 4 B), and (271{\degr}, +2{\degr}) in Vela (Model 4 C). Our objective is to
analyse data in a three-dimensional sub-space (eight of them) constrained by the values of $l$, $b$ and $d$ in \citet{2018MNRAS.474.4584G}.
{\it Gaia} DR2 \citep{2016A&A...595A...1G,2018A&A...616A...1G} provides extensive astrometric and photometric data ---namely,
coordinates, parallax, radial velocity, proper motions, blue, red and green magnitudes, and their respective standard errors--- that
can be used to perform the exploration required to confirm or reject the prediction made by \citet{2018MNRAS.474.4584G}. Out of
87\,733\,672 sources with strictly positive values of the parallax, we found 4\,831\,766 sources with values of the radial velocity,
$V_r$. This is our primary sample and we have been able to extract 475, 878, 102, 252, 247, 344, 199, and 150 sources, respectively,
for the eight patches of sky singled out by \citet{2018MNRAS.474.4584G}. The extraction process has been restricted to a square region
of side 10{\degr}, centred at the proposed locations; i.e. for Model 2 A, the ranges in $l$ and $b$ are, respectively, (9{\degr},
19{\degr}) and ($-$6{\degr}, +4{\degr}) with distances within 27$\pm$12~kpc. We believe that the size of our sampling window is large
enough to obtain robust results (see fig. 2 in \citealt{2018MNRAS.474.4584G}).
When needed and following the approach outlined by \citet{1987AJ.....93..864J}, we have used the input data to compute Galactic space
velocities and their uncertainties in a right-handed coordinate system for $U$, $V$, and $W$; axes are positive in the directions of
the Galactic centre, Galactic rotation, and the North Galactic Pole (NGP). The Galactocentric standard of rest is a right-handed
coordinate system centred at the Galactic centre. Axes are positive in the directions of the Sun, Galactic rotation, and the NGP. The
necessary values of the Solar motion were taken from \citet{2010MNRAS.403.1829S}. Averages, standard deviations, medians, interquartile
ranges (IQRs) and other statistical parameters have been computed in the usual way (see e.g. \citealt{2012psa..book.....W})
\section{Velocity distributions}
Fig.~\ref{vrdist} shows the radial velocity distributions of the eight {\it Gaia} DR2 samples associated with the possible locations of
the progenitor of the Monoceros Ring discussed by \citet{2018MNRAS.474.4584G}. The histograms displayed here have a statistically
meaningful bin width computed using the Freedman-Diaconis rule \citep{FD81}, i.e. $2\ {\rm IQR}\ n^{-1/3}$, where $n$ is the number of
sources; cumulative frequencies are plotted as dashed curves. The values of the IQRs (median values in parentheses) in km~s$^{-1}$ are,
respectively (top to bottom in Fig.~\ref{vrdist}), 110.27 (+31), 134.93 ($-$56.98), 32.51 (+105.22), 127.33 (+15.82), 138.47 ($-$58.01),
99.33 (+31.83), 32.99 (+106.14), and 44.56 (+75.78), and the bin sizes of the associated histograms are, respectively (also in
km~s$^{-1}$), 28.26, 28.18, 13.92, 40.31, 44.14, 28.35, 11.30, and 16.77. Given the fact that the values of the IQRs for the different
regions are quite different, our analysis is sensitive to finding kinematically cold substructures only in a few of them.
As the Galactic latitude of the various samples corresponds to the disc, the presence of a single peak in the radial velocity
distribution can be interpreted as the signature of a disc-like population; if additional peaks are found, they may correspond to
coherent structures different from the disc. As a reference, the solid lines in Fig.~\ref{vrdist} show the distributions of synthetic
samples of field stars generated from the \textsc{galaxia}\footnote{\url{http://www.physics.usyd.edu.au/~sanjib/code/}} model
\citep{2011ApJ...730....3S} within one square degree around the region's centre (thin red curve, entire synthetic sample, thick black
curve, synthetic sample with Galactocentric distance $>$ 15~kpc, normalized to the peak of the corresponding histogram). In general and
with the exception of the sample of location C of Model 4 in \citet{2018MNRAS.474.4584G} ---see Fig.~\ref{vrdist}, bottom panel--- it
is not possible to identify any statistically significant bimodal behaviour. The error bars in the bottom panel of Fig.~\ref{vrdist}
have been calculated using Poisson statistics ($\sigma=\sqrt{n}$) and applying the approximation given by \citet{1986ApJ...303..336G}
when $n<21$, $\sigma \sim 1 + \sqrt{0.75 + n}$.
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\begin{figure}
\centering
\includegraphics[width=\linewidth]{fdistVrxn.eps}
\caption{Radial velocity distributions of the {\it Gaia} DR2 samples associated with the eight possible progenitor locations of the
Monoceros Ring discussed in \citet{2018MNRAS.474.4584G}. Solid line distributions from \textsc{galaxia}
\citep{2011ApJ...730....3S} with warp and flare. See the text for details.
}
\label{vrdist}
\end{figure}
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\section{A bimodal distribution of radial velocities in Vela}
The bottom panel in Fig.~\ref{vrdist} seems to follow a bimodal Gaussian distribution produced by populations with central velocities
60$\pm$7~km~s$^{-1}$ and 97$\pm$10~km~s$^{-1}$ and consistent dispersions of about 10~km~s$^{-1}$. Assuming that the observed
distribution is mainly the sum of two normal distributions, the bimodality index (difference of means divided by dispersion) is 3.7,
and the ratio of the two populations is nearly 1:1. We interpret the population with the lowest central velocity as an extension of
the thin disc at average Galactocentric distances close to 30~kpc. This interpretation appears to be compatible with predictions from
\textsc{galaxia}, but the higher $V_r$ peak is neither compatible with the thin disc (thin curve) nor with the halo component (thick
curve), which should have values close to 150~km~s$^{-1}$. Fig.~\ref{pm} shows that both populations (low velocity in green, high
velocity in red) have similar distributions in terms of proper motions ($\mu_{\alpha} \cos\delta$, $\mu_{\delta}$), although the
high-velocity group exhibits a larger dispersion in the values of their proper motions in right ascension ---averages, standard
deviations, medians, and IQRs are (disc members) $-$3.9$\pm$0.7~mas~yr$^{-1}$, $-$3.78~mas~yr$^{-1}$, 0.77~mas~yr$^{-1}$ (right
ascension), 3.2$\pm$0.8~mas~yr$^{-1}$, 3.19~mas~yr$^{-1}$, 0.88~mas~yr$^{-1}$ (declination), and (higher $V_r$ peak)
$-$3.8$\pm$1.1~mas~yr$^{-1}$, $-$3.73~mas~yr$^{-1}$, 1.24~mas~yr$^{-1}$ (right ascension), 3.3$\pm$0.9~mas~yr$^{-1}$,
3.24~mas~yr$^{-1}$, 1.11~mas~yr$^{-1}$ (declination). As the values of their distances span the same range ---averages, standard
deviations, medians, and IQRs for the parallaxes are (disc) 0.039$\pm$0.009~mas, 0.039~mas, 0.018~mas, and (higher $V_r$ peak)
0.040$\pm$0.011~mas, 0.040~mas, 0.019~mas--- their tangential velocities are also similar. The colour-magnitude diagram in
Fig.~\ref{cmd} displays the intrinsic values (i.e. corrected for extinction and reddening using the data provided by {\it Gaia} DR2) of
these two populations, which follow similar distributions made of probable young stars, perhaps classical Cepheids, although the
high-velocity population seems to be slightly older and/or have different metallicity. Fig.~\ref{cmd} is similar to figs~19b and 20 in
\citet{2018A&A...616A...8A} in which the red giant branch (RGB) corresponds to an age of 1--2~Gyr. As a reference, three 50 Myr and one
1.5 Gyr PARSEC v1.2S + COLIBRI PR16\footnote{\url{http://stev.oapd.inaf.it/cgi-bin/cmd}} isochrones are also plotted. Most sources in
the higher $V_r$ peak of the bottom panel of Fig.~\ref{vrdist} are too bright to be RGB stars.
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\begin{figure}
\centering
\includegraphics[width=\linewidth]{fpmM4C2n.eps}
\caption{Proper motion components with 1$\sigma$ error bars for the {\it Gaia} DR2 sample of location C of Model 4 in
\citet{2018MNRAS.474.4584G}: green triangles, stars assumed to be part of the disc (45~km~s$^{-1}$$<V_r<$ 75~km~s$^{-1}$),
red squares, stars linked to the high-velocity peak in the bottom panel of Fig.~\ref{vrdist} (80~km~s$^{-1}$$<V_r<$
120~km~s$^{-1}$), others plotted as black circles.
}
\label{pm}
\end{figure}
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\begin{figure}
\centering
\includegraphics[width=\linewidth]{fhrM4C2n.eps}
\caption{Colour-magnitude diagram for the same sample in Fig.~\ref{pm}, similar to fig. 5 in \citet{2018A&A...616A..10G}. The
entire {\it Gaia} DR2 sample of location C of Model 4 is plotted in light grey. Three
isochrones of an age of 50~Myr and solar (cyan), solar/5 (blue), and solar/50 (purple) metallicities and one of an age of
1.5~Gyr (brown) and solar metallicity from \citet{2017ApJ...835...77M} are plotted as a reference.
}
\label{cmd}
\end{figure}
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\section{Statistical significance}
Fig.~\ref{vrdist}, bottom panel, shows that the sample from Model 4 C is bimodal with central values 60$\pm$7~km~s$^{-1}$ and
97$\pm$10~km~s$^{-1}$. The separation between the peaks is 34~km~s$^{-1}$, which is of order of IQR and well above the usual values of
the uncertainties in the values of the radial velocities (well below 5~km~s$^{-1}$ in most cases, see Fig.~\ref{full}, bottom panel).
The difference between the bin centred at 58.7~km~s$^{-1}$ and the one at 75.5~km~s$^{-1}$ is nearly 4$\sigma$; in addition, the excess
of the bin centred at 92.2~km~s$^{-1}$ with respect to the one at 75.5~km~s$^{-1}$ is about 3.5$\sigma$ ---in both cases using the
$\sigma$-value at the dip. If we repeat the same analysis for Model 2 B (Fig.~\ref{vrdist}, second to top panel), the single other that
might exhibit signs of bimodality, excesses of just 1.1$\sigma$ are found.
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\begin{figure}
\centering
\includegraphics[width=\linewidth]{m4cFULLn.eps}
\caption{Distribution in Galactic coordinates (top panel), distance (second to top panel), and tangential velocity (second to
bottom panel) for the entire {\it Gaia} DR2 sample of location C of Model 4 in \citet{2018MNRAS.474.4584G}. The colour map
shows the value of the colour index in Fig.~\ref{cmd}, computed as shown in \citet{2018A&A...616A..10G}. Hot stars appear
lighter in colour and cold stars darker. Points in green and red as in Figs~\ref{pm} and \ref{cmd}. The radial velocity
distribution with error bars is shown in the bottom panel.
}
\label{full}
\end{figure}
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\section{Discussion}
In general, the data from {\it Gaia} DR2 for sources beyond 15~kpc are affected by large uncertainties, particularly large in the case
of the parallax values (i.e. distances). Our main findings are based on a parameter, the value of the radial velocity, which is
definitely the least uncertain of the data set, if measured. The samples used in this investigation have very good values of the radial
velocity and also the proper motions, but the values of the distances derived from the parallaxes are rather uncertain. The values of
the tangential velocities and absolute magnitudes inherit these large errors. Therefore, our conclusions must be based on the value of
the radial velocity and those of the components of the proper motion. These are sufficiently reliable.
It may be argued that both peaks in the distribution of radial velocity in Fig.~\ref{vrdist}, bottom panel, could be the result of the
presence of small, distant, and decaying clusters such as E 3 \citep{2015A&A...581A..13D}. In such cases, we should observe very small
concentrations of relevant sources. Fig.~\ref{full}, top panel, shows the distribution of low- and high-velocity sources in Galactic
coordinates and no such obvious concentrations are observed. The same can be said about the values of the distance in Fig.~\ref{full},
second to top panel. In Fig.~\ref{full}, both the sample with radial velocities and the most general one are plotted using the
intrinsic colour index in Fig.~\ref{cmd} as third coordinate. Fig.~\ref{full}, top panel, clearly indicates that the effects of
extinction are very important in this region with the empty spaces corresponding to putative missing sources not present in {\it Gaia}
DR2 because the foreground population and associated gas clouds effectively block the light from these sources. The fact that the
sources surrounding the empty spaces tend to have the largest reddening values, clearly confirms that reddening is a major concern
here. It is however unlikely that the dominant role that interstellar extinction indeed has in this region may have a major impact on
our conclusions, which are based on the distribution of radial velocities for the most part. Fig.~\ref{full}, second to bottom panel,
shows the values of the tangential velocity, $V_t$, which exhibit a considerable spread due to the large uncertainties in the values of
the distances.
Fig.~\ref{galactovs} makes the issue of the uncertainties in the values of the distance explicit. In general, the Galactocentric
velocity components have very large error bars and this makes it virtually impossible to disentangle the two structures that may
contribute to the bimodal radial velocity distribution observed in Fig.~\ref{vrdist}, bottom panel. The prominent feature with
central value 97$\pm$10~km~s$^{-1}$, may signal the presence of the long sought progenitor of the Monoceros Ring, but the data might
also be compatible with the existence of an unrelated, previously unknown, kinematically coherent structure, such as a distant spiral
arm or a spur that connects two well-separated arms. In Fig.~\ref{full}, top panel, the distribution above the disc looks quite
stream-like, as opposed to cluster-like. The location in the sky of structure B in fig.~5 of \citet{2004MNRAS.348...12M} matches
somehow that of the feature discussed here, but the distance range appears to be different. The same can be said about the Argo
structure discussed by \citet{2006ApJ...640L.147R}.
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\begin{figure}
\centering
\includegraphics[width=\linewidth]{fuvwM4Cn.eps}
\includegraphics[width=\linewidth]{fuvwM4Cwen.eps}
\caption{Galactocentric velocity components with (top two panels) and without error bars (bottom two panels). Triangles in green
and squares in red as in Figs~\ref{pm}--\ref{full}.
}
\label{galactovs}
\end{figure}
%
%-------------------------------------------------------------------------------------------------------------------------------------------
%
\section{Conclusions}
In this Letter, we have investigated the plausibility of the predictions made by \citet{2018MNRAS.474.4584G} regarding the possible
location of the putative progenitor of the Monoceros Ring using data from {\it Gaia} DR2 and focusing on the distribution of radial
velocities. A statistically robust feature in the radial velocity distribution of stars in Vela has been identified. Our conclusions
are:
\begin{enumerate}[(i)]
\item Based on the distributions of radial velocities provided by {\it Gaia} DR2, the distant stellar populations located in the
regions proposed by \citet{2018MNRAS.474.4584G} appear to be compatible with single populations (i.e. no kinematically
heterogenous samples) in all but one case, that of the region towards the constellation of Vela.
\item We have identified a statistically significant bimodal Gaussian distribution towards Galactic coordinates ($l$, $b$) =
(271{\degr}, +2{\degr}), which is one of the present-day locations of the progenitor of the Monoceros Ring proposed by
\citet{2018MNRAS.474.4584G}. This feature has central values of the radial velocity of 60$\pm$7~km~s$^{-1}$ and
97$\pm$10~km~s$^{-1}$.
\item The prominent feature found towards Vela may signal the presence of the long sought progenitor of the Monoceros Ring, but
the data might also be compatible with the existence of an unrelated, previously unknown, kinematically coherent structure.
\item Interstellar extinction may be a major obstacle to disentangle the true nature of the two remote populations that appear to
share the patch of sky around ($l$, $b$) = (271{\degr}, +2{\degr}).
\end{enumerate}
\section*{Acknowledgements}
We thank the anonymous referee for a particularly insightful and constructive review, and A.~I. G\'omez de Castro for comments and for
providing access to computing facilities; RdlFM thanks L. Beitia-Antero for extensive discussions on {\it Gaia} DR2 data. This work was
partially supported by the Spanish MINECO under grant ESP2015-68908-R. In preparation of this Letter, we made use of the NASA
Astrophysics Data System, the ASTRO-PH e-print server, and the SIMBAD and VizieR databases operated at CDS, Strasbourg, France. This
work has made use of data from the European Space Agency (ESA) mission {\it Gaia} (\url{https://www.cosmos.esa.int/gaia}), processed by
the {\it Gaia} Data Processing and Analysis Consortium (DPAC, \url{https://www.cosmos.esa.int/web/gaia/dpac/consortium}). Funding for
the DPAC has been provided by national institutions, in particular the institutions participating in the {\it Gaia} Multilateral
Agreement.
\begin{thebibliography}{99}
\bibitem[\protect\citeauthoryear{Andrae et al.}{2018}]{2018A&A...616A...8A} Andrae R. et al., 2018,
A\&A, 616, A8
\bibitem[\protect\citeauthoryear{Carraro et al.}{2005}]{2005ApJ...630L.153C} Carraro G., V{\'a}zquez R.~A., Moitinho A., Baume G., 2005,
ApJ, 630, L153
\bibitem[\protect\citeauthoryear{Conn et al.}{2007}]{2007MNRAS.376..939C} Conn B.~C. et al., 2007,
MNRAS, 376, 939
\bibitem[\protect\citeauthoryear{Dame \& Thaddeus}{2011}]{2011ApJ...734L..24D} Dame T.~M., Thaddeus P., 2011,
ApJ, 734, L24
\bibitem[\protect\citeauthoryear{de la Fuente Marcos et al.}{2015}]{2015A&A...581A..13D} de la Fuente Marcos R., de la Fuente Marcos C.,
Moni Bidin C., Ortolani S., Carraro G., 2015,
A\&A, 581, A13
\bibitem[\protect\citeauthoryear{Freedman \& Diaconis}{1981}]{FD81} Freedman D., Diaconis P., 1981,
Z. Wahrscheinlichkeitstheor. verwandte Geb., 57, 453
\bibitem[\protect\citeauthoryear{Gaia Collaboration}{2016}]{2016A&A...595A...1G} Gaia Collaboration, 2016,
A\&A, 595, A1
\bibitem[\protect\citeauthoryear{Gaia Collaboration}{2018a}]{2018A&A...616A...1G} Gaia Collaboration, 2018a,
A\&A, 616, A1
\bibitem[\protect\citeauthoryear{Gaia Collaboration}{2018b}]{2018A&A...616A..10G} Gaia Collaboration, 2018b,
A\&A, 616, A10
\bibitem[\protect\citeauthoryear{Gehrels}{1986}]{1986ApJ...303..336G} Gehrels N., 1986,
ApJ, 303, 336
\bibitem[\protect\citeauthoryear{Grillmair, Carlin \& Majewski}{2008}]{2008ApJ...689L.117G} Grillmair C.~J., Carlin J.~L., Majewski S.~R., 2008,
ApJ, 689, L117
\bibitem[\protect\citeauthoryear{Guglielmo et al.}{2018}]{2018MNRAS.474.4584G} Guglielmo M., Lane R.~R., Conn B.~C., Ho A.~Y.~Q.,
Ibata R.~A., Lewis G.~F., 2018,
MNRAS, 474, 4584
\bibitem[\protect\citeauthoryear{Ibata, Gilmore \& Irwin}{1994}]{1994Natur.370..194I} Ibata R.~A., Gilmore G., Irwin M.~J., 1994,
Nature, 370, 194
\bibitem[\protect\citeauthoryear{Ibata et al.}{2003}]{2003MNRAS.340L..21I} Ibata R.~A., Irwin M.~J., Lewis G.~F., Ferguson A.~M.~N.,
Tanvir N., 2003,
MNRAS, 340, L21
\bibitem[\protect\citeauthoryear{Ivezi{\'c} et al.}{2008}]{2008ApJ...684..287I} Ivezi{\'c} {\v Z}. et al., 2008,
ApJ, 684, 287
\bibitem[\protect\citeauthoryear{Johnson \& Soderblom}{1987}]{1987AJ.....93..864J} Johnson D.~R.~H., Soderblom D.~R., 1987,
AJ, 93, 864
\bibitem[\protect\citeauthoryear{Kalberla et al.}{2014}]{2014ApJ...794...90K} Kalberla P.~M.~W., Kerp J., Dedes L., Haud U., 2014,
ApJ, 794, 90
\bibitem[\protect\citeauthoryear{Laporte et al.}{2018}]{2018MNRAS.473.1218L} Laporte C.~F.~P., G{\'o}mez F.~A., Besla G., Johnston K.~V.,
Garavito-Camargo N., 2018,
MNRAS, 473, 1218
\bibitem[\protect\citeauthoryear{Li et al.}{2017}]{2017ApJ...844...74L} Li T.~S. et al., 2017,
ApJ, 844, 74
\bibitem[\protect\citeauthoryear{L{\'o}pez-Corredoira \& Molg{\'o}}{2014}]{2014A&A...567A.106L} L{\'o}pez-Corredoira M., Molg{\'o} J., 2014,
A\&A, 567, A106
\bibitem[\protect\citeauthoryear{Marigo et al.}{2017}]{2017ApJ...835...77M} Marigo P. et al., 2017,
ApJ, 835, 77
\bibitem[\protect\citeauthoryear{Martin et al.}{2004}]{2004MNRAS.348...12M} Martin N.~F., Ibata R.~A., Bellazzini M., Irwin M.~J.,
Lewis G.~F., Dehnen W., 2004,
MNRAS, 348, 12
\bibitem[\protect\citeauthoryear{Michel-Dansac et al.}{2011}]{2011MNRAS.414L...1M} Michel-Dansac L., Abadi M.~G., Navarro J.~F.,
Steinmetz M., 2011,
MNRAS, 414, L1
\bibitem[\protect\citeauthoryear{Momany et al.}{2006}]{2006A&A...451..515M} Momany Y., Zaggia S., Gilmore G., Piotto G., Carraro G.,
Bedin L.~R., de Angeli F., 2006,
A\&A, 451, 515
\bibitem[\protect\citeauthoryear{Newberg et al.}{2002}]{2002ApJ...569..245N} Newberg H.~J. et al., 2002,
ApJ, 569, 245
\bibitem[\protect\citeauthoryear{Pila D\'{\i}ez}{2015}]{Pila15} Pila D\'{\i}ez B., 2015,
Ph.D. Thesis, Univ. Leiden
\bibitem[\protect\citeauthoryear{Rocha-Pinto et al.}{2006}]{2006ApJ...640L.147R} Rocha-Pinto H.~J., Majewski S.~R., Skrutskie M.~F.,
Patterson R.~J., Nakanishi H., Mu{\~n}oz R.~R., Sofue Y., 2006,
ApJ, 640, L147
\bibitem[\protect\citeauthoryear{Schlaufman}{2011}]{2011PhDT.......171S} Schlaufman K.~C., 2011,
Ph.D. Thesis, Univ. California, Santa Cruz
\bibitem[\protect\citeauthoryear{Sch{\"o}nrich \& Dehnen}{2018}]{2018MNRAS.478.3809S} Sch{\"o}nrich R., Dehnen W., 2018,
MNRAS, 478, 3809
\bibitem[\protect\citeauthoryear{Sch{\"o}nrich, Binney \& Dehnen}{2010}]{2010MNRAS.403.1829S} Sch{\"o}nrich R., Binney J., Dehnen W., 2010,
MNRAS, 403, 1829
\bibitem[\protect\citeauthoryear{Sharma et al.}{2011}]{2011ApJ...730....3S} Sharma S., Bland-Hawthorn J., Johnston K.~V., Binney J., 2011,
ApJ, 730, 3
\bibitem[\protect\citeauthoryear{Sheffield et al.}{2018}]{2018ApJ...854...47S} Sheffield A.~A., Price-Whelan A.~M., Tzanidakis A.,
Johnston K.~V., Laporte C.~F.~P., Sesar B., 2018,
ApJ, 854, 47
\bibitem[\protect\citeauthoryear{Smith}{2016}]{2016ASSL..420..113S} Smith M.~C., 2016,
Tidal Streams in the Local Group and Beyond: Observations and Implications.
ASSL. Vol.\ 420, Springer International Publishing, Switzerland, p.\ 113
\bibitem[\protect\citeauthoryear{Sollima et al.}{2011}]{2011ApJ...730L...6S} Sollima A., Valls-Gabaud D., Martinez-Delgado D., Fliri J.,
Pe{\~n}arrubia J., Hoekstra H., 2011,
ApJ, 730, L6
\bibitem[\protect\citeauthoryear{Wall \& Jenkins}{2012}]{2012psa..book.....W} Wall J.~V., Jenkins C.~R., 2012,
Practical Statistics for Astronomers.
Cambridge Univ. Press, Cambridge
\bibitem[\protect\citeauthoryear{Wang et al.}{2018}]{2018MNRAS.478.3367W} Wang H.-F., Liu C., Xu Y., Wan J.-C., Deng L., 2018,
MNRAS, 478, 3367
\bibitem[\protect\citeauthoryear{Xu et al.}{2015}]{2015ApJ...801..105X} Xu Y., Newberg H.~J., Carlin J.~L., Liu C., Deng L., Li J.,
Sch{\"o}nrich R., Yanny B., 2015,
ApJ, 801, 105
\bibitem[\protect\citeauthoryear{Xue et al.}{2011}]{2011ApJ...738...79X} Xue X.-X. et al., 2011,
ApJ, 738, 79
\bibitem[\protect\citeauthoryear{Yanny et al.}{2003}]{2003ApJ...588..824Y} Yanny B. et al., 2003,
ApJ, 588, 824
\end{thebibliography}
\bsp
\label{lastpage}
\end{document}