deletions | additions       diff --git a/quenching_time.tex b/quenching_time.tex  index 3a457a0..d48db3b 100644  --- a/quenching_time.tex  +++ b/quenching_time.tex 
...
We now translate the quiescent fractions in Figure~\ref{fig:quiescent_fraction} into the typical timescales over environmental processes quench satellites after they fall into a host halo, following the methodology of \citet{Wetzel2013}.  First, motivated by the dearth of isolated dwarf galaxies that are quiescent (see Introduction), our model assumes that all satellite dwarf galaxies were actively star-forming prior to firstvirial  infall. Because many galaxies $\mstar(z=0)<10^4\msun$ may be quenched at high redshift due to cosmic reionization \citep[e.g.,][]{Weisz2014a,Brown2014}, we do not model those masses.  Star-formation histories for satellites at $\mstar(z=0)=10^{4-5}\msun$ show a mix of them having completely quenched by $z\gtrsim3$ (e.g., Bootes I, Leo IV) and showing signs of star formation at $z\lesssim1$ (e.g., And XI, And XII) \citep{Weisz2014a,Brown2014} \textbf{XXX most importantly XII,  And XVI from Weisz et al. 2014c}, XVI) \citep{Weisz2014a,Weisz2014c,Brown2014},  so quenching at these masses may be driven by a mix of reionization and the host-halo environment. %Leo T had recent star formation, suggesting that galaxies at least down to logM_star ~ 5 can form stars today if not for environment.  Though, the 100\% quiescent fraction for satellites at this $\mstar$ means that if both processes are responsible, both are highly efficient, and that if satellites that were quenched by reionization have a similar virial-infall infall  time distribution as those that were quenched by the host-halo environment, it would not affect our results. Thus, we include this $\mstar$ in our modeling but label it distinctly to emphasize caution in interpretation.    Within each 1-dex bin of $\mstar$, we use the ELVIS simulations to compute the distribution of virial-infall infall  times that satellites at $z=0$ experienced. Assuming that environmental quenching likelihood correlates with time since infall, we designate those that fell in earliest as having been quenched, and we adjust the time-since-infall threshold for quenching until we match the observed quiescent fraction in each bin.  Several works have shown that this model successfully describes the dependence of satellite quiescent fractions on host-centric distance \citep[e.g.,][]{Wetzel2013, Wetzel2014, Wheeler2014} because of the correlation of virial-infall infall  time with host-centric distance \citep[e.g.,][]{Wetzel2015}. \textbf{XXX what about Rocha et al.? XXX} AW - they did not look at the correlation of infall time with distance, but rather as a function of orbital energy, unless there is something else that you had in mind.}  However, this correlation means that we should account for observed satellite's distances, including incompleteness for fainter satellites, in computing their infall times.  Thus, in selecting satellites in ELVIS, we only use those out to the maximum host-centric distance that they are observed from the MW or M31 at each $\mstar$ bin.  While this matters for the fainter satellites, it is most important for the highest masses, $\mstar=10^{8-9}$, at which all known satellites (M32, NGC 205, LMC/SMC) lie within $<61\kpc$ of the MW or M31.  Figure~\ref{fig:quench_times} shows the inferred environmental quenching timescales, that is, the time duration from firstvirial  infall to being fully quenched/gas-poor, as a function of $\mstar$. Blue circles show the satellite dwarf galaxies in the MW and M31, and we shade the lowest $\mstar$ bin lighter to highlight caution in interpretation, as explained above.  Error bars are derived from the 68\% uncertainty in the observed quiescent fractions in Figure~\ref{fig:quiescent_fraction}, which are typically larger than the host-to-host scatter in satellites' virial-infall infall  times in ELVIS. As explored in \citet{Wetzel2015}, many satellite dwarf galaxies first fell into a another host halo (group), typically of $\mvir=10^{10-12}\msun$, before falling into the MW/M31 halo.  Because the importance this environmental preprocessing in such lower-mass groups remains unclear, we present the inferred environmental quenching timescales both including and neglecting such group preprocessing. 
...
The red squares in Figure~\ref{fig:quench_times} show the timescales from \citet{Wheeler2014}, who used nearly identical methodology based on combining the the galaxy catalog from \citet{Geha2012} with satellite infall times (including group preprocessing) from the Millennium II simulation (REF).  They examined satellites with $\mstar\approx10^{8.5}$ and $10^{9.5}\msun$ around hosts with $\mstar>2.5\times10^{10}\msun$, which they found likely spans halos with $\mvir\approx10^{12.5-14}\msun$.  %$8.25<\log(\mstar/\msun)<8.75$ and $9.25<\log(\mstar/\msun)<9.65$  %\citet{Wheeler2014} defined the virial-infall infall  time of a satellite as the first time that it became a satellite, so their definition include group preprocessing, with the caveat that if a satellite orbits beyond its host, as defined by the FoF group, becoming a backsplash/ejected satellite, and then falls back into a host again, they include only the latter infall time. This is significantly higher than the MW/M31, which could mean that their quenching timescales are \emph{shorter} than for similar mass satellites of MW/M31-like hosts.  Similarly, the green curves in Figure~\ref{fig:quench_times} show the timescales for more massive satellites from \citet{Wetzel2013}, who also used identical methodology, combining galaxy groups from SDSS \citep{Tinker2011, Wetzel2012} with satellite infall times (including group preprocessing) from mock group catalogs in their cosmological simulation.