There are a couple of options (and I'm assuming that you're using this device in a switching application...so no need to go all the way into saturation):
1) In your application, look at the Rdson for the particular gate drive that you plan to use, and set the parameters accordingly. When you do this, it is very likely that the saturation current will be too high, but in a switching application, you shouldn't ever reach this point anyway. The Rdson may also not be quite correct for other values of gate drive, so there may be some inaccuracy that you have to live with using such a simple approach. For the capacitance model, you could consider using tabulated Ciss, Crss and Coss and just input the data from one of the plots on the datasheet.
2) If you're not happy with the fit from the simple square-law model (which is really for long-channel devices...and not so much for HV devices), you could also consider (mis-)using the N-Channel IGBT block, which has a tabulated option. In this block, you can enter tabulated data for the I-V and C-V curves. In principle, you can model any three-terminal, voltage-controlled device with this block if you have tabulated data...even though the icon doesn't really match the device that you're using.
If you really need to get the variation of Rdson with Vgs and any quasi-saturation effects, I would opt for the tabulated option. The alternative is to try an LDMOS model (which is MUCH more difficult to parameterize and simulate) or to add a bias dependent resistor in series with the drain. Both of these options are rather involved though and are probably unnecessary for a switching application.
