What can the alignments of the velocity moments tell us about the nature of the potential?

Sep 30, 2015
11 pages
Published in:
  • Astrophys.J. 816 (2016) 1, 35
  • Published: Dec 29, 2015
e-Print:

Citations per year

20162017201812
Abstract: (IOP)
We prove that, if the time-independent distribution function F(v,r){ \mathcal F }({\boldsymbol{v}},{\boldsymbol{r}}) of a steady-state stellar system is symmetric under velocity inversion such that F(v1,v2,v3,r)=F(v1,v2,v3,r){ \mathcal F }(-{v}_{1},{v}_{2},{v}_{3},{\boldsymbol{r}})={ \mathcal F }({v}_{1},{v}_{2},{v}_{3},{\boldsymbol{r}}) and the same is true for v(2) and v(3), where (v1,v2,v3)({v}_{1},{v}_{2},{v}_{3}) is the velocity component projected onto an orthogonal frame, then the potential within which the system is in equilibrium must be separable (i.e., the Stäckel potential). Furthermore, we find that the Jeans equations imply that, if all mixed second moments of the velocity vanish, that is, vivj=0\langle {v}_{i}{v}_{j}\rangle =0 for any iji\ne j, in some Stäckel coordinate system and the only non-vanishing fourth moments in the same coordinate are those in the form of vi4\langle {v}_{i}^{4}\rangle or vi2vj2\langle {v}_{i}^{2}{v}_{j}^{2}\rangle , then the potential must be separable in the same coordinates. Finally we also show that all second and fourth velocity moments of tracers with an odd power to the radial component vr being zero is a sufficient condition to guarantee the potential to be of the form Φ=f(r)+r2g(θ,ϕ){\rm{\Phi }}=f(r)+{r}^{-2}g(\theta ,\phi ).
Note:
  • accepted for publication in the ApJ; see also arXiv:1509.08794
  • Galaxy: kinematics and dynamics
  • methods: analytical