One of the most striking findings comes from a reconstruction of top-of-atmosphere radiation extending across the last 1,000 years. By combining evidence preserved in tree rings, corals, and ice cores with climate data and modeling, researchers estimated how much energy entered Earth’s system and how much escaped back into space over centuries. Their results indicate that the modern increase in Earth’s energy budget—the imbalance produced primarily by rising greenhouse-gas concentrations—is unprecedented in the reconstructed millennium. When more energy enters the climate system than leaves it, the excess warms the atmosphere, oceans, and land, drives glacier and ice-sheet loss, and intensifies many components of the hydrological cycle. Earlier long-term cooling trends were associated largely with volcanic eruptions, whose particles temporarily reflected sunlight and reduced the amount of solar energy reaching the surface.

The consequences of that energy imbalance are especially visible in forests across the boreal region and the western United States. A comprehensive review reports that the average annual area of forest burned has increased approximately threefold since 2000, accompanied by rapid growth in the number of very large fires and rising smoke exposure across much of North America. The study identifies aridity as the strongest climate-related control on burn area and fire-driven carbon emissions. Warmer air increases evaporative demand, drying vegetation and soils even where total precipitation does not decline sharply. The authors argue that the observed expansion of fire activity is more consistent with climate-driven drying than with changes in fire management alone. The result is a dangerous feedback: fires release carbon dioxide, destroy carbon-storing ecosystems, and produce smoke that can travel thousands of kilometers.

Climate signals are also emerging in tropical cyclone behavior, although the picture is more complex than a simple increase in storm numbers. Large-ensemble climate-model experiments suggest that rising greenhouse-gas concentrations may be contributing to a southward shift and an overall reduction in tropical cyclone frequency over the North Atlantic and eastern North Pacific. The simulations also indicate that aerosols—tiny particles emitted by industrial activity that can alter sunlight, clouds, and atmospheric circulation—would have produced a different pattern if they had remained the dominant forcing since the 1980s. Under that scenario, storms would have become more frequent and shifted northward. The findings highlight why cyclone risk cannot be judged by counts alone: fewer storms do not necessarily mean less danger, because warming oceans and higher sea levels can increase rainfall, storm surge, and the destructive potential of individual events.

Other research shows that climate impacts can travel far beyond the location of environmental change. Modeling experiments focused on deforestation in the Maritime Continent—the tropical region around Indonesia, New Guinea, and neighboring islands—suggest that forest loss can modify atmospheric circulation across the Pacific. The influence appears particularly strong during La Niña, when altered convection over the islands can generate wave-like changes in the jet stream and affect North American temperatures. As global warming changes the frequency and intensity of climate patterns such as El Niño and La Niña, disturbances in one part of the tropics may increasingly produce consequences in distant regions. The study does not imply that deforestation is the only driver of North American warming, but it demonstrates how land-use change can interact with ocean-atmosphere variability to amplify remote climate effects.

The same research collection points toward a new generation of tools designed to make weather warnings more locally useful. A machine-learning system called LiveCyc converts conventional tropical-cyclone forecasts, which often operate at scales of tens of kilometers, into probabilistic wind predictions at roughly neighborhood or kilometer scale. The system incorporates local influences such as terrain and topography, which can accelerate, weaken, or redirect winds around hills, coastlines, and urban areas. Instead of providing a single deterministic wind value, it generates a range of possible outcomes, allowing emergency managers to evaluate uncertainty. Another study examines whether lightning patterns can help identify disturbances that are about to become tropical cyclones. Using observations from the Geostationary Lightning Mapper, researchers classified lightning structures in five recent North Atlantic systems, four of which developed into named storms. Common features among the developing systems suggest that lightning may offer an additional early signal, although the sample is too small to establish a reliable operational forecasting method.

Rainfall trends reveal another divide between global averages and regional realities. Although climate change is generally increasing extreme precipitation worldwide because warmer air can hold more moisture, researchers examining eight arid regions found declining total and extreme precipitation in seven of them between 1980 and 2024. North Africa was the exception, experiencing an increase faster than the global mean. The apparent contradiction reflects the sensitivity of dry climates to temperature and atmospheric moisture. In many arid environments, warming increases evaporation and atmospheric drying even when humidity rises in absolute terms. With dry regions also warming faster than the global average, rainfall losses can have disproportionate effects on groundwater recharge, agriculture, ecosystems, and reservoirs. In Malawi, a separate multi-model study found that river flows may decline soon after meteorological drought begins, leaving little time between a lack of rain and reduced water availability for irrigation. Climate change is expected to intensify that overlap between peak water demand and shrinking supplies.

Water constraints also complicate efforts to use nature as a climate solution. An analysis of revegetation potential in China’s drylands found that planting forests or other vegetation can increase carbon storage, but the strategy may be unsustainable in areas where water is already scarce. Even less water-intensive plants may consume more water than the landscape can afford, reducing supplies for ecosystems and people. In wetter areas with a surplus, however, higher-water-use vegetation may be feasible. Research on China’s national parks offers a more encouraging example of land stewardship: establishing protected areas was associated with lower county-level carbon-emission intensity, partly through forest carbon sequestration and the growth of low-carbon sectors such as ecotourism and cultural services. The findings suggest that conservation can reduce emissions, but only when ecological goals are matched to local water budgets and economic conditions.

The physical changes documented by these studies are increasingly colliding with social vulnerability. Research involving Florida and Louisiana residents found that impact-based storm-surge alerts—warnings that describe consequences such as flooded roads rather than simply reporting water levels—were especially effective for people without previous evacuation or storm-surge experience. Yet focus groups with Spanish-speaking and economically disadvantaged residents in Miami revealed serious difficulties interpreting the National Hurricane Center’s forecast cone. Participants described the graphic as cluttered and struggled with its shape, symbols, colors, warning categories, time periods, legends, language, and measurement conventions. These findings help explain why excellent forecasts can still produce poor outcomes. Communication failures, unclear guidance, inadequate infrastructure, delayed emergency response, and ineffective evacuation systems can all convert accurate predictions into disasters.

The societal challenge extends beyond emergencies. A survey of 6,000 people in the United Kingdom found that 78 percent regarded current numerical weather prediction as accurate, while only 47 percent trusted forecasts produced with machine-learning methods. The gap suggests that technical performance alone will not determine whether artificial-intelligence forecasting is adopted. Public confidence may depend on transparency, explanations of uncertainty, accountability, and clear evidence that new systems improve decisions rather than merely automate them. Other studies in the collection show how climate risk is altering agriculture, sports, and ecosystems: future hail losses in the Great Plains may rise because cropland is expanding into hail-prone areas, while the Midwest and South could face more severe hail even without major land-use change. Chilean outdoor sports are confronting heat stress, wildfire smoke, declining snowpack, glacier retreat, permafrost thaw, altered river flows, storm surges, and coastal erosion. In Brazil, a weakened Atlantic Meridional Overturning Circulation could reshape rainfall seasonality in the Amazon, cool parts of the basin, warm central Brazil, and alter the boundary between rainforest and dry shrubland. The message running through the research is unmistakable: climate change is modifying not only weather, but the conditions under which societies plan, grow food, move, compete, conserve nature, and survive extreme events.

Subject of Research: Climate change, weather extremes, atmospheric circulation, tropical cyclones, wildfires, hydrology, forecasting technology, disaster communication, and climate adaptation

Web References: American Meteorological Society journals: https://journals.ametsoc.org/ ; https://www.ametsoc.org/

References: https://doi.org/10.1175/JCLI-D-25-0568.1 ; https://doi.org/10.1175/BAMS-D-25-0193.1 ; https://doi.org/10.1175/JCLI-D-25-0442.1 ; https://doi.org/10.1175/JCLI-D-24-0512.1 ; https://doi.org/10.1175/BAMS-D-25-0182.1 ; https://doi.org/10.1175/MWR-D-25-0270.1 ; https://doi.org/10.1175/JCLI-D-25-0615.1 ; https://doi.org/10.1175/JHM-D-25-0060.1 ; https://doi.org/10.1175/WCAS-D-25-0215.1 ; https://doi.org/10.1175/WCAS-D-25-0179.1

Keywords: Climate change, global warming, greenhouse gases, Earth’s energy budget, wildfires, drought, precipitation, tropical cyclones, hurricanes, La Niña, deforestation, machine learning, weather forecasting, storm surge, evacuation, hail, water scarcity, hydrology, AMOC, Amazon, climate adaptation

The Holocene epoch, spanning roughly the last 11,700 years following the final retreat of the last Ice Age, has long been considered a period of relatively stable and warm climate. Yet, fossil pollen evidence amassed over decades from various North American locations reveals a far more complex hydrological history. This new research exploits advanced machine learning techniques to analyze these pollen datasets, enabling the team to infer subtle, regional variations in moisture conditions throughout the Holocene. Their findings reveal a persistent deficit in moisture relative to modern levels, punctuated by diverse drought episodes lasting centuries to millennia.

Interestingly, the onset and intensity of drought did not spread evenly across the continent. According to lead investigator J. Sakari Salonen, an Academy of Finland research fellow, dryness first emerged in the northeastern United States and adjacent Canadian regions, traditionally among the wettest zones today. This anomalous early dearth of moisture peaked approximately 11,000 years ago, marking the beginning of a prolonged drought phase in these easternmost areas. Over the subsequent millennia, the drought shifted westward, culminating around 7,000 years ago in the modern prairie regions of the mid-continental United States. At this stage, the Atlantic coast had already begun to revert to wetter conditions, illustrating a migrating climate anomaly rather than a static, continent-wide drought.

Bryan Shuman, co-author of the study from the University of Wyoming, highlights that the severity of these historic droughts was comparable to the infamous Dust Bowl of the 1930s but extended over vastly longer durations. This insight is crucial not only for understanding past ecological transformations, including widespread forest dieback and shifts in fire regimes, but also for anticipating future vulnerabilities. As climate variability intensifies in the coming decades, unraveling the mechanisms controlling historical drought variability becomes imperative for improving societal resilience and resource management strategies.

The team’s reliance on fossil pollen data marks a significant advancement in paleoclimatology. Pollen grains, deposited layer by layer in lake sediments and peat bogs, serve as biological proxies for past vegetation and hence climate conditions. By feeding this rich dataset into sophisticated computational algorithms, including machine learning models, the researchers could reconstruct detailed moisture patterns with unprecedented spatial and temporal resolution. This approach surpasses traditional proxy analysis, providing nuanced insights into the timing, duration, and geographical progression of the Holocene droughts across North America.

To bolster their empirical reconstructions, researchers employed state-of-the-art numerical climate simulations running on supercomputers. These high-resolution models, operating at two to four times the resolution of prior attempts, allowed the team to probe the physical processes behind the reconstructed droughts. Frederik Schenk, atmospheric physicist and visiting scientist at the University of Helsinki, explains that the simulations elucidated two primary mechanisms: first, the persistence and migration of a high-pressure system linked to the massive ice sheets lingering in northern North America during the early Holocene; and second, the onset of widespread drought conditions across the continent as summer temperatures increased following the ice sheet’s disappearance.

The study also draws a striking parallel between past and future climatic conditions. As global temperatures continue to rise due to anthropogenic greenhouse gas emissions, much of North America is projected to experience heightened dryness by the century’s end. This paradox—where rising precipitation fails to counterbalance increasing evaporation due to warming—mirrors the Holocene drought dynamics identified by the researchers. Schenk emphasizes that although overall warming tends to increase global moisture availability, regional thresholds exist beyond which evaporation surpasses precipitation, triggering drought conditions akin to those that unfolded millennia ago.

However, the study carefully notes a fundamental difference between the drought drivers of the Holocene and those shaping today’s climate crisis. The ancient, multi-millennial droughts were precipitated by slow shifts in Earth’s orbital parameters—collectively known as Milankovitch cycles—including variations in axial tilt and orbital eccentricity. These orbital changes modulated the intensity and distribution of solar radiation, leading to progressively warmer summers and thus, drier conditions in eastern North America. In stark contrast, the rapid pace and scale of modern warming are predominantly fueled by human activities, particularly the accumulation of greenhouse gases in the atmosphere.

The research draws on a rich scientific tradition of studying Earth’s orbital influences on climate. For over two million years during the Quaternary period, Milankovitch cycles have governed the timing of glacial and interglacial periods. The peak of the last Ice Age approximately 20,000 years ago corresponded with an orbital configuration that reduced summer sunlight in the northern hemisphere, permitting the build-up of massive ice sheets. By around 10,000 years ago, orbital shifts reversed this pattern, triggering the melting of these ice sheets and ushering in the warmer Holocene interglacial, during which North America underwent significant hydrological transitions as revealed by this study.

The implications of these findings stretch beyond academic curiosity. As Jack Williams of the University of Wisconsin-Madison, another co-author, articulates, public perception in eastern North America often assumes water abundance as a constant. The revelation that the region has historically endured prolonged drought-induced ecosystem upheavals challenges this complacency and underscores the necessity for proactive water management policies grounded in a deep-time perspective. Such historical insights can inform adaptive strategies that better accommodate the growing risks of drought and ecosystem stress under future climate regimes.

Moreover, the use of cutting-edge computational tools—for both data analysis and climate modeling—demonstrates the powerful synergy between paleoclimatic proxy research and numerical simulations. Together, these methodologies enable scientists to transcend the limitations of fragmentary records, constructing cohesive, dynamic narratives of Earth’s climatic past. The increasing resolution and sophistication of climate models are particularly salient, as they reveal subtle circulation patterns and feedbacks that were previously too complex to decipher, thus enabling a transformative understanding of long-term drought drivers.

The study received generous support from several funding agencies, including the Research Council of Finland, the Swedish Research Council for Sustainable Development (FORMAS), the Swedish Research Council (Vetenskapsrådet), and the U.S. National Science Foundation, underscoring the international and interdisciplinary nature of this research endeavor. Led by J. Sakari Salonen, the team’s work exemplifies collaborative science tackling one of the most pressing challenges in climate research: understanding variability and extremes through both natural and anthropogenic lenses.

As the planet warms at an unprecedented rate, the echoes of ancient droughts may yet foreshadow troubling trends. Salonen warns that if current climate projections hold, North America might soon experience a rapid recurrence of the natural drought patterns last seen over ten thousand years ago, but compressed into mere decades. This warning adds urgency to global efforts aimed at mitigating emissions and developing adaptive measures to safeguard water resources and ecological stability in the face of inevitable climatic shifts.

Subject of Research: Not applicable
Article Title: Patterns and drivers of Holocene moisture variability in mid-latitude eastern North America
News Publication Date: 15-Apr-2025
Web References: https://www.nature.com/articles/s41467-025-58685-7
References: Salonen, J.S., Schenk, F., Williams, J.W. et al. Patterns and drivers of Holocene moisture variability in mid-latitude eastern North America. Nat Commun 16, 3582 (2025). DOI: 10.1038/s41467-025-58685-7
Keywords: Holocene drought, North America, climate variability, Milankovitch cycles, fossil pollen analysis, machine learning, climate modeling, Earth’s orbit, anthropogenic climate change, moisture reconstruction, paleoclimate, ecosystem transformation

The ENSO phenomenon, characterized by periodic fluctuations in sea surface temperatures and atmospheric pressure in the equatorial Pacific Ocean, is the most influential mode of interannual climate variability. Modern ENSO events have profound impacts on global weather patterns, ecosystems, and economies. However, its dynamics under drastically different climate regimes remained elusive until now. By leveraging novel paleoclimate proxies and advanced climate modeling, Abhik, Dommenget, McGregor, and their collaborators present a detailed reconstruction and analysis of ENSO’s behavior during the Early Eocene.

This study situates itself within a broader effort to decipher Earth’s climatic past to better anticipate future changes. The Early Eocene represents a “hothouse” period, marked by elevated atmospheric CO2 concentrations and global temperatures surpassing current averages by several degrees Celsius. Existing paleoclimate evidence suggests that oceanic and atmospheric circulation patterns during this era diverged notably from today’s, raising questions about the intensity, frequency, and duration of ENSO events amidst such conditions.

The research team combined empirical data gleaned from sediment cores, marine fossils, and isotopic markers with state-of-the-art climate models. These models were meticulously tuned to replicate Early Eocene boundary conditions, including paleogeography, greenhouse gas concentrations, and solar insolation. The integration of proxy data with simulations allowed the scientists to discern ENSO characteristics distinct from the modern Pacific climate oscillations.

One of the principal findings highlights that Early Eocene ENSO events were not only stronger in amplitude but also exhibited a prolonged duration, sometimes persisting for multiple years. This is in contrast to the typically episodic nature of contemporary ENSO cycles, which commonly span 9 to 12 months. The extended ENSO phases would have significantly modulated global climate patterns, intensifying and extending periods of drought and rainfall across various continental regions.

Mechanistically, the study attributes these changes to altered ocean-atmosphere feedbacks under higher baseline temperatures. The enhanced warming of the tropics intensified the thermal gradient between the western and eastern Pacific Ocean, thereby amplifying the oceanic wave responses and atmospheric convection patterns central to ENSO dynamics. Additionally, Early Eocene ocean stratification and altered thermocline structures played a pivotal role in modulating ENSO behavior.

Particularly intriguing is the implication that the Early Eocene’s prolonged warm episodes may have sustained ENSO events, propagating their climatic influence over extended timescales. This result challenges prevailing assumptions that warmer climates would dampen ENSO variability. Instead, it appears that under elevated greenhouse gas conditions, ENSO could become a more dominant driver of climate variability, with far-reaching consequences for biospheric and geospheric systems.

The study also probes the potential feedback mechanisms linking ENSO with global carbon cycles during the Early Eocene. Stronger and longer ENSO events could have influenced oceanic carbon uptake and release, modulating atmospheric CO2 concentrations and climate feedback loops. This introduces a complex interplay between orbital forcing, Internal climate variability, and biogeochemical processes that governed Earth’s past climate evolution.

Furthermore, the patterns revealed in this research provide a tangible analog for the future, as modern anthropogenic greenhouse gas emissions push global temperatures into uncharted territory. Understanding the response of ENSO—the planet’s most significant climate oscillation—to warmer climates aids in forecasting potential changes to modern weather extremes. Amplified ENSO events in the future could exacerbate droughts, floods, and heatwaves worldwide, posing unprecedented risks to human societies and ecosystems.

Technically, the team’s climate models incorporated coupled atmosphere-ocean general circulation models (AOGCMs) with high spatial resolution and sophisticated physical parameterizations to faithfully simulate Early Eocene climate dynamics. Calibration against proxy reconstructions ensured model fidelity, enabling the isolation of ENSO signals from broader climatic noise. The researchers deployed spectral analysis and statistical methods to quantify ENSO amplitude, frequency, and persistence.

The robustness of the findings stems from the convergence of multiple lines of evidence and the application of rigorous sensitivity analyses. For instance, variations in greenhouse gas forcing, paleogeographic reconstructions, and oceanic nutrient cycles were tested independently to assess their influence on ENSO characteristics. The consistency across simulations and proxies strengthens the conclusion that Early Eocene ENSO was indeed distinctively intensified and prolonged.

This work underscores the importance of paleoclimatology as a vital tool for decoding Earth’s climate system responses under extreme conditions. By probing a deep-time interval with heightened global warmth, the research provides a natural laboratory for exploring how fundamental climate modes, like ENSO, adapt or transform. Such insights become increasingly relevant in light of ongoing climate change and its anticipated impacts on global weather variability.

In essence, this study marks a significant leap forward in paleoclimate research, revealing that the ENSO phenomenon was not merely present but enhanced and extended during a greenhouse world. This reshapes our understanding of past climate variability and compels us to rethink how future ENSO dynamics might evolve. The interplay between oceanic processes, atmospheric circulation, and carbon-climate feedbacks identified here opens new avenues for multidisciplinary exploration.

The findings could also influence the way researchers interpret paleoclimate records from other epochs, as ENSO variability imprints distinct signatures on sedimentation patterns, isotopic compositions, and terrestrial ecosystems. Recognizing stronger and longer ENSO phases in ancient climates enhances the resolution of these reconstructions, allowing for improved correlations between climatic events and their geological archives.

Moreover, this enhanced ENSO model can inform more accurate and reliable predictions of climate extremes, helping policymakers and planners anticipate challenges linked to water resources, agriculture, and disaster preparedness in a warming world. The study’s implications extend beyond academic curiosity, touching on practical considerations for building climate resilience.

In summary, the investigation into Early Eocene ENSO dynamics reveals that global warming in Earth’s deep past amplified the strength and persistence of this critical climate oscillation. This groundbreaking research bridges deep-time climate science with contemporary climate challenges, illuminating pathways for understanding and managing the intensifying impacts of a changing world’s most powerful climatic driver.

Subject of Research: Early Eocene El Niño-Southern Oscillation dynamics under greenhouse climate conditions.

Article Title: Stronger and prolonged El Niño-Southern Oscillation in the Early Eocene warmth.

Article References:
Abhik, S., Dommenget, D., McGregor, S. et al. Stronger and prolonged El Niño-Southern Oscillation in the Early Eocene warmth. Nat Commun 16, 4053 (2025). https://doi.org/10.1038/s41467-025-59263-7

Image Credits: AI Generated