Around 55 million years ago, Earth experienced one of the warmest climates in its history. This period, known as the early Eocene, was marked by atmospheric carbon dioxide concentrations several times higher than today, no permanent ice sheets at the poles and global temperatures comparable to those projected for the coming centuries.
While this ancient world may seem distant, it offers a valuable natural experiment for understanding how Earth’s climate system responds under extreme warming.
In our recent study, published in the “Paleoceanography and Paleoclimatology,” we examined how monsoon systems acted during this ancient hothouse climate. By analyzing simulations from five state-of-the-art Earth system models participating in the Deep-time Model Intercomparison Project (DeepMIP), we explored whether monsoons existed during the early Eocene and, if so, how strong they were.
For decades, scientists have debated whether monsoon circulation could exist in the early Eocene. At that time, Earth’s geography was very different. The Indian subcontinent was still drifting northward and had not yet collided with Eurasia, meaning the Himalayas and Tibetan Plateau, features often considered central to the modern South Asian monsoon, had not yet formed. Despite these differences, our simulations consistently show the presence of monsoon-like circulation over the Indian Ocean. Seasonal wind reversals and organized low-level flows indicate that a proto-monsoon system did operate during the Eocene.
However, the most striking result is not the presence of an ancient monsoon, but it is its weakness. Compared to today’s monsoon, the Eocene systems were dramatically weaker, with circulation strength nearly three times lower than present day values. This weakness was especially clear in the monsoon low-level jet, the fast-moving winds near the surface that play a central role in transporting moisture from the ocean onto land.
In today’s climate, stronger contrasts between land and sea temperatures typically intensify monsoon winds. During summer, the land warms faster than the ocean, creating pressure gradients that drive strong low-level winds and enhance moisture transport. During the Eocene, temperature contrasts were indeed strong and in some cases even stronger than today. But paradoxically, the low-level jet weakened as atmospheric CO₂ increased. This finding challenges a simple assumption often made about monsoons that warmer climates automatically lead to stronger winds. Our results show that this is not always the case.
The explanation lies in how warming affects the vertical structure of the atmosphere. As CO₂ concentrations rise, the atmosphere can hold more moisture. When this moisture condenses, it releases latent heat, warming the upper troposphere more than the surface. This process increases atmospheric static stability, a measure of how resistant the atmosphere is to vertical motion. In a more stable atmosphere, rising air is suppressed, convection weakens and large-scale overturning circulation slows down. Our simulations show clear evidence of this process during the Eocene, with reduced vertical temperature gradients, weaker ascent near the equator and a slowdown of circulation. This weakened overturning limits the pressure-gradient force needed to sustain strong low-level winds. In effect, the atmosphere becomes warmer and moister, but dynamically less energetic.
Geography also shaped the structure of the Eocene monsoon. Although the Himalayas were not yet present, the models reveal low-level jets forming along other topographic features, particularly the Eastern African Rift and the Deccan Plateau in India. These ancient jets followed the contours of paleotopography, confirming that surface features influence monsoon circulation even in very different climatic and geographic settings. However, these ancient jets were shorter, weaker and less extensive than the modern monsoon jet. This highlights the point that while geography helps organize circulation, it does not override the fundamental role of atmospheric physics.
The early Eocene is not a direct analogue for today’s world. Continents, oceans and topography were very different. Most climate models project increased monsoon rainfall under future warming, largely because warmer air can hold more moisture. But increased rainfall potential does not necessarily imply stronger monsoon winds. Our results show that enhanced upper-tropospheric warming and increased stability can counteract the forces that typically strengthen circulation. Changes in the strength and position of the low-level jet affect moisture transport, surface cooling and regional climate variability. Observations already show shifts in monsoon wind patterns and warming of the Arabian Sea, raising questions about how circulation and rainfall will evolve together in the future.
These findings carry important implications for East Asian, including the Korean Peninsula. Summer rainfall is strongly influenced by large-scale monsoon circulation and low-level moisture transport from the surrounding oceans. If future warming leads to increased atmospheric stability similar to what we observe in the Eocene simulations, monsoon winds over East Asia could weaken even as the atmosphere becomes warmer and more humid. This may contribute to more variable rainfall, with a higher risk of short, intense downpours separated by longer dry spells, a pattern that has already begun to emerge in recent decades. Understanding how circulation and moisture respond differently to warming is therefore critical for anticipating future changes in East Asian climates.
The early Eocene teaches us that Earth can be warmer, wetter and yet dynamically weaker. Monsoon systems emerge from a delicate balance between thermodynamics and atmospheric dynamics. When that balance shifts, familiar climate patterns can behave in unexpected ways.
By studying ancient climates, we gain insight into processes that unfold slowly and subtly, but with far-reaching consequences. As humanity pushes the climate system toward conditions not seen for tens of millions of years, understanding these responses becomes increasingly important. The past does not provide a direct forecast of the future. But it does reveal what is physically possible and sometimes, those possibilities are counterintuitive.
Pratik Kad is a climate scientist based in Norway and the author of the study published in “Paleoceanography and Paleoclimatology.” Ha Kyung-ja is a professor of atmospheric sciences at the Pusan National University in South Korea. She is an expert on monsoon dynamics, and a co-author of the research.
