Observational results point to climate impacts, but conclusive findings from modelling prove elusive.

A group of 23 scientists from around the world have concluded that climate change increased both the likelihood and severity of the rainfall unleashed by Cyclone Gabrielle, but have been unable to put a definitive number on that impact. 

The rapid study by a team from the World Weather Attribution initiative, including several New Zealand scientists, produced a “mixed bag of results”, said Sam Dean, principal climate scientist at Niwa. While observational data indicated that climate change had exacerbated the downpours in the areas focused on, Tairāwhiti and Hawke’s Bay, modelling was not conclusive. 

“Based on the analysis of the observation records, we undertook a statistical analysis which showed that the rainfall due to Cyclone Gabrielle was increased by 20% to 30% and was made three to four times more likely as a result of the human emissions of greenhouse gasses into the atmosphere which has caused the planet to warm,” explained Dean at a media briefing. 

As well as analysing historical weather station data, Dean and his collaborators around the world ran thousands of simulations on supercomputers to compare the weather event in the climate as it is today with the climate before the impacts of warming. That modelling proved unable to quantify the influence of climate change on the rainfall produced by the cyclone, which ripped across the north and east of the North Island a month ago, wreaking massive destruction and 11 deaths. 

Rapid analysis of severe weather and climate attribution is a burgeoning branch of climate science. The World Weather Attribution group looks at specific events with a view to urgently assessing the extent to which climate change is more likely and more extreme, with the ambition of putting that information into the public domain while the impacts are fresh in people’s minds. 

“The [Gabrielle] event is challenging us as climate scientists,” said Dean, who had in recent weeks endured “a lot of sleepless nights trying to work on this”. Gabrielle was “a gargantuan event, and I have no doubt whatsoever in my mind, based on my experience as a climate scientist, that climate change has influenced that event. But do we know it’s exactly 30%? No, we don’t,” he said.

“We would never say that climate change caused this event,” Dean stressed. “This event could have happened without climate change. It would have been quite rare in a world without climate change. It is now not quite so rare. It is still an unusually extreme event.”

Researchers focused on the impact of climate change on rainfall. The study did not assess the role of wind, nor did it seek to measure how climate change might have exacerbated the damage. “Even small changes in intensity can play through to a much bigger depth of impact,” said Dean. “It might be that the last bit of water that breaches your stopbanks was what climate change added. That’s doing a lot of the damage. So if you did [assess] damages, then you might well find a much larger impact from climate change. It depends on the situation.”

Another of the scientists involved in the project, Luke Harrington, a senior lecturer in climate change at the University of Waikato, said that the observations showed clearly that extreme rainfall was occurring with greater regularity in the region studied, and even when La Niña and the marine heatwave were factored in, “we don’t have a clear alternative explanation for [the increase] other than climate change, particularly over this long period of time”.

He said: “I appreciate that there’s always a desire to come up with a singular answer as to what the impact [of climate change] in an event is, but in this situation we’d be much more comfortable providing a range, and acknowledging there is uncertainty there.”

When evaluating the modelling results, said Harrington, “we have quite a range of uncertainty as to what the real world is in nature, in terms of the statistics of extreme rainfalls in this part of the country. So our ability to discard models on the basis of them not performing well was limited. We need to spend a bit more time looking at the reasons why these different models show different answers.”

He noted that they were still awaiting rainfall data from stations which remained difficult to access as a result of the storm, and that some of those had longer records which would help to draw conclusions. 

Friederike Otto, a climate scientist at Imperial College and a key figure in the “rapid reaction force” of severe event weather attribution studies, said interest in and understanding of the branch of climate science continued to grow, accelerated by the IPCC Working Group 1 report “which put a huge emphasis on the advances made in event attribution”. The collaborative energy had similarly increased. “Finding scientists to work with in all countries has become a lot easier,” she said.

Another of the New Zealand scientists involved in the study released today, Dave Frame of the University of Canterbury, told The Spinoff that a commitment to “the integrity of the scientific process” meant that those involved were determined not to overstate the conclusions. “We know there are plenty of events where we can quantify the climate change contribution fairly well. We’re confident here that the obs are showing a trend. These models don’t really give much of a signal but this is the first step, the first piece of the puzzle.”

He said: “If we’d pulled that process because we didn’t like the answer, well, that would be shabby.” It might be a hackneyed phrase, he said, but “sometimes further work really is needed”.