View toward Redington Pass at 5:25 am this morning.
Plot of detected CG flashes for 24-hours ending 0733 UTC (above, from Atmo and Vaisala) shows a fairly large hole covering most of metro area and extending south into Santa Cruz County. ALERT observations for 24-hours ending 7:00 am this morning (below) shows rainfall mostly avoided the metro area. Yesterday's WRF forecast did not verify well, nor did the NWS forecast POPs and Flash Flood Watch (see previous post). There was rumbling thunder here in 4 to 5 pm time-frame, but only a brief sprinkle.
At 500 mb this morning (above) the anticyclone is centered over Oklahoma, with much of Arizona under the influence of the western US trough. The morning sounding from TWC/TUS (below) appears considerably more favorable for deep convection than did yesterday's - see previous post. Precipitable water and CAPE are both high, and winds southerly to southwesterly.
The GEFS plumes for TUS (from the 06 UTC runs - above) indicate chances for light showers through the Fourth, but overall forecast a fairly quiet week. The 06 UTC GFS, however, forecasts an active second week in July (below) with heavy rainfall across southeastern Arizona.
Not much June rain(0.17 inches) at my foothills home(near Campbell & Sunrise).
ReplyDeleteHello Craig - you're not very far way from here, so I guess I was fortunate to catch as much rain as I did! Bob
DeleteBob, a comment on forecast accuracy. You say "Yesterday's WRF forecast did not verify well". It depends on what forecast you look at. The various incarnations of the WRF-HRRR (12 and 15Z) did well, as precipitation was mainly over the higher terrain. Unfortunately, since I left, something is wrong with the 9Z run. The 6Z RR and GFS WRF runs are almost always too active.
ReplyDeleteHi Mike - this highlights one of the dilemmas I face in using WRF forecasts. There is such a wide variety of parameters, run times, and model variants that it is impossible for me to examine all, if I want to get a blog posted before 8 or 9 am. This morning I wanted to go out to the Fourth and the 12 UTC WRF-RR hadn't run out that far yet, so I defaulted to the 06 UTC WRF-GFS. When I discuss verification, I only deal with what
ReplyDeleteI showed on the previous blog post. June 30th was definitely a problem day - especially for the NWS forecasts. My main concern now that you're retired, is that the long-term fate of the WRF model and runs is an unknown! Bob
I had the same issues when I was forecasting for TEP and APS. Too many model runs to sort through. However, it was clear to me that the WRF-HRRR outperformed the others so I'd focus on it. I compared about 6 months of forecasts between the WRF-GFS and WRF-HRRR in ID a few years ago, and the study showed a small, but significant improvement in accuracy (T and precip) using the WRF-HRRR. My guess is that the stats would be even better during the Arizona monsoon, but now that I'm retired, we'll never know for sure.
DeleteHard to say what will happen with the runs. Now that Chris has left, there is no real champion at HAS. Runs must continue though, as they are critical for the utility renewable energy forecasts made by the Power Forecast Group.