Arizona Smoke Inhibits Convection? 6/8/11 & 6/9/11

[Peter Wharton]

Wondering if the smoke from the Arizona fires was the reason for the inhibition of convection on the two last days in what were two areas with all the parameters right for severe weather and tornados - SW of Des Moines, Iowa on 6/8/11 and SE Nebraska on 6/9/11 (10% tornado risk)?

On the satellite loop for 6/8/11 it seemed that as the smoke plume from Arizona advanced into the area SW of Des Moines, existing convection simply went flat. Again yesterday as the plume advanced into SE Nebraska, what looked like initial explosive towers suddenly collapsed with no further convection until forcing right along the frontal boundary.

3 possible inpacts of the particulate matter in the smoke could be hypothesised:

1) The obvious is the screening of the sun's rays reducing insolation, especially from late afternoon as the sun angle is through an increasing layer of smoke
2) The particulates themselves absorbing the heat of the sun rather than simply scattering it, thus adding to the ambient temperature and adding to the capping inversion in place
3) The wealth of particulates somehow interfering with condensation processes in the clouds and reducing heat being released from condensation - perhaps by water molecules attaching themselves to the particulates.

Could also be a more prosaic meteorological reason for the lack of activity in the two areas noted, but all the parameters seemed in place, and the arrival of the smoke plume seemed to coincide with the suppression of convection.

Does anyone here who is better trained in the physics of this have a view on this? Have found some articles indicating speculatively some effects from smoke, but what do you think?

 

[Dima Smirnov]

This is an interesting observation and I was thinking the same thing to myself as I was doing base support for a van chasing in the area. Of the three impacts you suggested, I think that (1) is probably a small effect, at least based on surface observations in the area. If the dust did have an effect on squashing moist convection, I think it was through your reason (3) and also due to a large increase in the number of cloud condensation nuclei, which would wreak havoc on convective clouds. If we can figure out what level the dust was located (I'm currently talking to a radiation expert regarding that), we should be able to more definitely say if it was a thermodynamic effect or a cloud physics effect.

 

[Aaron Kennedy]

We can estimate some impacts. The ARM SGP site in Lamont, OK has instruments to measure the impacts of aerosols (i.e. smoke from the wildfire) on the radiation budget. Here's a plot when the plume passed over the site on the 7th:

http://aeronet.gsfc.nasa.gov/cgi-bi...f_day=0&if_err=0&place_code=0&year_or_month=0

Plot on the right is the day, plot on the left is the time-series for June for aerosol optical depth. The variety of lines represent the AOD at different wavelengths (1640nm (near IR) to 340nm (UV)).

So what is optical depth? Basically a measure of transmittance of radiation. transmittance= exp(-AOD) As optical depth increases, transmittance decreases. In the case of the smoke plume towards the end of the 7th, AOD reaches > 2
Plug in 2, and you get a transmittance of 13% In other words, 87% of the radiation does NOT reach the surface. This radiation either is absorbed or scattered by the aerosols. Other than the height being a bit lower, this is very similar to cirrus moving over a target area and limiting heating. I've been pretty busy so I can't comment on this specific convective setup, but based on the plots at the ARM site, I would say it could have been a factor for convection not developing if this plume passed over areas during peak heating.

Satellite views of the plume:
http://cimss.ssec.wisc.edu/goes/blog/archives/8299

 

[Keith Wilson]

So basically it could have actually had somewhat of a "capping" effect on the convection?

 

[Zach Young]

1) The obvious is the screening of the sun's rays reducing insolation, especially from late afternoon as the sun angle is through an increasing layer of smoke

I would add the effect of the morning in there too. The day gets off to a slow start, and starts to lose effective insolation early in the afternoon. The effect is probably much smaller during the middle of the day, but with especially this smoke plumes I'm sure it's measureable. It should also be noted that the effect doesn't necessarily need to be that large to disrupt convection. A difference of 3-5 degrees could be enough to prevent parcels from breaking the cap, especially in scenarios with weak forcing.

 

[Aaron Kennedy]

So basically it could have actually had somewhat of a "capping" effect on the convection?

Taking away the heating makes the cap harder to break.

The effect is probably much smaller during the middle of the day

Solar insolation is greatest at noon, and then diminishes. The plume was substantial enough that I don't think looking through it at an angle matters too much (makes the optical depth larger). Given that heating is greatest from mid morning-early afternoon, this would be the best time to disrupt the warming of the boundary layer.

 

[Patrick Marsh]

Smoke and other aerosols tend to get trapped just below inversions, which act to prevent their ascent and dispersion.

Smoke will also act to warm the layer it is in and the layer immediately above it. This is because the smoke will absorb some of the incoming radiation and scatter/reflect some of it upwards. This means that the level/layer immediately above the smoke plume will be warming from above and below. Thus, you can get local enhancements of inversions which trapped the smoke in the first place. There was a sounding at Topeka sometime last week that captured this fairly well (comparing the 18/19 UTC sounding to the subsequent 00 UTC sounding).

 

[Peter Wharton]

Think I observed another instance of smoke inhibiting convection on 30th May - severe storms & tornadoes broke out largely outside of the moderate risk box posted by the SPC for the day - as I drove out to NW Oklahoma for what should have been a great chasing day, I noticed a fairly extensive plume of smoke in the mid to upper levels of the atmosphere - from the New Mexico fires that had broken out? - I said to my fellow chasers that day "I hate it when smoke is present" - sure enough convection, what there was of it, was starved/capped/skinny and didn't develop into severe weather until the derecho event later that evening, forced by the outflow from earlier severe storms in Kansas.

I know there were other factors that may have inhibited convection that day, yet all the parameters suggested storms should fire - and they did - to the south along the dryline and up in Kansas along the cold front - but not in NW Oklahoma where the smoke was visible. Perhaps it's a "chicken & egg" situation - does the smoke inhibit convection when parameters are "on the edge" or did capping &/or convergence in upper level winds that day lead to smoke being concentrated in that area?

Twas a weird day - should have been severe in NW Oklahoma with tornadoes, but something got in the way - the SPC moderate risk box seemed to be the area for nothing to happen.

 

[Warren Faidley]

The moisture quality / depth has also been a major factor this season, contributing to crazy LCL's and a lot of inverted V soundings. Even the tornadoes on April 14th, were amazingly high based. (See my pics on sc.com). It also seems there has been a lack of stronger shortwaves this season (and/or poor timing), with a lot of forcing depending on surface based lift. I don't recall the year(s), but smoke from massive fires in Mexico made chasing almost impossible before we had mobile radar as towers / storms were impossible to see.

W.

 

[Jeff Duda]

Think I observed another instance of smoke inhibiting convection on 30th May - severe storms & tornadoes broke out largely outside of the moderate risk box posted by the SPC for the day - as I drove out to NW Oklahoma for what should have been a great chasing day, I noticed a fairly extensive plume of smoke in the mid to upper levels of the atmosphere - from the New Mexico fires that had broken out? - I said to my fellow chasers that day "I hate it when smoke is present" - sure enough convection, what there was of it, was starved/capped/skinny and didn't develop into severe weather until the derecho event later that evening, forced by the outflow from earlier severe storms in Kansas.

I know there were other factors that may have inhibited convection that day, yet all the parameters suggested storms should fire - and they did - to the south along the dryline and up in Kansas along the cold front - but not in NW Oklahoma where the smoke was visible. Perhaps it's a "chicken & egg" situation - does the smoke inhibit convection when parameters are "on the edge" or did capping &/or convergence in upper level winds that day lead to smoke being concentrated in that area?

Twas a weird day - should have been severe in NW Oklahoma with tornadoes, but something got in the way - the SPC moderate risk box seemed to be the area for nothing to happen.

I think the causes for the lack of initiation in NW OK on Wednesday were more strongly tied to poor moisture quality and being on the wrong side of the OFB left behind by the morning storms. Also the surface warm front never really formed, as the upper-level support lagged behind (hence all the activity in KS/NE), whereas the subtropical jet was pretty strong in TX, thus leaving OK mostly in between areas of forcing. It would be hard to conclude that smoke aerosols from fires had anything to do with it.