Tuesday, July 31, 2018

20180731

Previous Forecast
It was a difficult day for forecasting.  Phoenix got walloped during the early evening as severe storms moved off of the high country into the valley.  The 0Z Phoenix sounding showed that the PBL was mixed to almost 700mb with a weak cap and 1300 J/kg of MLCAPE. Storms were over the higher terrain to the north all the way around to the southeast.  Multiple outflows were observed on RADAR entering the valley thus providing the local convergence and lifting necessary to initiate scattered deep convection. Not that much rain fell in Phoenix proper as the strongest storms remained just north of the city.  However, as the storms collapsed, severe winds were generated with a G74 at Glendale and G67 at Luke.


The 12Z model runs were unable to forecast this outbreak, but they were close.  The WRFRR was the best as it had scattered strong storms just north and east of Phoenix but coverage wasn’t enough and they weren’t able to propagate into the valley.  The forecast vertical profile was excellent though. So close!

While none of the 12Z WRF nor the 12Z HRRR moved activity into Phoenix, it was a different story with the 15Z HRRR and WRFRR.  The 15Z HRRR moved scattered storms through the Phoenix area accompanied by very strong winds.

The 15Z WRF had a broken line of strong storms that moved through Phoenix around 6pm.  However, this was not correct as most activity died as it moved through. The 15Z WRFRR had too much moisture and heating and ended up with around 3000 J/kg of CAPE.  So what actually transpired was more than the 12Z and less than the 15Z. The 15Z HRRR was the better run.


Day 1
Drying has occurred over much of the eastern ½ of the state but very wet air continues over southwestern and far western Arizona.

CAPE is quite low for most areas except for the high country and the far SW part of the state.  It looks like only the high terrain will see any activity today. The one exception is that little finger of CAPE over central Pima County.  Does this feature really exist? The WRFGFS doesn't have it while the WRFNAM is similar looking but has lower values.

The problem is that the 15Z WRFRR came along and made the forecast complicated.  It develops late afternoon and early evening activity farther north into eastern Pima County.  I took a close look at forecast vs actual PW as of 17Z and the 15Z WRFRR is very close to observed PW so I can’t discount it for being too wet.  What I thought was going to be a quiet day may not be so quiet for areas around/in Tucson later today and this evening.

Day 2
A similar story tomorrow as dry air continues over NE Arizona and wet over SW Arizona.


The 500mb anticyclone center is over northern Arizona which places southern Arizona in weak, generally easterly flow.  So, some storms could be steered towards the lower elevations of Pima and Pinal Counties.

The atmosphere recovers and CAPE is much higher over much of the southern ½ of the state thus the potential for more activity is present.


Tucson is mixed fairly deeply by late afternoon and CAPE is around 1600 J/kg so plenty to support deep convection.   Steering is weak but does have an easterly component so some storms may move towards the lower elevations. Anvil shading from storms to the north may also be an issue.

Phoenix is again capped with only a moderately deep mixed layer thus storms will struggle to get into the lower elevations there.


The runs vary quite a bit with the success of storms making it into the lower elevations of SE Arizona.   The WRFNAM is quite active while the other two runs keep most storms over the higher terrain.

The 15Z WRFRR is now available (as I started the discussion so late) and it’s quite active, similar to the WRFRR so it’s hard to say what will actually happen tomorrow.


Monday, July 30, 2018

20180730

Previous Forecast
Activity remains over the higher terrain of the state.  Southern Pima into Cochise and eastern Arizona were especially active.
The 12Z WRFRR was one of the better runs but didn’t have any activity out in western Pima county. None of the runs had any activity in either Phoenix or Tucson which verified well.  

The 12Z HRRR did a pretty good job for the most part, similar to the WRF except for later in the evening when it moved storms into the lower elevations of eastern Maricopa and Pinal counties.

Initializations
The 500mb high center is located over Nevada and a weak inverted trough is somewhere over southern Arizona into northern Sonora and weak cyclonic motion can be seen over southern Arizona in the visible imagery.  A MCV is moving out over the central GofC. The RR has about everything initialized. The NAM is missing the MCV but that is no factor for Arizona today.

Both the RR and NAM did an OK job initializing the scattered morning clouds across the state but the RR was too far south for the thicker clouds that are west of Phoenix.  The NAM had some large, 4-8mm, wet bias errors over the SW and into NW Mexico (below). The RR has mainly small errors except for far NW Mexico where there was a bit of a wet bias.  The WRFRR is the favored model today.



PW is about the same as yesterday with 42-50mm in the lower elevations and very wet air continues to be transported into the state as flow continues from the surface to above 850mb from the GofC.   There is quite a moisture gradient across the state as dry air has moved into NE Arizona and NW NM.

CAPE is again high to very high over much of central and SE Arizona so any storms that are able to form will be very strong and produce heavy rain.  This backdoor dryline that is over NE Arizona and western NM may act as a convergence line and help initiate and enhance storms later today.


By mid-afternoon, scattered strong storms are located over the higher terrain of the state.

The plotting routine went a little crazy with the streamlines, but eastern Arizona into SW NM appears to be in difluent/divergent flow aloft.

Some storms move off the high terrain towards the Phoenix area while Cochise county again experiences very strong convection. This may become a MCS as there is some upper support.

The vertical profiles look very similar for both Tucson and Phoenix as CAPE is around 1500 to 2000 J/kg along with a good wind profile.  However, both are not mixed deeply and have significant inversions/caps on top of the mixed layer. So it’s the same old story of needing strong lifting from outflow boundaries in order to trigger deep convection for the lower elevations.

Big storms are just north of the Phoenix valley during the early evening and result in multiple outflows into the valley only from the north and northeast  It appears that linear outflows do not supply enough mesoscale lift for widespread deep convection but may be enough for a few storms. If a big outflow boundary comes in from the south or southeast and they intersect, it may be more active but none of the model runs are forecasting this situation.

The storm complex in Cochise county moves to the west and threatening the Tucson area.


The WRFNAM plays it a bit different as it moves a strong line of storms towards Pima county from the NE.

The WRFRR moves a strong outflow boundary into the Tucson area later in the evening and this appears sufficient to trigger at least a few storms but not a widespread outbreak.

The WRFNAM moves widespread storms into the Tucson area around midnight and typically, these “rim shot” storms come in earlier in the evening so the timing may be a bit off. The WRFRR is probably more accurate but the WRFNAM can’t be ruled out.  I’ll be watching to see how active Cochise county is later this afternoon and evening as the WRFNAM kept it pretty quiet which enabled the Rim storms to propagate into an unmodified air-mass.

The 6Z WRFGFS looks much like the 12Z WRFNAM with Rim storms moving into Tucson during the evening.  The 6Z WRFNAM looks similar to the 12Z WRFRR with less activity. So, my take is a 50-50 shot at big night time storms for Tucson.

I waited around to see what the 12Z WRFGFS has had to say and it is quite interesting as it develops a broken line of strong to severe storms from NE of Phoenix all the way into far SW NM.  

This line moves into the Tucson area during the late night hours.  It looks likely that SE Arizona is going to experience a big outbreak tonight.


Day 2
Drier air moves into much of the eastern ½ of the state.  I’m thinking that all the above storm action was aided by not just the upper divergence but this westward moving dryline.  In any case, with such an active day forecast for today plus drying, little or no activity is expected for most of tomorrow.

Storms are mainly active from Flagstaff and northwest from there.


Sunday, July 29, 2018

20180729

Previous Forecast
The model runs struggled yesterday as most of them had storms developing in the Phoenix valley but no storms materialized.  It was close, as the 0Z Phoenix sounding was mixed deeply and it appeared that any outflow boundary would trigger convection.  Areas just to the north and west of Tucson had a lot of activity.

The best run was the 15Z WRFRR which turned out to be very accurate.  It kept storms out of the Phoenix area and was able to forecast the early morning activity across central Arizona.

Of interest was the comparison to the 15Z HRRR which uses mostly the same RR initialization.  It moved a huge squall line through Phoenix.

The WRFRR kept storms to the north and west of Phoenix but a little too intense.  It was also able to correctly forecast the scattered deep convection NW of Tucson.


Initializations
The center of the 500mb anticyclone has moved north into Nevada in response to a digging trough over the midwest.  A broad inverted trough is located over SW Arizona and into NW Mexico and is responsible for the early morning activity and widespread clouds across much of Arizona.  Visible imagery indicates an MCV over far SW Arizona which was not resolved in the upper air data. Both the RR and NAM initialized the IT okay and the RR even had a bit of a closed circulation at 600 in SW Arizona.  


High and mid clouds cover much of southern Arizona along with some continuing showers in the vicinity of the MCV.  The RR was able to initialize both the clouds and showers well but the NAM didn’t have enough of either. The NAM also had a 4-6mm wet bias over much of eastern Arizona while the RR initialized well.  A quick look at the 6Z WRFGFS shows that it also forecast the morning clouds and showers well but did have some moderate to large PW errors over central Arizona. The WRFRR has initialized well and will be the model of choice.  

Day 1
PW is very high across southern Arizona (43-50mm) in the wake of the overnight activity.  It continues this way during the day with maybe even an increase over western Arizona. Moist air has pushed eastward into SW NM so maybe far SE Arizona will have some decent rains today/tonight.

There is an impressive area of 50mm+ PW all the way down the Gulf of California and even very wet air well out into the Pacific.

CAPE is very high over SE Arizona and NW Arizona thus those areas are likely to be the most active if it gets warm enough but CAPE is also quite high over the lower deserts.  The Tucson 12Z sounding had quite a bit of MLCAPE, around 1100 J/kg, and wasn’t that worked over, so just going by that SE Arizona should be quite active today.

The WRFRR closes off the far SW AZ IT and moves it to around Yuma while another weak IT is over SE Arizona.  500mb temperatures are around -7 C to a warm -4 C.

Storms take their time getting going today and by mid-afternoon, there are a few scattered about over the high terrain and far southern Arizona.

By early evening, Tucson is mixed deeply and near the LFC.  PW is very high at around 44mm and CAPE is around 1000-1400 J/kg so storms that form will produce very heavy rain.  The vertical wind profile couldn't be much better: good low-level shear, 15-knot mid-level steering, and light westerly aloft to push anvils away from the storm movement.


The 6z WRFGFS is also fairly quiet as only a few storms are active over the higher terrain around the Tucson area late this afternoon.  This is a head scratcher as the Tucson vertical profile looks so good. I guess the model’s inversion at the top of the PBL is too strong to overcome at lower elevations as the very wet and slightly cooler air moves in from the west.

It’s quite obvious this is the case at Phoenix as the PBL is not mixed very deeply plus there is a pronounced cap just above 850mb.  Other than that, it looks great!

This evening, the WRFRR develops some storms over SE Arizona and moves them toward Tucson area but they die out before arriving. Perhaps the outflows will be enough to kick off some activity?

For what it’s worth, the 6z WRFGFS does develop a few storms to the east of Phoenix this evening.  So, again, there is quite a bit of spread and no model consensus thus confidence is low.

Day 2
I’m reminded of this song when I’ve started to write this section the past few days.  Just replace “sun” with “storms”. Day 2 forecasts have been too active as I think they are not handling the low-level cooling in the 24+ hour time frame but that is just a guess. Maybe upper air data from Mexico would help this situation?

Moisture remains very high over southern Arizona while dry air has moved into much of northeastern part of the state.

Once again, CAPE is moderate to very high over much of the southern ½ of the state.  A very pronounced boundary exists between the two air-masses and sometimes these can act as convergence zones and help trigger deep convection.

Tucson becomes well mixed by late afternoon with around high CAPE and a very good wind profile thus it looks like a very good situation for big storms.  Both the WRFNAM and RR are very similar.

Both model runs develop strong/severe storms over SE Arizona during the late afternoon and move them into the Tucson area during the early evening.  In spite of both runs being very active, I’m skeptical.

The WRFRR develops storms in Pinal County on the edge of the strong outflow but Phoenix only gets a big dust storm.