Ancient Warm Period Reveals the Hidden Forces Behind India’s Monsoon Swings

Some 400,000 years ago, the Indian subcontinent was cloaked in lush tropical forest, sustained by a summer monsoon far stronger than anything the region experiences today. Then, in a matter of roughly 1,600 years, that forest collapsed by more than 20 percentage points, replaced by grasses and drought-tolerant shrubs, before rebounding just as dramatically. This extraordinary episode, buried in seafloor mud from the Bay of Bengal, is now telling scientists something unsettling about the future: even under globally warm conditions remarkably similar to our own, the monsoon can fail catastrophically.

The discovery comes from a new study published in the journal Climate of the Past, in which an international team of researchers led by Dulce Oliveira of the University of the Algarve and the University of Bordeaux reconstructed vegetation and monsoon variability across Marine Isotope Stage 11, or MIS 11, an interglacial period spanning roughly 426,000 to 374,000 years ago. MIS 11 has long fascinated paleoclimatologists because its orbital configuration resembles that of the Holocene, the current interglacial, and because it combined prolonged carbon dioxide-driven warming, higher-than-present sea levels linked to polar ice-sheet loss, and an unusually long duration. If scientists want a preview of how tropical ecosystems and the Indian summer monsoon might behave in a warmer world, MIS 11 is one of the best natural laboratories available.

To read that ancient landscape, the team turned to pollen grains preserved in sediments from IODP Site U1446, a core drilled during Expedition 353 on the northeastern Indian margin, offshore the Mahanadi River basin. The site was strategically positioned to capture rainfall signals from India’s Core Monsoon Zone, the central belt of the peninsula that receives 80 to 90 percent of its annual rainfall during the summer monsoon months of June through September. Because the narrow continental shelf near the river mouth efficiently delivers terrestrial material to the deep sea, the pollen extracted from the marine sediments provides an integrated picture of the vegetation across the watershed, from semi-evergreen coastal forests to dry deciduous savannas. The researchers analyzed 62 samples spanning late MIS 12 through early MIS 10, achieving an average temporal resolution of about 1,200 years, and grouped the pollen into ecologically meaningful categories, with moisture-loving wet evergreen forest taxa such as Elaeocarpus, Gnetum, Mallotus and Syzygium serving as the primary indicator of monsoon rainfall intensity.

The record reveals a striking transformation at the MIS 12 to MIS 11 transition. Semi-arid steppe, dominated by grasses and xerophytic plants like Amaranthaceae, gave way to tropical forest as the percentage of arboreal taxa surged from about 2 percent to 58 percent, a response the authors attribute to a steep northward shift of the Intertropical Convergence Zone and a rapid intensification of monsoon rainfall. During the interglacial substage MIS 11c, wet evergreen forest flourished for roughly 32,000 years, peaking around 414,000 to 407,000 years ago, when the monsoon reached its maximum intensity. Notably, the forest record follows an asymmetric M-shaped pattern with two peaks, mirroring similar structures seen in Antarctic methane concentrations and southern European pollen sequences, suggesting that shifts in the ITCZ’s mean position linked low- and mid-latitude hydroclimate and left a common imprint on tropical wetland methane emissions.

But the most dramatic moment in the record is not an expansion. Around 406,000 years ago, at the height of full interglacial conditions, the tropical forest abruptly contracted by 22.3 percentage points in roughly 1,600 years, with no subsequent recovery to previous levels during the optimum. The researchers link this event, labeled Fe-1, to a millennial-scale weakening of the summer monsoon, and they note a striking parallel: an abrupt cold and dry event of the same age is documented in marine pollen records off southern Europe, where it similarly terminated the interglacial optimum. Because the event occurred when ice volume was low and carbon dioxide concentrations were high and stable, the authors propose that a prolonged or more frequent El Niño-like state, possibly connected to North Atlantic circulation changes, may have weakened the monsoon and desiccated both regions simultaneously.

The second half of MIS 11 tells a different story. As ice sheets grew and atmospheric carbon dioxide declined during substages MIS 11b and MIS 11a, the forest record diverged from insolation trends and instead tracked carbon dioxide and sea-level reconstructions. Three further abrupt forest contractions, centered near 387,700, 383,300 and 374,600 years ago, coincide within dating uncertainties with North Atlantic cooling events and disruptions of the Atlantic meridional overturning circulation, the great ocean conveyor that redistributes heat across the hemispheres. The mechanism is well understood from younger records: when massive ice sheets discharged icebergs into the North Atlantic, the overturning circulation weakened, the ITCZ shifted southward, and the Indian summer monsoon faltered. Each contraction was rapidly followed, within 600 to 1,000 years, by a sharp forest rebound, coinciding with Dansgaard-Oeschger-like warming events, abrupt carbon dioxide jumps and methane overshoots recorded in Antarctic ice, all pointing to a sudden reinvigoration of ocean circulation and a poleward ITCZ shift that swelled tropical wetlands.

To disentangle the competing drivers, the team compared their pollen record with transient climate simulations performed with the LOVECLIM Earth system model, run with time-varying insolation and carbon dioxide over two precessional cycles. The comparison confirmed that during MIS 11c, the pattern and amplitude of the wet evergreen forest closely tracked precession-paced boreal summer insolation, peaking nearly in phase with the strongest insolation maximum while carbon dioxide remained high and sea level continued to rise. In warm climates with elevated carbon dioxide and reduced ice volume, insolation is the dominant control on the monsoon, acting through the thermal contrast between the Indian peninsula and the equatorial Indian Ocean. Conversely, during the glacial inception, expanding ice sheets and falling carbon dioxide overshadowed insolation, favoring carbon-hungry C4 grasses and reducing atmospheric moisture transport to the peninsula. The mismatch between data and model during late MIS 11, where the simulations continued to follow insolation, highlights a critical gap: models lacking dynamic ice sheets cannot capture the millennial-scale variability that dominated the record during that interval.

The findings carry a pointed message for the twenty-first century. Projections consistently indicate an intensifying Indian summer monsoon under continued warming, with the potential for denser vegetation and deciduous trees replacing grasslands across the peninsula. The MIS 11 simulations suggest that central India was considerably wetter and more forested at that interglacial’s climate optimum than during the pre-industrial period, thanks to much higher boreal summer insolation, offering a possible analogue for the greening expected under future scenarios. Yet the abrupt forest collapse at 406,000 years ago demonstrates that extreme monsoon weakening can strike even under globally warm, high-carbon-dioxide conditions, precisely the state the planet is entering now. The tropical forests of India are a key component of both the global carbon and methane cycles, and their sensitivity to sudden hydroclimatic shocks has now been demonstrated for a warm world.

What makes the study especially significant is that it fills a long-standing gap in the monsoon archive. While East Asian summer monsoon variability during MIS 11 has been documented in Chinese speleothems, records from India’s Core Monsoon Zone older than the Holocene were previously limited to lower-resolution marine sequences focused on glacial-interglacial cycles. By resolving millennial-scale events within a single interglacial, the Site U1446 pollen record shows that the Indian and East Asian monsoon subsystems responded heterogeneously to the same forcings, and that neither a purely insolation-driven nor a purely ice-volume-driven explanation suffices. Instead, the dominant control shifts with the baseline climate state, and abrupt, high-latitude-triggered events can override the orbital pacemaker altogether. As the authors note, future research should prioritize the atmospheric circulation patterns, particularly El Niño-like dynamics, that can synchronize drought across low and mid-latitudes, because understanding those teleconnections may prove essential for anticipating how the monsoon and the forests it sustains will behave as the planet continues to warm.

Subject of Research: Indian summer monsoon and tropical vegetation variability during Marine Isotope Stage 11

Article Title: Complex interplay of forcings drives Indian vegetation and summer monsoon variability during MIS 11

Article References: Oliveira, D., Desprat, S., Yin, Q., Zorzi, C., Wu, Z., Anupama, K., Prasad, S., Alonso-García, M., & Martinez, P. (2026). Complex interplay of forcings drives Indian vegetation and summer monsoon variability during MIS 11. Climate of the Past, 22(9), 1691-1709. https://doi.org/10.5194/cp-22-1691-2026

Image Credits: AI Generated

DOI: 10.5194/cp-22-1691-2026

Keywords: Indian summer monsoon, MIS 11, paleoclimate, pollen analysis, IODP Site U1446, Intertropical Convergence Zone, tropical forest, AMOC, insolation, carbon dioxide, Bay of Bengal, interglacial