Trans-Tasman seesaw?

You can’t watch snow weather in our part of the world for long without noticing that when things are good in Australia, they’re often not so good in New Zealand, and when they’re bad here, they usually look much better over there. Take this season — in late June and early July we had a series of good snowfalls while NZ struggled; then in late July our snow weather pretty much shut down, and ever since it seems to have been snowing steadily in NZ (especially in the North). Is this a real effect, or just a nice illusion? Lets look at the data.

I’m not aware of any high quality long-record snow depth data for New Zealand similar to our Snowy Hydro and AGL Hydro records. The best available records appear to be those from the resorts, particularly the ones kept by NZ website Here’s their record for Mt Hutt:

Mt Hutt snow depth

Mt Hutt snow depth graph, from archive of


There’s a litte problem with the plotting colours there (is orange 2013 or 2004, blue 2012 or 2003?). They promised to send me the raw data last season, but so far no luck (bump…). Fortunately it’s easy to disambiguate via the web archive (of course 2013 was the big one). The nzski records are short (mostly post-2000), and they are resort depths, so presumably affected by resort activities like snow making, grooming and, err, other imperatives. And it looks like they usually don’t measure after the resort has closed, hence the “cliffs” at the end of some year’s data.

Whatever … they’re what we’ve got. Let’s look at just the peak depths against Spencers Creek in Australia (midway between Perisher Valley and Thredbo):

Mt Hutt peak snow depth vs Spencers Creek


Whoops, that’s what I call a scatter plot. The dubious correlation shown is weak and slightly positive, and on a simple test there’s nearly 30% likelihood that it’s just the result of chance. Not quite a seesaw…

If there is a seesaw effect it seems to me that it’s likely to arise from some sort of phase locking of the waves in the polar circulation, which again is anecdotally popular. “All the systems are peaking over Perth and NZ at the moment”, that sort of thing*. If so, we probably need to look for that reverse correlation in shorter duration data. For that I need the full NZ dataset.



* That’s pretty much the old LWT notion. More recently some people have been suggesting it may be linked to the (circum-equatorial) Madden-Julian oscillation. Strong phase 5&6 MJO certainly correlates with more snow in Australia, but does it also correlate with less in New Zealand? That seems doubtful; it’s a broad-scale thing.