A ground-breaking study combining machine-learning-based hydroclimate reconstructions with new high-resolution climate simulations has resolved a long-standing scientific issue on the causes of millennial-long megadroughts in Northeast America. The research, conducted by an international team of scientists, confirms that warming summer temperatures and evaporation—not changes in precipitation—were the primary drivers of past droughts.

For over two decades, climate models and geological reconstructions have disagreed on past hydroclimatic trends in the region. However, the new study, published in Nature Communications, demonstrates that new climate models can now accurately reproduce past moisture variability over the last 12,000 years. Two of the three climate simulations were performed at the Bolin Centre for Climate Research, Stockholm University, while a third simulation was performed at the Max-Planck Institute for Meteorology in Hamburg, Germany, which confirmed the findings.

“The new climate model simulations are a clear breakthrough,” says co-author Frederik Schenk, climate modeller and co-lead for the Research Theme “Past Climates” at the Bolin Centre for Climate Research, Stockholm University.

For over 20 years, climate simulations and reconstructions fundamentally disagreed on past hydroclimatic long-term trends for Northeast America. Hence, we could not identify the main drivers between the wet periods before ~11,000 years ago and in the last ~2,000 years and the millennial-long megadroughts in between,” Schenk explains.

The study reveals that the unusually wet early Holocene was driven by colder summers and persistent high-pressure systems over the ancient Laurentide Ice Sheet, which directed moisture into today’s dry prairies. As the ice sheet melted, rapidly warming summers—caused by changes in Earth’s orbit—led to strong evaporation that exceeded annual rainfall, triggering megadroughts. The last ~2,000 years saw a return to wetter conditions due to cooling summers and increased precipitation.

The findings have significant implications for future climate projections. “This is a concerning result as there seems to be no efficient enough mechanisms to increase rain amounts to compensate for the warming-driven evaporation,” Schenk explains. “This supports the IPCC results projecting a drying for Northeast America—again driven by warming temperatures.”

With ongoing global warming, evaporation is rising rapidly, now driven by greenhouse gas emissions rather than natural orbital changes. The study suggests that future warming could lead to a return of megadroughts and an expansion of the prairies unless climate change is mitigated.

Read the press release from the University of Helsinki: Waxing and waning prairie: new study unravels causes of ancient climate changes | University of Helsinki

The study was published in the Journal Nature Communications: Salonen, J.S., Schenk, F., Williams, J.W., Shuman, B., Lindroth Dauner, A.L., Wagner, S., Jungclaus, J., Zhang, Q. & Luoto, M.: Patterns and drivers of Holocene moisture variability in mid-latitude eastern North America | Nature Communications

Contact Information:

• Sakari Salonen, Docent, PhD, Academy of Finland Research Fellow, Paleoclimatology Department of Geosciences and Geography, University of Helsinki
  sakari.salonen@helsinki.fi | +358 50 530 5542
• Frederik Schenk, PhD, Research Fellow for Paleoclimate Modelling & Statistics and Research Area Leader for “Past Climates” at the Bolin Centre for Climate Research, Stockholm University
  frederik.schenk@geo.su.se | +46 8 16 47 41

Photo: Private

Climate Modelling at the Bolin Centre
Two of the three climate simulations of this study were performed by Qiong Zhang (Department of Physical Geography) and Frederik Schenk (Department of Geological Sciences) from Stockholm University on the supercomputer Tetralith (picture). The supercomputer forms part of the Swedish National Infrastructure for Computing (SNIC) and National Academic Infrastructure for Supercomputing in Sweden (NAISS) at the National Supercomputer Centre (NSC). It is partially funded by the Swedish Research Council institutional grants 2018-05973 and 2022-06725 to the Bolin Centre for Climate Research. Resources | NAISS

Climate modelling of past climates
To simulate past climates, the same climate models are used as for the present or for future climate projections like in the IPCC reports. In this way, climate models used for the future can be compared and validated against climate reconstructions in the past. This is a crucial part of climate model validations: Climate models must demonstrate that they produce correct results for climates that are very different from today, e.g., much colder or warmer than the 20th century. Paleoclimate modelling is one of the core tasks of the Bolin Centre as part of the basic research goals to better understand climate change on different timescales and improve and validate climate models used for IPCC projections.