快猫短视频

When temperatures turn extreme, we rush to adjust our heating or cooling systems in an attempt to remain comfortable. With severe weather events becoming ever more frequent due to climate change, so too does our fiddling with the thermostat, which in turn is causing spikes in daily carbon emissions, leading to yet more climate change. Failing to break this vicious cycle will mean the problem only gets worse.

at Tsinghua University, China, and his colleagues have been tracking global daily CO2 emissions since January 2019 to examine how seasonal cycles, time of the week, the weather and other factors influence them. They used data collected between 2019 and 2022 to train a machine learning algorithm to estimate daily CO2 emissions back in time to 1970, at a global level and at a national level for key countries.

How climate change has pushed our oceans to the brink of catastrophe

For decades, the oceans have absorbed much of the excess heat caused by greenhouse gases. The latest observations suggest they are reaching their limits, so how worried should we be?

The results, unsurprisingly, revealed a large increase in global daily CO2 emissions from 1970 to 2022, from 50.6 megatonnes a day on average in 1970 to 106.9 megatonnes a day in 2022. We already know emissions have risen dramatically over the past 50 years as countries around the world have industrialised.

More interestingly, Liu鈥檚 team also identified a previously overlooked trend in increasing emissions due to extreme hot and cold days, defined as days where temperatures fall below the 5th percentile or above the 95th percentile over 1970-2022 for each country.

For example, daily emissions during extremely cold days jumped by 11 per cent in Germany over the study period, compared with the average for the corresponding month, and emissions increased by 4 per cent in Japan during extremely hot days.

The increase in emissions due to extreme heat events accelerated over the 2010s, when the average number of days of extreme heat experienced each year more than doubled compared with the 1970s. Between 2019 and 2022, daily emissions during extreme hot days in Japan were 12.3 per cent higher compared with the monthly average, for example. 鈥淢ore frequent extreme temperature is resulting in more emissions,鈥� says Liu.

Liu warns the emissions spikes are a consequence of poor adaptation to a rapidly changing climate. This will create a feedback loop where extreme weather drives higher emissions, which in turn intensifies climate change. 鈥淚t鈥檚 definitely not good news for the world鈥� it鈥檚 difficult to break,鈥� he says. Even an accelerated switch to renewable energy is not a complete solution, as power plants, including renewable energy units, become less efficient in extreme temperatures, leading to an uptick in fossil fuel use to make up the difference.

The climate impact of feeding ourselves is getting worse and worse

If we want to reduce the greenhouse gas emissions from food production, we have to make it more efficient as the global population grows 鈥� and that isn't happening

The team also identified a 鈥渃ritical temperature鈥� for each country, where emissions are generally at their lowest level. In China this temperature is 19.5掳C (67.1 掳F), in the US it is 15掳C (59掳F) and in Japan it is 18.2掳C (64.8掳F). For most countries, when temperatures fall below or climb above this critical point, emissions start to increase.

Team member at the Center for International Climate Research in Norway says governments will need to consider the impact of increasingly extreme weather and the way people respond to more and more uncomfortable conditions when designing future energy systems. 鈥淚f you have an increased demand for cooling, and more people buy air conditioners, then it can amplify the effects,鈥� he says.

鈥淲orsening climate impacts can affect societal emissions of greenhouse gases that are driving the warming of climate in the first place,鈥� says at the University of Reading, UK, leading to a 鈥渧icious cycle鈥� of escalating emissions. 鈥淭he new study is the first to really dig into day-to-day fluctuations and construct regional relationships between long-term changes in emissions and temperature.鈥�

at Imperial College London says the study underscores the need for urgent climate policies to cut emissions and ensure low-carbon adaptation. 鈥淚t just shows the importance of going to net zero, both to limit the amount of heat extremes, but also to limit the amount of additional CO2 that gets produced because of those heat extremes,鈥� he says.

One limitation of the work is that the data the researchers used to train their algorithm include two years of the coronavirus pandemic, which saw emissions dip globally. The algorithm also does not account for differences between 1970s society and the present day. Levels of access to air conditioning, for example, would have been a lot lower in the 1970s, meaning extreme heat days would have a lower emissions impact back then. Even so, at Columbia University in New York says using machine learning is a good way to 鈥渇ill in鈥� data we would have liked to have collected in the past.

Both Peters and Allan say that even with uncertainties around the precise figures, the general trend towards more extreme temperatures causing higher emissions is clear. 鈥淒espite limitations, the findings do amplify existing knowledge that global warming will make emissions reductions even more challenging,鈥� says Allan.