Made-to-Meaure Galaxia is an easy way for anyone to generate mock samples of stars from made-to-measure models of the Milky Way. The made-to-measure method adjusts an N-body model attempting to match it to a set of constraints. Usually this means adjusting an N-body model until it fits data in a maximum likelihood sense.

The modelling process is flexible and so can handle the non-axisymmetric bar and bulge, a region that is
otherwise very difficult to model. All the models available here were fitted to Milky Way data from the central
~5kpc, with the result that they hopefully are good models of the inner Galaxy. Importantly the models are all
*dynamically self-consistent*, meaning that the particles move in the gravitational potential that they
themselves generate.

This model is based on the best dynamical model from Portail et al. 2017a. Each N-body particle was
split into five different metalicity components represented as Gaussians. The fraction of mass in
each metalicity component was adjusted using the made-to-measure method to fit chemokinematic data
from the ARGOS and APOGEE surveys. For use in Galaxia the model was resampled to be composed of
10^{8} particles of roughly equal mass. Every particle has a Gaussian log-age distribution
centered at 10^{10} years with a width of 0.3dex. Full details of the model can be found in
the original paper :
Chemodynamical modelling of the galactic bulge and bar

These dynamical models were fit to a range of data on stars in the central 5kpc of the Galaxy:
spectroscopy from the ARGOS, APOGEE and BRAVA surveys, together with star counts from VVV, UKIDSS
and 2MASS. In addition constraints on the density of stars in the disk and rotation curve were
included. A grid of models were run varying the key unknown parameters of the models: The pattern
speed of the bar/bulge, the mass of the nuclear star cluster, and the mass per red clump star.
Because the models were fit to star counts of red clump stars, this mass per red clump star (or
mass-to-clump ratio) is the equivalent of the mass-to-light which is often a key ingredient in the
modelling of external galaxies. The best model had a pattern speed of 40km/s/kpc, a nuclear star
cluster mass of 2x10^{9}M⊙, mass-to-clump ratio 1000M⊙. For use in Galaxia
every model in the grid was resampled to be composed of 2x10^{7} particles of roughly equal
mass. Every particle has a Gaussian log-age distribution centered at 10^{10} years with a
width of 0.3dex and a Gaussian metalicity distribution with <[Fe/H]>=0 and stddev([Fe/H])=0.3. Full
details of the model can be found in the original paper : Dynamical modelling of
the galactic bulge and bar: the Milky Way's pattern speed, stellar and dark matter mass
distribution.

These dynamical models are superceeded by the Portail et al. 2017a models. They are included here
for reference, but the newer models should be preferred. They were fit to spectroscopy from the
BRAVA survey, together with star counts from VVV. The models contain a range of dark matter
fractions in the inner Galaxy. The total mass in the bulge was very well constrained to be (1.84±0.07)x10^{10}M⊙,
however the split between dark and stellar matter was degenerate, with all the different models
fitting the data well. For use in Galaxia each model was resampled to be composed of
2x10^{7} particles of roughly equal mass. Every particle has a Gaussian log-age distribution
centered at 10^{10} years with a width of 0.3dex and a Gaussian metalicity distribution with
<[Fe/H]>=0 and stddev([Fe/H])=0.3. Full details of the model can be found in the original paper: Made-to-measure models of
the Galactic box/peanut bulge: stellar and total mass in the bulge region

These are the default models included in Galaxia,
which aim to be close to the Besancon models of Robin et al. 2003.
They are included on this website only to allow for example easy comparison to an axisymmetric model. Full
details are in Sharma et
al. 2011.

Note that these are N-body *dynamical models and ***not** simulations. Simulations evolve initial
conditions using physical laws, giving the simulator insight on how the final simulation state came
about. However, this final simulation state cannot be expected to well fit data on our Milky Way
very well. Instead the models available here were tailored to fit Milky Way data as closely as
possible. They aim to accurately describe the current state of the Galaxy, and not its history.