When  the front and rain-band went South east Oz  late last week approximately (00 UTC 30 January or Thursday morning),  the southerly flow cleared the smoke cloud over Victoria for a day or so.  Since then, however, we have had another wide-spread smoke cloud over the state once again, with a spectacular red circle around the sun, persistent haze, visibilities occasionally down to a couple of kilometres and the EPA issuing haze alerts.

 I have gone through the MODIS site and grabbed all the images of South East Oz for the past week and have put them on my web-page along with each morning's mean sea level analysis.

Going through the sequence, on 30 january the frontal rainband is lying across eastern Victoria. The Modis image shows the large scale stratiform band with what looks like embedded Cu in the vicinity of the fires, as discussed extensively last week when we were all excitedly looking at the pyro Cu (Discussion of 30 January and 1 February)

Modis 30 January 0355 UTC
 

On 31 January there is a southerly flow across Victoria.  The MODIS image shows cellular Cu across Melbourne and Southern Victoria with no smoke cloud at all.  To the North-east the fires are still blazing away and a relatively small scale smoke cloud can be seen trailing eastwards from the fires.

Modis 31 january 0025 UTC
 

On February 1 st we have reverted to a light wind situation and the smoke cloud is back.  The flow is light southerly, however, so there is no smoke over Melbourne/southern Victoria.  Rather the cloud extends northward from the fires.

Modis 01 Feb 0340 UTC

2nd February a weak easterly dip has developed.  The winds are now northeasterly and a very impressive large scale smoke cloud lies across the eastern half of the state
Modis 02 February 0425 UTC

3rd February  I'm not sure what's happening:  Weak north-easterly  flow and a good smoke cloud over the state.  But there is also a very solid smoke plume out further to the east advecting thick smoke-haze over eastern Bass Strait.  Clearly there is a shear line to the east of the state that doesn't quite jump out at you from the large scale manual analysis.  By this date (Monday) we were getting used to smoke, haze and a red sun over Melbourne.
Modis 03 February  0055 UTC

4th February yesterday.   A very warm day with smoke haze over Melbourne all day. Good northerly flow over the western half of the state, but light winds over the eastern half so we still had a pretty good smoke haze.  I was out running that night (last night) along the Merri Creek Trail with the Collingwood Veterans... there was a specatular red haze around the sun while it was low in the sky. There are two Modis images for yesterday and you can now see quite a thick smoke plume across eastern Bass Strait down to north east tassie.
Modis 04 February 0000 UTC
                                                                                                                                 04 February   0410 UTC
 

This ongoing widespread smoke situation has no doubt had an inhibiting effect on daily maximum temperatures and no doubt has made the aviation shifts interesting.  It would be valuable for building corporate knowledge of these events if any of the duty or aviation forecasters during the period could take the time to make a few comments on matters they had to think about or on what they may have learned from the event.

Once the cold air behind the front had cleared near the start of the sequence, the air was fairly stable so there was no lightning or pyro Cu in the smoke-fires region.  I I have put a link to a loop of the GPATS superimposed on the IR for the last day or so on  my web-page; and you'll see there are no lightning strikes/convective activity anywhere near the smoke region.  Thus as discussed in the mail from Mike Fromm of the US navy last week the pyro Cu only seem to be set off when there is an external large scale destabilisation mechanism: viz on the day we had the pyro-Cu there was large scale uplift ahead of the front.  Interestingly, this didn't lead to wide-spread Cb's but rather just over the location of the fires.  As far as I can tell we don't have any pyro-Cu today ahead of the current front... any comments on this?

On a related issue these fires are largely/mainly a legacy of the lightning event of 7-8 January I believe.  Could someone check the facts on this?

 cheers

John McB

Paul Fox-Hughes

the smoke haze was very distinct in Hobart yesterday morning, and generated a lot of media interest.  Serendipitously, Agnes Apostolou from SBOB had sent over a number of nicely enhanced images from NOAA 12 and 17, and particularly a 22:30 UTC image from FY-1 showing the smoke streaming from NE Victoria over Bass Strait, and then Hobart.  Some of the smoke was probably locally generated, but the bulk was from across the water.

My recollection of the aviation discussion, while sitting across the way at the fire weather desk, was concern at the increased likelihood of low cloud formation in the presence of lots of  condensation nuclei about the north coast.

Rgds
Paul

John McBride
re: the smoke as a source of CCN, I was wondering the same thing a few days ago when there was a bank of low-level stratus up the east coast that looked like it had the fire/smoke region as a source.  You can see this on the MOdis image for 2 February that I already have on my web-page.  It looks even better on 3 Feb but it is off the image on the MODIS image; so I have put up a conventional DIFACS-collected vis image for that day.

John McB

 Milton Speer

"................on the day we had the pyro-Cu there was large scale uplift
ahead of the front.
Interestingly, this didn't lead to wide-spread Cb's  but rather just over
the location of the fires.  As far as I can tell we don't have any pyro-Cu
today ahead of the current front... any  comments on this?"

Much lighter prefrontal low level winds today over the fires than last week.
Implication being that in the lee of the northwesterlies last week where the
fires were (i.e.just on the southeastern side of the main range), could be
that local mesoscale circulations were induced in the valleys near the fires
which aided pyroCB formation along with the large scale lift and instability.
Fire behaviour described to me in a 400 metre deep valley just east of the
main range last week while I was down in Jindabyne near Perisher and Thredbo
would certainly support the existence of such locally induced circulations.

Also, perhaps a paper by J.T. Steiner 'Blowup Fires - The Byram Wind profile
'   AMM 24 (3) Sept. 1976  might be relevant. It describes various low level
wind profiles suggested by Byram that are conducive to blow up fires.

Milton

Tony Bannister
On a related issue these fires are largely/mainly a legacy of the lightning event of
7-8 January I believe.  Could someone check the facts on this?

Yes John,
All the fires apart from the ones near Beechworth and the ones started by the pyro-CB near Gelantipy were started on the night of Tues 7 Jan,  the next morning DSE were reporting about 100 going fires.  From hearing the cursing going on at the aviation desk, this event has been an aviation pain in the butt, trying to pick where the smoke will get to, how bad it will be, is it up high and will mix down during the day or is it down low and mix out during the day.  The very first smoke event in Melb I think was a combination of the NE fires and also the leftover from a fire the evening before near the Kilmore Gap. It was
interesting the next time smoke came down over Melb it started out at a much higher level and then mixed down during the day.

Mike Fromm

Hello John,

Thanks you for the great email and posting!  Wonderful stuff.  I'd like to offer some information
that might correlate with your chronology of smoke, synoptics, and convection.  You can be
the judge as to whether to post it to your discussion group.

The thing that usually alerts me to these extreme smoke events is the TOMS
aerosol index data.  NASA's web site with these maps is usually my first "stop" every
morning at work. When I see the TOMS AI blobs go off their color scale, then persist 1000s
of km away from the source, I figure something really extreme has occurred.  But the TOMS data
on the www are processed and filtered in such a way that you can't see the true details of
the extreme events. So I get their individual pixel data and make my own plots, showing the
full range of AI values. These maps are perhaps not even totally sanctioned because they include
data that the TOMS retrieval algorithm flags as errors.

What I have found, and shown to the NASA TOMS folks, is that the extreme AI values are certainly (albeit qualitatively)
believable.  They signify aerosol abundances that most likely relegate the aerosol "cloud" altitude to levels unexpected in
normal conditions, i.e. very high. The persistence of these clouds confirms (in my mind) that the aerosols are
above scavenging altitudes. It is these intense, persisting AI plums that alerted me to your area.  Jan
18 was the first blowup, but your chronology has identified a few other potential ones.  I'm attaching a
few to show you what TOMS saw after the first blowup.  (toms-ai_030117_aus.gif, toms_ai_030118_aus.gif, toms_ai_030119_aus.gif)   I hope by tomorrow to have data through Feb 4 and will be able to send a complete daily chronology.

The maps show AI in the range from barely detectable (about 0.7) up to the  max value possible (12.8).
Certain pixels have a little black dot in the center.  These pixels were assigned an error flag by the TOMS
algorithm.  At these pixels, the algorithm could not retrieve  ozone.  Interestingly, it is the presence of these
error flag values that imply to me an event worth looking at!

You'll notice a nice agreement between the TOMS maps and MODIS.  The TOMS  data near Australia are collected
at approximately 1 UTC.

I'll be back in touch with more unless you tell me otherwise. Thanks again for your valuable insights and info.

mikef

John McBride

Mike,

 Thanks for the interest.  I expect our Australian mets will be quite interested in your analysis so I shall certainly put your note out to my/our discussion list.

 For me (and I suspect many of our operational mets) our experience and knowledge of cloud physics and in particular of aeroseol science is very, very rusty.  Given that, could you expand on your sentence, "The persistence of these clouds confirms (in my mind) that the aerosols are above scavenging altitudes."

Regards

John McBride

 Mike Fromm

Hi John,

>very rusty.  Given that, could you expand on your sentence, "The
>persistence of these clouds confirms (in my mind) that the aerosols are
>above scavenging altitudes."

The "clouds" I am referring to are the blobs of enhanced TOMS aerosol index. By "persisting" I mean that these AI clouds were trackable from Australia on Jan 18 to the S. Atlantic on Jan 26.  Usually an enhanced area of TOMS
AI disperses to the point of disappearing a couple of days downwind of its source place and time.  One process that certainly effects the persistence of an aerosol cloud is scavenging by precipitation.  Another might be nucleation into cloud droplets (to the extent that smoke aerosols might act as CCN).  The potential for smoke aerosols to act as CCN is, I think,
still an unknown.  But it IS fairly well known that precipitation and scavenging by precip is pretty effective at removing aerosols so that TOMS can't detect them.

So, the evidence that I saw that so impressed me was:
1. Very high values of AI
2. AI blobs that held together for over a week and 1/2 a world away
   from Aus.
Both qualities are consistent with an aerosol cloud at very high altitude.  I don't think for a moment that all the smoke aerosols congregated in a single, narrow, very high altitude range...they were probably distributed in some messy way throughout the troposphere. But the topmost part of the plume had to be pretty high for the two conditions above to be satisfied.

There's yet another piece of the puzzle supporting the high-altitude claim.  My back-of-the-envelope calculation of the average speed of advection of the TOMS cloud between Jan 18 and 26 showed that the blobs moved at roughly 28 m/s.  That's a pretty stiff breeze.  My quick look at gridded analyses indicated that those winds were pretty much
relegated to altitudes above 400 mb.

That's probably much more than you bargained for:)

Thanks again,
mikef