John,

I figure I'll have a better chance to do this from home as tomorrow will be a bit messy in the RFC. I was on nights over the weekend that you're interested in.

I don't know what sort of strength in the northerly winds you think is necessary for the extreme temps

"but the gardient wasn't all that strong..."

but the early morning wind flights from Melb airport on both Saturday  and Sunday mornings had 45 knot+ northerly winds at gradient level or below.

 On Saturday night a shallow trough pushed through southwest Victoria, reaching Melbourne and stalled overnight.  To check out the shallow nature of the cold air, see the 10Z temp trace Saturday night.  Winds were southerly over port Phillip until the early hours of Sunday night  then northerly flow re-developed.  However the high in the Bight was strengthening and a trough was passing through Tasmania to recharge the southerly flow over Victoria.

So, at 16Z there were 20knot southerlies in the south of port Phillip and the upper winds at Melbourne Airport had 45knot+ winds in the low levels - not that common a combination. The strong northerlies mixed the very warm air back to the surface and Melbourne's temp rose again, peaking at just under 33  just before 5am - the issue time for the Melbourne forecast as luck would have it.

Now all we well trained mets know that the forecast temps are verified  9am to 9am and that works fine most of the time.  But it can get a bit confusing for everyone, especially those ordinary folk who think the day  runs from midnight to midnight.

The upshot is that I was unsure how long it would be before the cooler  air reached the city so the 5am forecast went out with a max temp of 33 - and I knew an amendment would have to be done in a relatively short  time.  After all how could you issue a forecast max of 24 when the current temp is 33?  The amended forecast temp was issued an hour or two
later.

The various locations around the city presented a problem with a moving  trough.  How long would it take the cool air to reach the eastern suburbs, how long before the depth of the cool air became sufficient to affect Mt Dandenong?  Hence the range of temperatures forecast.  I don't know how they all eventually verified - post 9am that is, not the
verification of the temp on the actual calendar day which we all  dutifully ignore in this sort of circumstance. I wonder which
temperature the public might think more appropriate?
 
Regarding the misinformation about the record(?) temp for Melbourne.  I  don't know how the 1904 temp came to be missed but the second 1939 temp  may have been unnoticed because of the way the climate temps are displayed on screen in the Vic climate section's display.  One can see the highest for the month or all the temps in the month , but not all
the info at once.  When scanning quickly under time constraints it's normal to look at the highest for the month.  It was a weekend. There were no climate and consultancy people to help with the background while all in the RFC dealt with the immediate, and there was quite a bit of immediate.

Hope this helps a bit

Ward