When does our snow fall?

I guess most people look at snow depth charts for an idea of how good the cover is — or might be on some date or other. Looks like there’ll be about a metre at Spencers Creek (near Charlotte Pass, half way between Perisher and Thredbo), so there should be decent cover and good skiing. But it’s not quite that simple. A metre at Spencers Creek in late June probably means extensive fresh cover; a metre in late September more likely means pretty sparse spring slush.

You can get more information by looking at Snowy Hydro’s other published snow depth records at lower elevations … but there’s more information in the Spencers Creek record too, just by looking at the week-on-week change:

Spencers Creek inferred weekly snowfall and melt

Spencers Creek inferred weekly snowfall and melt

The chart shows the season 2013 changes in snow depth between consecutive readings (adjusted to cm/week) against a background of all the weekly depth changes1 for the full record back to 1954 (2013 from the raw data; background points from the weekly resampled record). You can see that about now the season usually flips from net weekly snowfall to net weekly melt2, on average. That “on average” is important. In our erratic climate there can be significant snowfall or melt in almost any week between late April and November3. The “net” is important too. Lots of recorded weeks would include both significant melt and significant snowfall.

I find that the best skiing — resort and backcountry — usually occurs after there’s been two or three weeks of substantial net gains, in the 20-30 cm territory. That generally ensures both extensive cover and a decent surface. The blue median curve shows that a weekly net gain is more likely than not (what median means) only in late June, in July, and in most of August. Good snow is fragile here.

Notes:

1. Actually there’s one missing. The record gain of 109.5 cm for the week ending Thursday 8 August 1996 is off the top of the chart.
2. Melt and/or consolidation. Both affect simple snow depth (as opposed to water-equivalent depth).
3. It can of course snow in any month (even quite heavily), but settled snowfall persisting long enough for the Hydro to bother measuring a net gain has only occurred in April through November.