An early warning signal for the loss of plants in the Amazon rainforest can be seen in the temperature difference between seasons, according to the latest climate models.
The sixth generation of these climate models, which were used to inform the 2021 climate reports from the Intergovernmental Panel on Climate Change, incorporate a wide range of data and environmental mechanisms. They are the first to include dynamic vegetation, which looks at how plant populations change and interact over time. Predicting forest dieback, where rainforest turns to savannah, is important because it could indicate when large volumes of carbon dioxide will be released into the atmosphere.
Isobel Parry at the University of Exeter, UK, and her colleagues used six of these models to predict how the Amazon will change in response to rising carbon dioxide levels and a warming climate. As well as finding evidence of forest dieback events in north-western Brazil, southern Colombia and northern Peru, Parry and her team noticed that a strong indicator of dieback events was when the gap between the highest and lowest annual temperatures in a certain region grew larger.
Advertisement
鈥淭he difference between those two [temperatures] is increasing before an abrupt shift [in vegetation],鈥� says Parry. 鈥淭his is significantly different from the sort of changes that you see at points that don鈥檛 experience an abrupt shift, so it鈥檚 a promising early warning signal.鈥�
鈥淚f you鈥檙e starting to see the difference between the seasonal temperature difference between seasons, you may be at the edge of the cliff of an abrupt shift,鈥� says at the University of S茫o Paulo, Brazil.
Given that the models don鈥檛 take into account human-made drivers of deforestation, such as forest fires and land use, the levels of dieback could be even greater, says Nobre.
The signal could be particularly useful for monitoring areas of the Amazon that currently have good forest cover, such as the north-eastern Amazon and regions of Suriname and Guyana, he says.
Some of the climate models that Parry and her team used also predicted localised dieback events, where certain areas in the northern region of South America appeared to show forest death. This is the first time that these climate models have predicted dieback events like these, says Parry.
Observations from satellite data have already shown the Amazon becoming less green over the past two decades, which could be a sign that the region has become less able to restore itself to a stable state after being affected by events like wildfire or drought.
Reference:
Sign up to our free Fix the Planet newsletter to get a dose of climate optimism delivered straight to your inbox, every Thursday
