First, a look at yesterday's east winds. The WRF-GFS forecast from Atmo yesterday morning was quite good wrt the gusty east winds - which continued at the airport into the early morning hours today. Plot above shows observed wind time-series at airport (blue), with gusts indicated by the black dots. Max gust at airport was 33 mph; DM 31 mph; and Mt. Hopkins 42 mph.
The 06 UTC GEFS TUS plumes for QPF, above, are interesting for several reasons. The timing for the New Year's Day event has shifted into the daytime hours. The GEFS models continue to forecast 100 % probability at the airport for both tonight and New Year's Day - but that POP is only for at least 0.01". Even though the plumes are clustered tightly in time, there is still considerable spread in amounts of precipitation forecast by the ensemble members, especially on New Year's Day (range of amounts is from 0.40 to 1.00 inches). The past three mornings the operational GFS operational version (blue) of the model has been very inconsistent, going from driest, to close to average (black), to wettest now. Current NWS morning forecast today is shown below for the airport grid.
Even though the operational GFS has become wettest forecast, the WRF-GFS high resolution, QPF forecast has decreased substantially during the past 24-hours. Forecast above is for total precipitation through 07:00 am MST on January 2nd. Heaviest amounts are now forecast for the Rim Country and for the Sky Islands. Decrease in forecast amounts at low elevations in eastern Pima County since yesterday are mostly due to the closed low off the west coast taking a more northerly track across Arizona tonight.
The SREF plumes from SPC include a nice graphic that shows the change in the average QPF amount from the 26 members as a function of changing forecast initialization times (a sort of dprog/dt). This graphic is shown below (at TUS) for the last four initialization times - red 29/15 UTC; purple 29/21 UTC; green 30/03 UTC; and blue 30/09 UTC. The first two runs were fairly consistent (the grid lines are at 0.10" intervals); however, the models have become more inconsistent as the event nears. The 03 UTC forecast today was much drier than the first two runs, but then the current 09 UTC run becomes much wetter for the New Year's Day event. A nice illustration of two things here: QPF remains very difficult; and the models can have considerable inconsistency even at short time ranges. So, we will have to wait to see what the New Years brings.