Michael Foley

Hi All,

last night there was a long lived squall line which developed out of storms west of Borroloola yesterday evening (around 0830 UTC) and then reformed as  a more coherent feature just west of Katherine at 1430 UTC (local midnight), from where it headed out across the Top End and into the Timor Sea (41 kt gust at Port Keats (1900 UTC) and 37 kt gust at Troughton Island (0100 UTC)), leaving a big blotch on this morning's satellite pictures.  It looks like it is still going, and is nearing Browse Island off NW Australia about now.  The line was passing over pretty remote country, but I managed to talk to a few people along its path, who remarked on the winds as the gust front came by.  At Dorisvale Station some large old trees were knocked over, and the observer there said there was a "dust storm" before the line passed (it had been very dry there until last night).

Some interesting questions:
1. See the attached radar picture of the line just as it reformed.  It is remarkably thin and linear even for precipitation in a squall line.  Roger  Atkinson was speculating that what was visible here could be the dust being lifted by the gust front which then also led to development of more normal looking convective cells as it propagated west.  Is this feasible?  (It would certainly match with the observation from Dorisvale.)  Or are we seeing a true line of almost-non-cellular convection as the squall line takes off.

2. My understanding of squall line motion is that they propagate perpendicularly to the line, rather than in the direction of the midlevel flow i.e. like so:  | | | | | |.  However the propagation of this line seemed more to be in the direction of the mid-level winds (which was E to ESE at 30 to 35 knots) i.e. like  / / / / / / .  Does this fit with people's ideas of squall line behaviour?  Is there some preferential development on the left hand side of the line occurring?

Bye,

Michael.

Harald Richter
Michael,

> 1. See the attached radar picture of the line just as it reformed.  It is
> remarkably thin and linear even for precipitation in a squall line.  Roger
> Atkinson was speculating that what was visible here could be the dust being
> lifted by the gust front which then also led to development of more normal
> looking convective cells as it propagated west.  Is this feasible?  (It
> would certainly match with the observation from Dorisvale.)  Or are we
> seeing a true line of almost-non-cellular convection as the squall line
> takes off.

The line doesn't look too unusual to me (looks like a young string of convective bubbles).  It is close to the radar leading
to the appearance of finer features in the reflectivity.  The beam might also undershoot the trailing stratiform shield (ifv present) given the line proximity to the radar site.  Were there any rain gauges in the line's path? Could this be dust?  Dunno - the beam 50 km out at .5 deg. is at 600m or so, does dust give off 40dBZ (yellow) at that height??

> 2. My understanding of squall line motion is that they propagate
> perpendicularly to the line, rather than in the direction of the midlevel
> flow i.e. like so:  | | | | | |.  However the propagation of this line
> seemed more to be in the direction of the mid-level winds (which was E to
> ESE at 30 to 35 knots) i.e. like  / / / / / / .  Does this fit with
> people's ideas of squall line behaviour?  Is there some preferential
> development on the left hand side of the line occurring?

I am not quite sure whether you talk about squall line motion relative to some ambient shear vector, or cell motion within a squall line? Were there any appreciable gradients in the BL moisture field?
Harald

Michael Foley

 "The line doesn't look too unusual to me (looks like a young string of convective bubbles).  It is close to the radar leading
      to the appearance of finer features in the reflectivity.  The beam might  also  undershoot the trailing stratiform shield (ifv present) given the line
      proximity  to the radar site.  Were there any rain gauges in the line's path?  Could this be dust?  Dunno - the beam 50 km out at .5 deg. is at 600m or
      so,  does dust give off 40dBZ (yellow) at that height??"
 

The stratiform trailing precipitation hadn't developed at that stage (the line was just getting going), but was certainly evident on radar later on.  I've no idea about how dust would show up on reflectivity radar--Roger A. might be able to comment with regard to other radar observations he has made around Tindal.

Rainfall: 24 mL at Dorisvale but the observer there had been much heavier falls around the place, filling up the waterholes on his station;10 mL at Peppinmenarti and 12.4 mL at Port Keats with the line--another 3 mL with the trailing precipitation over the next couple of hours.

' I am not quite sure whether you talk about squall line motion relative
      to some ambient shear vector, or cell motion within a squall line?"
I mean squall line motion relative to the ambient shear vector: the line was oriented SW-NE, the ambient shear vector would have been roughly easterly, and the line propagated to the west.

      "Were there any appreciable gradients in the BL moisture field?"
No--surface dewpoints around 20-21 across the inland region--moistening up to 24 around the west coast, though.
 

Jon Gill
Hi Michael,

(putting my musty tropical hat on)

movement of tropical squall lines in the direction of the mean middle level flow is pretty typical. In fact, it is intimately tied
to the dynamics of how these these things work. As the initial cells develop and mature, middle-level momentum is entrained by the convective scale downdrafts and transported into the boundary layer cold pool outflow. New cell development occurs on the leading edge of the cold pool, where localised convergence is enhanced. The system evolves into a meridionally-oriented line that moves in the direction of the middle-level flow. In other words, the squall line propagates by
preferential development on the cold pool outflow, in the direction dictated by the middle-level momentum that has been entrained into the cold pool.

There are some useful papers around that show the airflow structure of tropical squall lines in vertical perpendicular
cross-section, and the rear/midlevel to front/lowlevel flow-through is a feature. This flow acts to overturn the high thetae air
ahead of the line, replacing it with low thetae behind, like a giant mixer.

In the tropics, there is high theta-e everywhere ahead of the line, so the line can grow very long. By contrast, in the
midlatitudes, the high theta-e air tends to be equatorward, so the individual cells develop preferentially on the equatorward end of the line, and the line as a whole seems to deviate from the middle level flow.

To get the long-lived lines, there needs to be a careful balance between the strength of the rear-inflow and the strength/density of the cold pool outflow, i.e. the right balance between shear and buoyancy.

Something like that.

Cheers...Jon

--
Jon Gill
Acting Regional Director (Victoria)
Bureau of Meteorology

Roger Atkinson
My suggestion was not that dust caused the reflectivity returns early in the piece, but that flying leaves and sticks had (I've seen this before up here), but as I've since discussed with Ian Shepherd, I no longer think this was a case of that. (The
squall - I crudely estimate about 30odd kt at our place - came through Katherine about half an hour before the line echo developed. This made me suspect that the line echo might be due to flying foliage (it'd take about 60kt to do that). Since
then I was able to check the timing of everything, and the rainfall figures nearby to the west, and it all suggests the radar echoes were most probably rainfall echoes.) The remarkable features of this line were the speed with which it blew up,
and that early in the piece (i.e. at the stage the line of echoes had just developed and was still strong and thin - >50mm/hr everywhere and >100mm/hour in small cells along the line) the lightning display was spectacular. I've seen lots of vigorous
electrical storms up here but this was the grandaddy of them all. The lightning was continuous for more than 60 minutes (i.e. no trouble reading the shorter oxford in the back yard - something I think we all do regularly up here to relieve the monotony of a tediously dry buildup. No?)

> I mean squall line motion relative to the ambient shear vector: the line
>   was oriented SW-NE, the ambient shear vector would have been roughly
>    easterly, and the line propagated to the west.
>
>          Were there any appreciable gradients in the BL moisture
>           field?
 

>     No--surface dewpoints around 20-21 across the inland region--moistening
>     up to 24 around the west coast, though.
 

Not so, I think. Mid-teens around here at the time of the development, and probably
23-24 on the coast.
Hypothesis: there may in fact have been a strong horizontal gradient of low level moisture to drive the development. The line of convection (much less organised) had been evident earlier in the day as it tracked across the SE Top End from an origin
to the SW of Borroloola (check out the satpix and radar data), where low level moisture was quite high. As it approached Katherine from the ESE it appeared (radar) to die out, most likely due to the dryness at low levels (dewpoints 11-14 degrees).
Then the dry squall came through Katherine. Then the convection blew up about half an hour later, about 25nm to the west of Katherine. Now, at this time of year the seabreeze often arrives at Katherine from the Bonaparte Gulf late in the evening
(the "Point Keats undertaker"? It must be about the only seabreeze in the country that is hated: just as you're hitting the sack the temperature increases from the mid-20s to 30-31 and the dewpoint goes from mid teens to low 20s). OK, I reckon that, about 25nm west of Katherine, the squall from the east met the seabreeze coming in from the west. If so, I'd guess at a horizontal moisture gradient in the B/L of up to 10 degrees in 5nm, causing the explosive development of the line
disturbance. Seem to make sense?

Harald Richter
El Rogelio Atkinson opined:

> >No--surface dewpoints around 20-21 across the inland region--moistening up
> >to 24 around the west coast, though.
>
> Not so, I think. Mid-teens around here at the time of the development, and
> probably 23-24 on the coast.

> Hypothesis: there may in fact have been a strong horizontal gradient of low
> level moisture to drive the development. The line of convection (much less
> organised) had been evident earlier in the day as it tracked across the SE
> Top End from an origin to the SW of Borroloola (check out the satpix and
> radar data), where low level moisture was quite high. As it approached
> Katherine from the ESE it appeared (radar) to die out, most likely due to
> the dryness at low levels (dewpoints 11-14 degrees).

I am thinking top end in November, I am also thinking daytime T/Td = 30++/11-14,  I am thinking very high-based convection that is "no good" other than for potential gust front production ...

> Then the dry squall came through Katherine.

I believe that is the gust front from the presumably high-based convection?

> Then the convection blew up about half an hour
> later, about 25nm to the west of Katherine. Now, at this time of year the
> seabreeze often arrives at Katherine from the Bonaparte Gulf late in the
> evening (the "Point Keats undertaker"? It must be about the only seabreeze
> in the country that is hated: just as you're hitting the sack the
> temperature increases from the mid-20s to 30-31 and the dewpoint goes from
> mid teens to low 20s). OK, I reckon that, about 25nm west of Katherine, the
> squall from the east met the seabreeze coming in from the west. If so, I'd
> guess at a horizontal moisture gradient in the B/L of up to 10 degrees in
> 5nm, causing the explosive development of the line disturbance. Seem to
> make sense?

Lotsa sense.  A gust front plowing into a pre-heated Td~20s airmass in the presence of "buildup-season" CAPE must be a good thing.  Could a careful  surface analysis reveal that initiation took place precisely along the aged sea breeze front at the time the gust front made first contact?  This event sounds like a good case study for tropical squalllinogenesis.

Harald in Melbourne, Td~zilch, CAPE~zilch

Roger Atkinson

> > Then the dry squall came through Katherine.
>
>I believe that is the gust front from the presumably high-based convection?

Yes, I think so, but from the old activity rather than the renewed event, which didn't fire up until west of here.
(It's even possible, though I'm not sure how plausible, that the 'gust front' which came through Katherine that evening was what was left of the  seabreeze front from the SW Gulf of Carpentaria. Ordinarily I'd suggest it was of convective origin, but in that case the convection responsible died out about 70nm away to the ESE a few hours before the front reached
Katherine - it seems to me rather a long way for the front to propagate without being continuously fed by more active convection.)

> > Then the convection blew up about half an hour
> > later, about 25nm to the west of Katherine. Now, at this time of year the
> > seabreeze often arrives at Katherine from the Bonaparte Gulf late in the
> > evening (the "Point Keats undertaker"? It must be about the only seabreeze
> > in the country that is hated: just as you're hitting the sack the
> > temperature increases from the mid-20s to 30-31 and the dewpoint goes from
> > mid teens to low 20s). OK, I reckon that, about 25nm west of Katherine,
> the
> > squall from the east met the seabreeze coming in from the west. If so, I'd
> > guess at a horizontal moisture gradient in the B/L of up to 10 degrees in
> > 5nm, causing the explosive development of the line disturbance. Seem to
> > make sense?
>
>Lotsa sense.  A gust front plowing into a pre-heated Td~20s airmass in the
>presence of "buildup-season" CAPE must be a good thing.  Could a careful
>surface analysis reveal that initiation took place precisely along the
>aged sea breeze front at the time the gust front made first contact?
>This event sounds like a good case study for tropical squalllinogenesis.

Hah! That'd be the day! Data? Are you serious? Unfortunately, doing a careful surface analysis over the Top End is about as viable as doing a daily analysis of 100m ocean currents in the central Indian Ocean. (It's one of the beauties of forecasting the weather up here. One can propose all manner of hypotheses that can't be disproved due lack of surface data.) The only surface wind observations to analyse between here (Katherine/Tindal) and Port Keats are Douglas River Research Station (too far north), Mango Farm (ditto), Timber Creek (too far south) and Bradshaw (on the southern end of the squall liney thingy but only feeling it once it was beginning to dissipate at that end - the AWS issued a SPECI for wind at
1547Z, 2 hours after the radar signature became evident). John Price is currently downloading the 10 min AWS data from Bradshaw for that day, but I wouldn't expect it to be very revealing - it was too far from the centre of activity. At Tindal, the wind blew light ENE'ly all day til the squall came through at 1300Z (wind 100/14G27kt), then the 'seabreeze' arrived at around 1700Z (wind went wnw'ly, dewpoint jumped from 15 to 21), much later than usual, possibly because of the fun and games out west.