We present a wide-field ( 4.5 deg ^ { 2 } ) photometric and spectroscopic survey of the Leo I dwarf spheroidal ( dSph ) galaxy to explore its extended morphology and dynamics .
As in previous papers in this series , we take advantage of photometry in the M , T _ { 2 } , and DDO 51 filter system to select Leo I red giant branch star candidates , and , so far , this selection technique has proven 100 % reliable in selecting actual Leo I members among more than 100 M < 21.5 Leo I giant candidates having previous or new Keck DEIMOS spectroscopy to a radius > 1.3 times the limiting radius of the fitted , central King profile .
The two-dimensional distribution of all similarly-selected Leo I giant candidates is well fitted by a central single-component King profile of limiting radius 13.3 arcmin , but many giant stars are found outside this newly derived King limiting radius .
The density profile thus shows a break at a major axis radial distance of \sim 10 arcmin produced by an excess of stars at and beyond the King limiting radius ( spectroscopically confirmed to be made of true Leo I members ) , and primarily along the major axis of the main body of the rather elongated satellite .
This spatial configuration , a rather flat velocity dispersion profile and an asymmetric radial velocity ( RV ) distribution among the Leo I members at large radii together support a picture where Leo I has been tidally disrupted on at least one , but at most two , perigalactic passages of a massive Local Group member .
We demonstrate this hypothesis using mass-follows-light , N -body simulations of satellites in a Milky Way-like potential that reproduce the observed structural and dynamical properties of Leo I remarkably well .
These models include \sim 3 \times 10 ^ { 7 } solar mass , tidally disrupting dSph systems on bound orbits with rather high eccentricity ( 0.93–0.96 ) and small perigalactica ( 10-15 kpc ) .
The simulations allow the first observationally constrained orbit for Leo I without the measurement of its proper motion and show that the observed RV distribution is more consistent with a two Milky Way orbit history for the satellite while ruling out a Leo I orbit that includes a previous association with M31 within the last 10 Gyr .
Given the overall success of tidally disrupting mass-follows-light satellite models to account for the observed properties of Leo I , we conclude that there is no need to invoke an extended dark matter halo around the satellite ( e.g. , as one explanation of the velocity dispersion and radial profiles at large radii ) , and that an overall modest M / L for the satellite is consistent with the available data .
That a satellite on such a large ( apogalacticon of \sim 450 kpc ) , long period ( P \sim 6 Gyr ) orbit as Leo I can experience tidal disruption suggests that similarly structured satellites with even smaller ( eccentric ) orbits will experience even greater tidally-induced mass loss rates .