ABC Weather - Kate Doyle
As cyclone Marcus ripped through Darwin and fires
burned homes to the ground in Victoria and New South Wales, it was just a
matter of time until the climate change discussion reared its head.
This week the Greens drew a link between the two,
with comments made in the Senate and in the media, saying that fires are
now more severe and frequent because of climate change.
The Prime Minister Malcolm Turnbull later commented that he
was reportedly "disappointed" the Greens would politicise the events.
"You
can't attribute any particular event, whether it's a flood or fire or a
drought or a storm — to climate change," he said while on a visit to
Bega, near the bushfire-affected areas.
Two people who have
protested Mr Turnbull's remarks are Dr Sarah Perkins-Kirkpatrick from
the University of New South Wales and Dr Andrew King of the University
of Melbourne.
They are two of Australia's leading researchers in
climate change's role in extreme events and have spoken to the ABC
regarding the field of 'climate change attribution'.
How does climate change attribution work?
The
basic principle behind climate change attribution is comparing the
world as it is, with how it would have been without human-induced
greenhouse gasses.
To do this researchers use climate models
which work like computer-based virtual worlds, to recreate the real
world as closely as possible.
They then look at two sets of model
experiments, one which is as close as possible to the current world,
and one where the human introduced greenhouse gasses have been removed.
"We
look at the frequency of that specific event between those two
simulations and then compare how often it occurs now, compared to the
natural world — as we think it used to be," said Dr Perkins-Kirkpatrick.
"Essentially what we're doing is looking between
those two groups of models and how the probability or intensity of
extreme weather events [such as] heat waves, heavy rainfall events and
droughts, are changing just between those two ensembles," said Dr King.
In
the past, these attribution studies have found a link between the
Canberra and Sydney Heatwaves of February 2017 and climate change.
For
more complex weather events, for example Cyclone Debbie, or the 2011
floods of south east Queensland, Dr King says "it's harder to tell."
Part of the reason is that there is a lot of complexity in this analysis.
According to Dr Perkins-Kirkpatrick, "It's not that easy.
"A lot of time, a lot of blood, sweat and tears go into actually defining the event and making sure you've got it right.
"It's taken me weeks before to make sure I've captured the event as well as I can."
How
the event is defined, the model used, which parameters are included and
how the data is analysed statistically, can all change the outcome, so
these simulations are often repeated many times to ensure a robust
result.
What are the limitations?
Dr Perkins-Kirkpatrick said one of the limitations of climate change attribution studies is that they are heavily model-based.
"You have to be certain, or very confident at least, that that model is a good representation of the actual climate," she said.
"Unfortunately,
however, no matter how good your model is, we simply can't really
measure what the climate used to be like before climate change actually
started."
There are some old observations, mainly in Europe, and paleo records, that are helping to improve knowledge in this area.
But our understanding of pre-industrial revolution conditions are not as good as our understanding of the current climate.
The type of event matters too.
Dr King said that hot or cold
extremes are quite easy to attribute, especially on a big scale; other
more variable events like rainfall, are more difficult.
"More
complex events like fire weather, which has a combination of high
temperatures, low humidity, strong winds — those are a lot harder."
Dr Perkins-Kirkpatrick also puts individual cyclones in the variable pile.
"It doesn't necessarily mean with those events that there is no anthropogenic signal," said Dr Perkins-Kirkpatrick.
"It just means that we can't yet detect it because there's a lot of variability overlying that particular signal."
'Weather' vs. 'climate'
Communicating
the difference between long term-climate trends and individual extreme
weather events is where all of this gets messy.
The weather is what is going on day-to-day; the climate is what is happening over time.
Using the wardrobe analogy: climate is all of the clothes in your closet, while weather is what you wear each day.
Which raises the question: does every event need to prove or disprove climate change?
According to Dr Perkins-Kirkpatrick the answer is no.
"A
lot of the events — I'm just saying a lot, I'm not saying all on
purpose — would have probably occurred without climate change, but now
they're occurring more frequently," she said.
"And that's exactly
what attribution looks at: whether or not a particular event is
occurring more frequently because of climate change,.
"We've always had tropical cyclones for example, they're always going to occur.
"It doesn't necessarily mean that every single tropical cyclone needs to be attributed to climate change."
"And we look at every event separately because they're all very different and very individual."
Likewise, not every cold snap means climate change is wrong.
"It's
quite frustrating as a climate scientist to hear people saying that,"
Dr King said. "Especially if it's the president of the United States,
it's not very helpful."
"We're always going to have that variable weather — even in a hundred years.
"It's
just that the warm extremes are a bit warmer, the cool extremes are not
quite as cold as they would've been in the past."
Both Drs King and Perkins-Kirkpatrick are trying to move away from
publicly assigning a number or percentage to how much climate change has
contributed to an individual event, despite media pressure.
Dr
Perkins-Kirkpatrick said she understands the pressure from the media,
"They want that analysis because it is interesting and it does show that
climate change is actually happening now."
The researchers explain there is only so fast they can get such analysis done.
"We're reasonably confident in the method, but sometimes the media wants these results yesterday," Dr Perkins-Kirkpatrick said.
A fully quantifiable scientific analysis takes time.
"When we do put a number on it, it's not necessarily one that we've plucked out of thin air," Dr Perkins-Kirkpatrick said
"We've
looked at the event, we've repeated the analysis thousands of times, so
we have a range, and we never give the number that's the most scary, or
the highest number — we give a very conservative number."
"So even if we say an event is, for example, twice as likely due to climate change, that's probably the lower estimate."
For Dr King, it is about making sure the context is there.
"We
always give uncertainties with those numbers but often it's lost in the
communication from the media sources to the public," he said.
So last weekend's events?
Last
weekend's fires and cyclone may have been later than most but they
still occurred within both the official NSW Rural Fire Services Bush
Fire Danger Period and the NT's cyclone season.
With all of this confusion, is climate change attribution useful?
Dr King said it was the extreme events that people remembered.
"Understanding how [extreme weather events] are changing resonates with people I think," he said.
"So it's a really good communication tool."
Dr Perkins-Kirkpatrick said that ultimately, it is useful.
"Especially
if you have all your methods pointing towards yes," she said. "You can
say with confidence — with high confidence — that there's a signal
there."
"Conversely, certain events you can't necessarily attribute to climate change.
"And
if all your methods say the same thing, that climate change did not
affect this event, you can say that with confidence as well."
Links