Season 2015 snow depth prediction

Updated: Corrected error in the recent SST plot, and updated it for February which is now available from HadISST (refresh browser to see). Also the NOAA optimum interpolation SSTs previously used for recent guidance in fact bear little relation to the HadISST series used in the model, so I’ve switched back to BOM weekly SST chart. The snow depth prediction is unchanged.

It’s time to give this a try.   The Spencers Creek¹ season peak snow depth for 2015 will be  141 ± 44 cm.

Last time I described my new mark-IV prediction model with an extended “training period” and incorporating the effect of volcanic aerosols. The new formula is:

Spencers Creek peak depth (cm) = 899 – 0.311 x year – 13.8 x AAO + 2.09 x SOI – 8.80 x IOD – 5.36 x PDO – 143 x SST + 534 x aerosol

Once more to the parameters, which are explained in the notes at the end. Click on the graphs for sources. This is a longish read, so maybe grab a coffee.


Calendar year

We’re up to 2015, so take off yet another centimetre for gross obfuscation and political inaction.²


Antarctic oscillation (AAO)³

This is always the hard one. AAO exhibits moderate persistence (months), but it can also change suddenly. There is no meaningful numerical prediction that I’m aware of (this prediction is short range — it’s just AAO calculated from the 15-day GFS output). Here’s how AAO is looking (dynamic link, will update):

Antarctic oscillation

We want the winter average, the “3-month running mean” to the end of August. I think it’ll be positive (which is negative for our snow), but perhaps not hugely so. I’m going with +1.0.


Southern oscillation index (SOI)⁴

SOI reflects ENSO, which just like last year seems to be headed into El Niño (meaning SOI generally below about -10, i.e. minus one monthly standard deviation, and staying there for several months). The Australian Bureau of Meteorology’s (BOM’s) POAMA model has this prediction (will update; remember a positive Nino3.4 temperature anomaly correlates with a negative SOI):

Internationally, the various model predictions look like this (the 2015 winter average that we want is at the second tick in from the right edge on each; static image links):

Hmmm … most of the dynamic (physics-based) models that IRI plots seem to be looking for an El Niño, but the statistical models are not. (I have an inherent bias towards physics, but I can’t exactly say I dislike statistical models … given that this is one.)

Once again this ENSO thing could be pretty bad this year, but given last year’s experience it might not pan out. You’d have to say we’re overdue though. (April-May is the time of year when ENSO prediction models have their lowest skill, and that’s never particularly high at the best of times.) Like last year, I’m adopting a winter average SOI of -10, which is bad but not terrible.


Indian ocean dipole (IOD)⁵

BOM’s model has this (will update):

It looks like we can expect near neutral IOD conditions again this winter. I’m adopting -0.1, which is very slightly positive for our snow.


Pacific decadal oscillation (PDO)⁶

There is now strong evidence that the PDO has flipped back into its positive phase, which is negative for our snow. The model uses the two year average to the end of August, which of course lags the flip and so hasn’t yet turned strongly positive (but next year…). Most of that two year average is already in. The current average from September 2013 to February 2015 is +0.86. I’m adopting +1.0.


Sea surface temperature (SST)⁷

Don’t think there’s much global warming? Then what’s this:

The trend in sea surface temperatures in our area of interest is relentless, and the correlation of that with our season peak snow depth is the strongest of the lot. One degree Celsius rise in local sea surface temperature loses us at least a metre of expected peak snow depth.

Fortunately our winter SSTs have been near trend in recent years (they could easily have been much worse). SSTs have been a little above trend lately, but not by much:

SST_box SST_monthly_recent

That data is from the Hadley Centre’s HadISST dataset through February. Where is this headed? Here’s what NOAA’s CFSv2 model predicts for June, July and August (anomalies vs 1999-2010, add about 0.5°C for 1951-1980 in our region; will update):

CFSv2 SST anomaly prediction for June-July-August 2015
(vs 1999-2010, add 0.5°C for 1951-1980)

Wow, look at that El Niño (hot water in the eastern equatorial Pacific). The classic positive PDO pattern of a sideways-U shaped band of warmer water in the north Pacific is also obvious. That model probably has negligible skill in our little area from this far out, nevertheless the local pattern it predicts is one we’ve become used to in the last couple of years — a very warm western Tasman and a nearer “average” Great Australian Bight. But if that Tasman SST hotspot were to pan out as shown (+2.5°C !!), we could be looking at a wipeout season.

Here is what BOM has for SSTs over the last available week (anomalies vs 1961-1990, add 0.1°C for 1951-1980; will update):

Latest BOM weekly SSTs vs 1961-1990, add 0.1°C for 1951-1980

So as I said, things have been warm in the Tasman but not so much in the Great Australian Bight. This is a really tough call, but I think the chances are our local SST anomalies will stay near trend, which is of course strongly negative for our snow. We’ll know more as we get closer, but for now I’m adopting a winter average SST of +0.65°C vs the 1951-1980 average (no trend adjustment is needed this year because the detrending used rotates around 2015).


Southern Hemisphere stratospheric aerosols⁸

There have been no major volcanic eruptions affecting the stratosphere, but low levels of aerosols have persisted lately (paywalled text) from minor eruptions and pollution. Global coverage aerosol data is available from satellite remote sensing (e.g. from MODIS), but that is of course total column, not stratospheric (and the tropospheric aerosols swamp the signal). For a real-time guide to the stratospheric situation, the high-altitude Mauna Loa data (Hawaii … yes I know, not SH) provides a rough indication:

Optical thickness is the negative log of transmission, so the left panel is an inverted view. The right panel is a direct stratospheric aerosol measurement by laser probing (LIDAR). It’s looking to me like stratospheric aerosols may have fallen back a bit since the last published NASA GISS Southern Hemisphere estimate of 0.0036 in late 2012. I’m adopting 0.003.



In the new model, those seven parameter choices give a 2015 best-estimate peak snow depth of 141 cm.  Sorry guys; that’s just how it is. The 1-σ error is 44 cm, so the range 97 – 185 cm would be expected to include about two-thirds of likely outcomes, if the parameters were perfectly known (they’re not, of course).

For comparison, a statistically thorough naive prediction would yield a 2015 peak depth of 176 cm, with a ±1-σ range from about 120 – 240 cm. On that basis my prediction is for a well below average peak depth (whatever “average” means these days with our alpine climate changing so rapidly). Also, for what it’s worth, my mark-III model would have predicted a peak depth of just 130 cm. Here is the comparison:

Spencers Creek peak snow depth pre-season prediction for 2015

Spencers Creek peak snow depth pre-season prediction for 2015

My suggestion: ski the second half of July or very early August.



  1. Spencers Creek near Charlotte Pass, NSW, Australia, midway between Perisher Valley and Thredbo; data courtesy Snowy Hydro Limited.
  2. The current raw downtrend would be about 1 cm/yr allowing for the non-linearity. Only 0.3 cm/yr appears in the calendar year term of the prediction model, because it’s a linear model and because there are global warming trends in some of the other parameters.
  3. Antarctic oscillation (AAO), also called “southern annular mode” or SAM, is a measure of how tightly the circumpolar winds (“polar vortex” in one usage) blow around the pole. A loose pattern (negative AAO) leads to more polar storms reaching southern Australia, and more snow. The winter average AAO is used — the average of June, July and August.
  4. Southern oscillation index (SOI) is the difference between Tahiti and Darwin surface atmospheric pressure expressed as monthly standard deviations times ten. SOI is an indicator of the El Niño Southern Oscillation (ENSO), an east-west quasicycle in equatorial Pacific Ocean surface temperature and wind patterns which correlates with precipitation across much of Australia, including with alpine snow. A positive SOI is associated with more (and wetter) Australian snow. The winter average is used.
  5. Indian ocean dipole (IOD) is an ENSO-like variation in the smaller Indian Ocean, which correlates with winter precipitation across southern Australia, including with alpine snow. Negative IOD is associated with more snow (the sign is consistent with SOI, but we’re on the opposite side of the Indian Ocean). The winter average is used.
  6. Pacific decadal oscillation (PDO) is a long-cycle, largely north-south variation in the Pacific Ocean, which interacts with (and perhaps partially mediates) ENSO. Negative long-average PDO is weakly correlated with more Australian snow. The 2-year average to August is used.
  7. Sea surface temperature (SST) is that in the Great Australian Bight and northwest Tasman Sea, averaged over the box: latitude 30-37°S, longitude 115-160°E, and expressed as degrees Celsius anomalies from the 1951-1980 mean, detrended about 2015. Cool SSTs correlate strongly with more snow, but unfortunately local SSTs are rising rapidly with global warming. The winter average is used. More at my post here.

    (I detrend the SSTs to make the model equation parameters appear more sensible, and to make them more comparable with previous versions. Because of the model linearity, linear detrending doesn’t alter the regression outcome — it just shifts the SST trend effect across to the “calendar year” term.)

  8. Stratospheric aerosol optical thickness is the average for the Southern Hemisphere at 550 nm wavelength (greenish visible light). Large optical thicknesses from big volcanic eruptions correlate with big snow seasons. The winter average is used. More at my post here.