Models continue to be quite similar in their forecast of the evolution of the large-scale pattern through the weekend, so the smaller scale details will become important re what happens when and where. The forecast above is from the 06 UTC WRF-GFS forecast run at Atmo and is for total precipitation through 11:00 am MST on Sunday the 19th. The event is widespread for Arizona, but the model forecasts a considerable minimum over much of Pima County, with the Sky Islands of southeast Arizona capturing the heavier precipitation.
The 10-m wind forecast (below) indicates very strong, south-southeasterly winds at 02:00 pm on Saturday the 18th. These winds are downslope from the higher terrain along the Borderlands and appear to be the primary factor producing the donut hole in the precipitation pattern. The forecast sounding for Tucson (2nd below) shows a deep, dry boundary layer at the same time, with only a fairly shallow layer producing rainfall. The depth of vertical motion above the downslope winds will be very important as the event develops - deeper vertical motion below 500 mb favoring higher rainfall.
I also took a look at the total precipitation forecast from the 00 UTC WRF-GFS forecast, on the 5.4 km grid, and forecast above is of total event precipitation through noon on Monday, February 20th. Not much difference for Arizona in this forecast. The amounts of precipitation in California, however, are very significant. The structural failures at Oroville Dam have the state rushing to draw down the reservoir as quickly as possible, before the heavy precipitation begins with this event. The WRF-GFS forecasts extreme amounts above the dam (location shown below).
The model also forecasts very heavy amounts of rainfall in the northern portions of the LA Basin, across the Santa Barbara and Santa Maria areas and well north up the coast, reflecting the unusual influx of high moisture from the south-southeast. Much of southern California is under Flash Flood Watches as the event approaches.