normalized units via SI:
$x_0 =$ reference length, in $m$.
$t_0 =$ reference timescale, in $s$.
$m_0 =$ reference weight, in $kg$.
$q_0 =$ reference charge, in $C$.
$T_0 =$ reference temperature, in $K$.
normalized using (?):
$x_0 =$ reference length, in $m$.
$m_0 =$ reference mass, in $kg$. Is this defined in the simulation, or is it derived as $m_0 = B_0 q_0 t_0$?
$q_0 = e =$ reference charge, in $C$.
$B_0 =$ reference magnetic field strength, in $T = \frac{kg}{C \cdot s}$. Should this be derived from other constants, as $B_0 = \frac{m_0}{q_0 t_0}$?
$n_0 =$ reference number density, in $\frac{1}{m^3}$. This would scale with length units, as $n_0 \propto \frac{1}{x_0^3}$? Particle number density can be calculated as $n = \frac{\rho}{m}$.
...and derived reference values:
$d_D = \sqrt{\frac{\epsilon_0 u_0^2 m_0}{n_0 q_0^2}} =$ Debye length, in $m$.
$u_0 = \sqrt{\frac{d_D^2 n_0 q_0^2}{\epsilon_0 m_0}} =$ reference velocity, in $\frac{m}{s}$. This is circularly dependent on Debye length, so one of these has to be fixed in the simulation and the other derived.
$t_0 = \frac{x_0}{u_0} =$ reference timescale, in $s$.
$E_0 = c B_0 =$ reference electric field strength in $\frac{kg \cdot m}{C \cdot s^2}$. Should this be derived, just as $B_0$ should?
$d_L = \frac{m_0 u_0}{q_0 B_0} =$ Larmor radius, in $m$.
$T_0$ is not needed, as temperature units are not used.
So either $m_0$ is provided and $B_0$ is derived, or $B_0$ is provided and $m_0$ is derived (the paper hints at this).
Likewise, either $u_0$ or $d_D$ is provided and the other is derived. I suspect that $d_D$ is provided.
normalized using 2014 Abgrall, Kumar:
$x_0 =$ reference length, in $m$. Referenced in the paper but never explicitly defined.
$m_i =$ ion mass, "is assumed to be 1" (section 1), but later stated (in section 3.1) to be set to 2 ... Then we have $m = \frac{m_i}{m_e}$, makes me think this means $m_0 = m_i$ and $m_e = \frac{1}{1836.1526734311} m_i$, though the simulation assigns it $m_e = \frac{1}{25} m_i$ (section 3.2).
$v_0 = v_i^T =$ reference thermal velocity of ions, used to compute Debye length where $u_0$ is used in (?) above, so I'm assuming this is the equivalent of the reference velocity.
$\hat{l}_r = \frac{l_r}{x_0} = \frac{m_i v_i^T}{q_i B_0 x_0} =$ Larmor radius, is initialized as $10^{-2}, 10^{-4}, 10^{-6}$ in simulations (section 3.2).
derived reference values:
$t_0 = \frac{x_0}{v_0} =$ reference timescale, in $s$.
If $\hat{l}_r$ is provided and $x_0$ is provided then we must derive $l_r = \hat{l}_r x_0 = \frac{m_i}{v_i^T}{q_i B_0}$. From this, $m_i$ is already stated, so we are left with $v_i^T, q_i, B_0$