23-26 July

Outcome: Perisher Valley claims 65 cm; Falls Creek 45 cm


Friday night surprised on the high side in NSW, with Perisher Valley claiming a 30 cm snowfall. The excellent winterised automatic weather station there recorded over 30 mm of sub-zero precipitation (at 1738 m elevation), so the claimed fall looks reasonable. In Victoria, Falls Creek claimed 16 cm and Mount Hotham 15 cm, more in line with my expectations for this part of the event.

Saturday night – Sunday produced about 30 cm right across the alps, also well above what I expected. (Probably I should just give up predicting snowfall depths — I’m always wrong, usually on the low side.) Here’s the event in animation; note the big blob of JAXA-inferred alpine precipitation as the main front passed through on Sunday:


JAXA precipitation inferred from satellite imagery

BOM surface chart archive

Update #4 (22 July):

Well Wednesday is certainly wet, though with a bit more snow in it than I expected by this stage (late morning). We may see a depth gain at tomorrow’s measure at Spencers Creek (remembering that about 15 cm fell after the measure last week).

Friday afternoon / night’s precipitation is looking a bit colder than today’s, so using today for calibration, expect snow down to at least 1500 m , heavy at times. I think upwards of 20 cm 10 cm. Then Saturday night / Sunday looks very cold and windy. There’ll be snow to low levels, but unless the projected track of the low moves north, not a large fall … probably less than 20 cm.

Then next week BOM’s ACCESS-G model is right out on its own showing a little follow-up low Monday night, with still more snow. Does it have “local knowledge” … we shall see. Add that, and you’re looking at another official “dump” (50 cm gain) at the next measure.   Nope, that bit’s not happening — a few centimetres Monday, that is all (23 July adjustments).

Update #3 (20 July):

Wednesday looks wet as foreshadowed at update #2, although there may be some snow in the evening — particularly BOM’s ACCESS-R model has atmospheric thicknesses* and temperatures suggestive of snow at higher elevations from late in the day. But the main action is showing on the weekend — mainly Sunday — so I’ve extended the date range by a day. It starts with some more marginal precipitation Friday night (I think snow at higher elevations), then the major cold is back on Sunday, though maybe with a little less moisture this time:

GFS for Sunday, 26 July (Levi Cowan)

ECMWF for Sunday night (Levi Cowan)

(* Atmospheric thickness provides a rough-and-ready guide to average atmospheric temperature. It’s just simple physics — gas volume (“thickness”) is proportional to absolute temperature, all else being equal. A height difference or “thickness” of 5400 m between the 1000 and 500 hPa pressure levels usually indicates a lower atmosphere temperature cold enough for snow in our alps — that’s the blue “540” line on the first chart.)

Update #2 (18 July):

The models are now suggesting two burst of activity for this interval. As foreshadowed in update #1, the first peaks too far west, arriving about Wednesday night (22 July). The US GFS model and BOM’s ACCESS-G have it slipping south and triggering a warm trough in the wake of the high, which could bring widespread rain. The ECMWF scenario is more favourable for snow:

GFS for Wednesday, 22 July (Levi Cowan)

ECMWF for Wednesday night (Levi Cowan)

The second burst of activity is shown arriving about Saturday (25 July), and currently looks cold and snowy, if not especially strong.

Update #1 (13 July):

This action is well over a week away, so plenty could change. There’s still a clear node on the GEFS spaghetti, but it might be peaking just a little early — more Adelaide than alps:

Original post (10 July)

After next week sorts itself out, the next burst of snow weather is showing around the middle of the following week. Enjoy it while you can; on average August is not the snowy month it once was, and El Niño is forecast to strengthen further (click the chart for a thorough explanation and the source link):