i've been working with heidi newberg and also with collaborators from georgia state university who made the data set that we're going to be using and we're going to be talking about two massive radio mergers in local solar neighborhood dwarf stars perfect so the movie made it in so this is what a radio a radio merger looks like and so it goes quickly but there's going to be a dwarf galaxy that comes straight into the center of milky way like galaxy and then it's going to pass through and make these shells as it comes through and then over time these shells are going to mix as different energy levels of the radial merger pass back and forth between the sides of the galaxy and then shells become less noticeable as uh mixing process continues and then finally at something like say eight giga years which you can see in the top right corner there is going to be basically no shell structure left it's going to be essentially completely phased mixed if you looked for shells you weren't going to find any and so there's a phase mixing argument to be made which is uh is there substructure in the inner halo do we see that and if so then how do we explain that that substructure uh and that can tell us a lot about the time of the accretion of these things so uh you got a little bit of this from jesse's talk but just to jog your memory again there's the gaia sausage enceladus merger event which is currently thought to be basically all of the inner halo and it was this massive merger with a sizable dwarf galaxy that crashed into the milky way at early times and supposedly puffed up what was the milky way's proto disk and made this thin disc and so based on the ages of thick disk stars you say that this had to be about eight or more giga years ago it was discovered independently by a few different groups but it's characterized by this big sausage velocity structure which i've circled and read here which you saw in jesse's cuts and so this is like the canonical gaia sausage because apparently to some people that looks like a sausage and a meatball and so um we are going to be talking about another merger event that is possibly in the halo that we call the virgo radial merger and this was actually discovered in virgo over density stars which is the over density that jesse was talking about in the north so if you take stars in that over density and actually run their orbits and you can fit an end body to that orbit then you get something that looks like this in the local cellar neighborhood so instead of this big smooth sausage structure you get this kind of what we call dumbbells or double lobed thing and so it is puffed out on either side at about the same radial velocity and positive and negative and then there's fewer stars in the middle but the weird thing was that we were able to do this with just a two billion year end body simulation and in fact if you did it for too long if you did it for more than five the substructure went away in velocity and you didn't really get anything except just a double gaussian in the middle and so it's weird that we were able to get this sausage with something so young with virgo over density stars but uh so it raises the question of whether the virgo radio merge and the gaia sausage are the same thing are they different what's going on here so when we have this type of tension we go back to the data and so bucky and kim at georgia uh state university constructed a catalog of gaia edr3 dwarf stars within two kiloparsecs of the sun and you see clearly that there's this thick disk in red which we've cut out some of it by a proper motion cut and then there's this gaia sausage in the bottom that you see so we've got the sausage structure in the local solar data and for this data set we only have proper motions but that's okay because proper motions are going to be tangential to your line of sight and you get the two directions and so if you just look at the poles up and down those two tangential velocities are going to be oriented uh in your radial velocity and your rotational velocity so you can get your sausage plot back just from the proper motions if you're looking in the right way and then you get a metallicity for each star by making isochrones and then you plot those isochrones on a color magnitude diagram and then uh you take where each dwarf star lies on that cmd and you extrapolate its metallicity based on that grid and so we now have 5d space so no radial velocities but we also have metallicities for these and so we wanted to see what happens if you start to break these up in metallicity because there should be a clear metallicity signature of the gaia sausage so we do that so each of these panels is a split in fv on h which you see there up above each panel and what you get is really interesting is that if you ignore the thick disk in that red circle and you just look at those radial stars in that in that gray band you actually get that there's kind of like almost a smooth kind of sausage looking thing in the lower metallicities and then when you get up to minus 2.2 on the top right you get these clear velocity lobes you get that dumbbell structure that kind of pops out at you and then when you go back up to about minus one in effion h it goes away again and so what's happening here because in jesse's plot they had a peak on at fe on h at minus one but we don't see that we see that this this sausage structure is actually more prominent at about the minus 2.2 to minus 1.4 and so that's kind of weird and so we went back to the theory where we ran a whole bunch of radio merger simulations and so we said all right we maybe aren't quite sure what a radio merger event actually looks like if you're just looking at velocities in the solar neighborhood so each one of these is varying in certain parameters of radial mergers we vary things like progenitor mass we vary things like inclination um the final epigalactic and distance which is essentially energy of the collision and then we cut it all to the solar region and then we plotted it in a velocity so it's kind of like that sausage plot on the bottom panel and then there's a histogram of that up above and turns out inclination doesn't matter too much but when you vary the progenitor mass you either get as you go very high you get this kind of just single gaussian thing that kind of fits uh your your whole kind of halo but it looks almost circular the sausage plot doesn't give you like a sausage but as you go down in mass you get just this double lobed dumbbell thing and if you increase the distance or the energy from uh the galaxy then it kind of separates your double lobes and if you increase time it doesn't really mix very much which is the bottom row except in the very early times you see that the two red sides aren't the same height okay which means that it's still phase mixing so it looks very anisotropic there and so we're able to fit this very well uh with a double gaussian distribution which is in blue to the data which is in red and then we go back and we're going to start fitting those double gaussians to the data that we see so that's what we do uh we go and we fit two double gaussians to the data and we tried fitting more and less and we used a bayesian argument to justify doing two and you see that it it wants to fit uh two two double lobes or two double gaussians there's going to be one at about plus or minus 230 kilometers per second in vu velocity and then there's going to be one that is actually at about plus or minus 60. so it's not quite at zero but it looks almost like a single gaussian at zero and what you get is that there's some mixture of the two at very low metallicities between minus 2.2 and minus 1.4 it's very dominated by this this dumbbell shape and then as you go higher especially over -1 you just get this almost single gaussian shape and so it looks a lot like you have two things here because you can't get uh something that looks like uh just uh in in the previous simulations you just get either something that's in the middle or you get something that's not in the middle you get that double lobe we get a mixture which makes it seem like we have something that is uh there's one that's in the middle and there's one that is the double load structure and you can actually because we have this separated out by metallicity you can just look at the mdf for these structures and you get that there's that that big double lobe velocity structure that peaks at about minus 1.7 or so and then you get that single one that has a very not really dwarf galaxy looking mdf but it peaks at like -1 which is what you saw in jesse's data and so you can imagine if you took that that blue mdf and you kind of shifted it over to minus one then you would have a little bit of thick disk uh contamination and then you would have some junk over on the left at very metal pore that could potentially just be an accretion background of a very small minor mergers that were very metal poor but so the blue is actually consistent with our virgo radial merger and the orange is consistent with a gaia sausage along with other contaminations but the the key thing here is that actually the virgo radio merger and the gaia sausage stars are comparable in the vocal solar neighborhood they contribute roughly the same number of stars and so the conclusion is that uh this is a cautionary tale right you have to be really really careful about what you're pulling out when you're saying let's just go for the sausage thing let's do metallicities um because we believe that if you grab that sausage in velocity space you're really grabbing two different structures you're grabbing the thing on the side which is the virgo radio merger that actually makes it look like a sausage and you're grabbing that potentially very old thing that might have fluffed up the thick disk that is in the middle and so that's it thank you [Applause] what are your photometric metallicity errors and what bands do you use yeah so they're uh the bands are the gaia bprp and so then you make a cmd in like gaia g versus some color that you derive from gaia color um and the errors are something like 10 it can go up depending on how bright the star is um but it's not really too bad especially when you're doing like photometric metallicities and when you're doing a large number of stars too and you're just looking at relationships um with the large bins that we have in metallicity it tends to not matter too much thanks for the awesome talk um so one point that i had to want to ask you about was if you thought about the connection of the virgo over density in the hercules circular cloud so hse is another over density in the galactic south that um we think to be um along with vod the two apocentric polyps of stars from the gse merger and if you think that gse actually had an initial metallicity gradient because it is a dwarf galaxy or is it a galaxy then you could have a situation where the two lobes are the two over densities resulting from the first and fall will have a different metallicity peak than the rest of the stars yeah yeah that's a really good point so um when we fit star to the virgo over density the orbits actually do go and pass through the hercules equal cloud so you recover that second over density um and interestingly enough you actually get that the orbits pass through the herkus equal cloud in the north which is above the disk which is a large component of that over density that isn't actually explained by a big triaxial orientation of the halo so you can have something that is aligned in a triaxial halo like what you have and on top of it you can have substructure that's aligned with it because apparently everything the magellanic clouds are all aligned with it too um that that populates the same regions but in order to get a hercules of quilt cloud in the north you need it to not be very mixed because otherwise you just get this big triaxial thing like what you're seeing i just wanted to follow up that like we do have a in-house simulation of a radial merger like gsc and we do get like the two dod and hcc um like things but we can chat later in the yeah no i'd love to talk with the agency very nice talk uh i should talk to you too afterwards i have lots of thoughts but um the question i have is related to what what your are you near the turn off and what the magnitude of your stars are yeah so um the magnitudes are i don't remember off the top of my head what the magnitudes are uh i think we limit it to be brighter than i want to see a 18th mag but i'm not positive about that so the reason why i asked that question is because i'm going to say what i always say which is that chemistry is very helpful you can imagine a situation right where if you go to your highest metallicity bin and you look at this structure and you say your prediction is that the middle part is the sausage and the outer lobes are something different and even if you just look at alpha metallicity guy enceladus should be relatively large your lobe should be much smaller based on your metalistic distribution and so you'd expect a different alpha distribution which is a prediction that you can test automatically and in fact the data might already be in the h3 survey to do this yep yeah for sure right so i would check in with some of the alpha argument right so what i would do is plot the sausage in the highest middle city bin color code by alpha you should see alpha pop out differently in the lobes compared to in the center right yeah yeah yep that's a really good thing to do and we we didn't have access to alpha abundances when we did this but that's absolutely the next step for it for sure and i also mentioned this mostly because as well like we could get data for you when in a low resolution spectrograph we can go to g of 19 and still get alpha abundances right yeah it might have been i don't know what it was but um cool yeah work really fun jesse asked my first question keith asked my second question i was banking on keith asking a question so i get a third question um or kind of it's uh exciting connections kind of thing right so i remember you and heidi talking about this a year or two ago right when we were you were looking at the velocity separations and the little velocity structure in uh the virgo evidence so it seems to me like um there's the chemistry um there's i love jesse's idea of maybe it's a two-component single galaxy i love your idea of two so it does seem like the age and the dynamics but the dynamical aging is going to be key for those shell structures and i'm wondering like you're talking about how they fade i think i think what we want to do probably form a huge collaboration and you want to look at very high resolution simulations because then you don't want to run out of um you don't want to run out of of resolution to see the shells and similarly in the data so if we have enough data on high enough simulations i'm hoping that substructuring velocity in the shells might help us distinguish between two separate events or the same event that's a thought what do you think because you've done some of these i think no i think that is totally different it's tricky because you very quickly run into the resolution limits in the data for your shells you need very fine resolution in your radial velocities to um and when i say radial velocities i mean from the center of the galaxy so it depends on all three of your velocities and the errors propagate um to get to isolate these shells and so we actually did that i kind of glanced over it but um that's something that we looked at and we were able to find shells and kind of time based on how much they had mixed to say that this has to be a young merger event but with higher resolution comes more data and you're going to be able to find more shelves but they mix very very quickly so i really don't anticipate being able to find clear shells for an eight giga year old structure i would expect that it would be a mixed distribution