Closing in on cyclone forecasts

Stu Nettle
Swellnet Analysis

Feared for their unpredictable nature and immense power, tropical cyclones have the capacity to devastate tropical communities. It's estimated 10,000 people per year die from tropical cyclones, while the associated wind, rain, swell, and storm surge can cause widespread damage to infrastructure and crops.

Between lives lost and the cost of damages, the need to better understand cyclones is compelling. The old adage of 'forewarned is forearmed' means an early warning system can save lives and property.

In 2015, residents of Tanna, the southern-most island of Vanuatu, were given hours warning before the eye of Tropical Cyclone Pam passed overhead. TC Pam was the most destructive cyclone in Vanuatu's history.

To date, the best science can offer is a pre-season outlook per ocean basin, heavily weighted by probability and statistics. For instance, on average four cyclones per season form in the Coral Sea with the BOM saying there was a 65% chance of fewer during season 2019-20.

As it happened, there was just one, TC Gretel, so the Bureau was correct but the information is too vague to be of much use.

This week, University of Newcastle in collaboration with New Zealand's National Institute for Water and Atmospheric Research, announced a new predictive model called Long-Range Tropical Cyclone Outlook for the Southwest Pacific (TCO-SP).

Here at Swellnet, we were informed about the model via an ABC article that made it sound almost too good to be true. Cyclone swells are notoriously hard to forecast, yet TCO-SP could allegedly forecast them better than swells from everyday lows and highs. Of course we were excited.

Yet it pays to adopt a skeptic's position. The claims were indeed too good to be true. A fact we discovered after speaking to lead scientist Dr Andrew Magee. No, the model couldn't forecast cyclones from years out, nor could it predict the exact track of cyclones - which is the aspect that tickled us as forecasters.

However, surfers won't be entirely disappointed with the capability of TCO-SP. After describing how the model works, Dr Magee said the South West Pacific basin - often treated as one region by weather agencies tracking cyclones - is treated as eleven "island-scale" regions in the new tool.

TCO-SP creates specific forecast data for eleven regions. It does so to narrow down warning information - Cairns and Rarotonga are both in the South West Pacific basin but at 6,000kms distant one size does not fit all.

Rather than a basin-wide forecast, TCO-SP creates specific forecast data for eleven separate regions (

Compartmentalising the region will help Queensland and northern NSW surfers. The best swell-producing cyclones usually form well off the QLD coast near New Caledonia or Fiji and remain distant from the mainland, thereby enabling a large spread of groomed swell free of meddling winds.

If, as Dr Andrew Magee claims, TCO-SP can tell us how many cyclones will form in the New Caledonian region, or the Fijian region, then we're heading towards something tangible. Still with many moving pieces in between outlook and realisation, but closer than we currently are.

Expectations can be further narrowed with TCO-SP giving monthly updates, as opposed to the BOM's lone cyclone prediction every October.

It's all theory at present, but come November we'll be watching TCO-SP very closely and marrying it up with actual cyclone swell.

There's a slightly perverse element in using TCO-SP to predict good surf - it was, after all, created to save lives and economies. Yet almost all the tools we use in modern surf forecasting have their roots in safety. Surfers are fortunate in that we (mostly) don't rely on them for safety but can utilise them for leisure.

Read the study in scientific journal 'Nature'
Keep abreast of cyclone predictions for the forthcoming season


freeride76's picture
freeride76's picture
freeride76 commented Friday, 24 Jul 2020 at 1:54pm

I honestly don't see any value in this at all for surf forecasting.

If they can't predict track, speed and direction of movement or intensity then it's nothing more than a piece of general knowledge.
A slightly educated guess about what kind of season may lie ahead.

Love to be proven wrong though.

stunet's picture
stunet's picture
stunet commented Friday, 24 Jul 2020 at 2:02pm

It's incremental, no doubt about it.

For surfers, it'll be a heads up that we're heading into a period where cyclone swell is more likely. Not definite by any chance, just more likely. 

Not unlike the Long Wave Trough, in a way.

The doc was adamant about its capacity to help island communities. For example, if they know there's a high likelihood of TCs then they'll look at an early harvest, or bringing fishing boats in, or securing loose matter.

Very keen to see how it works against real life.

freeride76's picture
freeride76's picture
freeride76 commented Friday, 24 Jul 2020 at 2:18pm

I dunno, where is the Coral Sea prob?

if its a La Nina year then you'd expect greater chance of CS cyclogenesis as opposed to SWP cyclogenesis.

Yes, SWP TC's tend to be the best ones (Wati , Sose etc etc) but not always (Anne/Agi, Betsy, Dinah etc etc) or marginal calls like Rewa, Roger etc etc.

Classic Noosa swells like TC Nancy tend to be CS systems, coast huggers.

I don't see an analogue between these forecasts and the position of the LWT.

stunet's picture
stunet's picture
stunet commented Friday, 24 Jul 2020 at 2:24pm

Analagous in that both cant pinpoint specific forecasts but can suggest a period that will have increased swell activity.

freeride76's picture
freeride76's picture
freeride76 commented Friday, 24 Jul 2020 at 2:30pm

I think the difference between them is greater than the similarity as far as useful forecasting tools go.

But, I'm happy to be proven wrong.

Anyway, according to the map it looks like quite an active South Pac season.

And I'm more than happy to keep the good wave boards on the top of the pile.

davetherave's picture
davetherave's picture
davetherave commented Friday, 24 Jul 2020 at 3:38pm

The BOM haven't come close recently in regards to Cyclone predictions. With sea temps off the east coast being what they were though, I think we all thought there would be more activity, especially moving north to south along our coast, rather than away or trying to holiday in nz.
Tying in with the mysto swell thread though, i think any information over a broader area can help.
I remember just before i left Mackay, i had been sacked at the radio station and i went out to north wall just to clear my emotional state.
Whilst locally there had been no real indicators and strong wind generation to create swell, i was greeted by lovely 2-3 foot peaks which i had to myself.
At the time I believed it was Huey telling me things would be ok, but obviously there were other synoptic states out wider at play and on the big tides the swell was able to get in.
But looking at the local synoptic chart, it couldn't be seen.
Still think it will be hard to be accurate, end up being closer to Cyclocane.


Fluiddreams's picture
Fluiddreams's picture
Fluiddreams commented Saturday, 25 Jul 2020 at 9:23am

Better technology = more crowds.

truebluebasher's picture
truebluebasher's picture
truebluebasher commented Tuesday, 20 Oct 2020 at 9:26pm

swellnet La Nina switch > East Coast Big Picture

Qld crew > "Oz 'La Nina' Cyclone Watch 2020/21"

[ WARNING ] Cat 5 (vid) Cyclone Tracking OZ
4 Tracks + 1 CAT - 5 + Sea Temp
WA 5 < > NT 3 < > Qld 6
BOM Names (etc...)

French :
Fiji : (eg: Oma /Yasi )
More: Naming + Wiki 2020/21 Cyclone Watch > (re: BOM Vid below + Full detail)

Fiji Met ~ 2020/21 Cyclone Alley (re: Wider version of Track 3 in CT OZ vid)
Papua > Solomon's > Fiji > Vanuatu > New Caledonia > New Zealand
Qld (+) 3-6 < > Polynesia (-) 1-2
(Meteo) Madagascar 3-5

BOM (vid) +Text version
WA +63% (7) < > NW +63% (5) < > NT +57% (3) < > Qld +67% (4)

TCO (Coral Sea)
WA 5-8 < > NW 3-6 < > NT 2-5 < > Qld 3-6