The last month has seen record-breaking high temperatures hitting the northern hemisphere, with simultaneous heat waves in Japan, Greece, France, the US, Ireland, Canada, and Sweden. This ferocious and devastating heatwave is now considered a natural disaster in many parts of the world. And according to new research, the fever’s not over yet.

The study, published today in Nature Communications, reports the use of a new probabilistic system that predicts mean temperatures; according to its predictions, the forecast is a global, four-year hot period from 2018–2022. The study was led by Dr Florian Sévellec, a CNRS researcher at the Laboratory for Ocean Physics and Remote Sensing (LOPS) (CNRS/IFREMER/IRD/University of Brest) and at the University of Southampton.

The method used to make these predictions uses an entirely new system, deviating from traditional simulation techniques. It searches 20th and 21st century climate simulations for similar or equivalent records of current climate conditions, and, based on this information, makes predictions about future possibilities. It’s at least as effective as current methods, particularly for the simulation of global warming hiatus of the beginning of this century.


The data were obtained from the Coupled Model Intercomparison Project Phase 5 (CMIP5), which is a global effort to examine climate predictability. It has close ties with the Intergovernmental Panel on Climate Change, which is the international body for assessing the science related to climate change.

The data that we used to train the method are simulated data of the 20th and 21th centuries from the CMIP5”, says Dr Florian Sévellec, “Because they are plenty of models – 10 in our case – and of realization, around 5 per model, there is hopefully plenty of data, even of extreme events.”

It’s processing speed is also something to swoon over; predictions can be made in a few hundredths of a second on a laptop, whereas supercomputers require a week using traditional simulation methods.

“At this stage, we do not know where the anomalous warming will occur”

-Dr Florian Sévellec

The new forecast system has predicted a low probability of intense cold events, which is, in part, responsible for higher temperatures. With the heat comes drought, and with drought comes an increased risk of disastrous wildfires, as recently seen in California, Portugal, and, of course, Greece. The next hot four years are likely to completely change landscapes worldwide, but forecast systems can at least allow us to prepare for the worst. According to the published study, we can also expect to see abnormally high air temperature, sea surface temperature changes, and more frequent and severe tropical storms.

Screen Shot 2018-08-14 at 16.50.42.png

Screen Shot 2018-08-14 at 16.50.51.png

For now, the method is generating global mean temperature predictions, but the researchers would like to make adaptions that make regional predictions possible – sure, it’s going to get hot, but where? “At this stage, we do not know where the anomalous warming will occur – this is something we need to work on”, says Dr Sévellec, “[Making regional predictions] is far from easy. This is a huge development. Even if the principle/method will be equivalent, we don’t yet know if we will still have skilful prediction at the regional level.” In addition to temperatures, this method could also be potentially developed to estimate precipitation and drought trends.

QUOTE SOURCE: personal communication