The magneto-rotational instability (MRI; Balbus & Hawley 1991) is the most likely mechanism for producing turbulently-driven accretion in many astrophysical disks. However, because of the low ionization of protoplanetary disks in most of the planet-forming regions renders the MRI ineffective, at least in the midplane. Moreover, recent numerical simulations including non-ideal MHD effects suggest that the MRI can be completely suppressed between ~0.5 and 20~AU (Bai & Stone 2013; Gressel et al. 2015).
Low-mass stars form through gravitational collapse of molecular cloud cores. Not all the infalling material directs to the protostar at the center, but some portion of it falls onto a disk surrounding the protostar in order to conserve angular momentum of the natal cloud core. During this infall phase, the disk can become gravitationally unstable because of the mass loading. The gravitational instability (GI) can initiate, trying to redistribute the disk material so that the disk becomes stable again. When GI triggers, the spiral waves launched by the instability not only transport material but also shock the disk gas, raising the disk temperature. If the GI lasts long enough, the disk can be hot enough to thermally ionize the disk material, producing ions that can couple with the magnetic fields. This can trigger the MRI that would otherwise be not operational, enhancing the efficiency of disk accretion significantly.
The above figure shows the moment the GI-driven spiral arms heat the disk hot enough to thermally trigger the MRI. As shown in the middle panel, the disk temperature inside of a couple of AUs have increased well beyond 1000K due to the PdV (i.e., compressional) heating by the GI-driven spiral arms. On the right panel shows the time evolution of the accretion rate at the disk inner boundary, and you can see that the accretion rate is skyrocketing from about to about . See also an animated version here.