CISL-powered climate analysis of Southeast Asia reveals fascinating historical insights

Note: this story updates a previous article first published on February 26, 2025.
 

"We can now take a bird's-eye view of our planet, and see connections we couldn't see with our own two eyes.” —Shawn (Shouyi) Wang, lead researcher

 

A recent study, entitled "Quantifying the Internal and External Drivers of Southeast Asian Rainfall Extremes on Decadal Timescales" and published in the September 2024 issue of Climate Dynamics, explores whether extreme rainfall changes in Southeast Asia are primarily driven by external drivers or by internal climate variability in the Pacific and Indian Oceans. These extreme rainfall changes refer to both droughts and pluvials—that is, extremely dry or wet rainfall events. External drivers refer to events such as volcanic eruptions, while internal variability indicates natural fluctuations in ocean temperatures and atmospheric patterns. 

To conduct this research, the authors—a team of scientists from four institutions—leveraged NSF NCAR computing and data resources. The research was led by Shawn (Shouyi) Wang of the Woods Hole Oceanographic Institution (WHOI) and the MIT–WHOI Joint Program in Oceanography, as well as WHOI’s Caroline Ummenhofer (both pictured). The team used a Data Analysis project allocation from CISL to access NSF NCAR-curated datasets, as well as CISL's Casper analysis cluster to investigate climate patterns in Southeast Asia. Stated lead author Wang: 

“NSF NCAR and CISL have provided really great resources for the community. There’s so much data and so many resources publicly available.”

Wang explained that the research emerged from collaboration with dendrochronologists, or tree ring scientists, from the Lamont-Doherty Earth Observatory at Columbia University. The paper’s co-authors used the width of tree rings to infer historical soil moisture and rainfall and to reconstruct droughts and extreme rainfall periods in the past. However, Wang added, “there's only so much you can tell from these individual tree ring records from the past. They can give you individual point records of change, but to get the full picture, a climate model is really valuable. And that's where the Community Earth System Model Last Millennium Ensemble (CESM-LME) model really came into use.” 

He continued: “We dove in and tried to isolate the periods in the climate model simulations that looked a lot like these really extreme historical droughts that we know happened. Using the model, we explored what the atmosphere and oceans were doing during this period to try and establish these relationships.”

Southeast Asia, a region dependent on monsoon rains, has historically experienced extreme shifts in rainfall. To further understand these shifts, the authors examined both effects stemming from internal climate factors—that is, natural climate variations, such as El Niño and La Niña events—and effects originating from external forcing factors, such as volcanic eruptions and changes in the sun’s irradiance. 

 

“What motivated us to pursue this was trying to understand—to really pin down—what sort of mechanisms led to these prolonged periods of droughts or enhanced rainfall, which have affected societies and people in really significant ways,” said Wang. 

 

“Climate and society are very much intertwined. Water, habitability, and agriculture are important aspects of society. When they change, they put strain on a society, an environment, and its people. There's no reason to suspect climate and society wouldn't influence each other in the future as well." 

"A period of drought that lasts for 10 years can certainly stress and affect a society's ability to exist in the way that it has been, as we've seen in this study. By better understanding the climate system and its drivers, we can achieve better prediction or attribution of certain big societal changes,” he continued.

 

The authors highlight that several of the lengthy wet and dry interludes identified here coincide with periods associated with “societal change and civil unrest throughout Asia” in previous studies, spotlighting the following historical events within the context of climate:

• In the late 14th to 15th century, droughts, punctuated by severe flooding, contributed to Angkor Wat’s collapse as the capital city of the Khmer empire.
• The prolonged Ming Dynasty Drought from 1637–1643 (which coincided with a 1641 volcanic eruption) played a major role in the fall of the Ming Dynasty in 1644.
• In 1756–1768, the unprecedented Strange Parallels Drought hit numerous societies hard in South and East Asia. The authors partially attribute this drought to a 1761 volcanic eruption.
   

The research team used the CESM-LME to simulate climate over the last thousand years. The dataset is developed and made available by the CESM Paleoclimate Working Group, and is housed within CISL. The CESM-LME allowed the team to differentiate between internal and external climate factors. According to the authors:

 

“The CESM-LME consists of multiple realizations of global climate under identical external forcings which allows for the separation of internal and external drivers... Additionally, each LME simulation spans over 1000 model years (850–2005 CE), thus providing a sufficiently large sample size for robust probabilistic analysis.”

 

“NSF NCAR and CISL produced this dataset in 2016. We're still writing papers and discovering new things from it in 2025. It's been nine years, and there's still new discoveries being made.” 

The CESM-LME simulations showed that in Southeast Asia, most droughts and pluvials are primarily driven by internal climate variability, meaning natural fluctuations in ocean temperatures and atmospheric patterns, especially in the Pacific and Indian Oceans. These internal variations affect how air circulates and moisture is transported from ocean to land as part of the monsoons, leading to periods of drought or excessive rain.

 

However, the study also found that external factors, particularly volcanic eruptions, can play a role in some extreme rainfall events, especially droughts. Volcanic eruptions release particles into the atmosphere that can cool the planet and disrupt rainfall patterns. The simulations suggested that volcanic activity may have contributed to some historical droughts, like the one that coincided with the fall of the Ming Dynasty in China. Interestingly, droughts seemed more susceptible than periods of excessive rain to external forcings.

“One thing that our study brings as a novelty is this ability to try to quantify different types of drivers statistically. We looked at droughts and said, quantitatively, what are the contributions from the Pacific Ocean, the Indian Ocean, volcanoes? We're able to put numbers to some of these relationships that people have qualitatively found in the past. I think that for many scientists and many stakeholders in the future, being able to quantify these relationships is important for predictions.” 

Furthermore, these relationships could have relevant real-world applications—for example, their findings provide the first step in potentially being able to predict droughts and floods in Southeast Asia over the next decade: “We can plug in the states of the Pacific Ocean, the Indian Ocean, the atmosphere, solar forcings, volcanic eruptions. We can use our framework to find the percent probability that we might be in a drought in the next ten years. We're really able to start quantifying some of these relationships.”

Regarding the 1756–1768 Strange Parallels drought—so named because it affected numerous areas across the Asian monsoon region—the team’s research has contributed to the ongoing scientific discussion about its causes. Using the model, the team was able to reproduce the Strange Parallels drought. Based on the climate model, they argue that it is partially attributable to volcanic eruptions.

“Paleoclimatologists, archaeologists, and historians have researched and found evidence of the Strange Parallels drought. It's very interesting—the climate model also produced a drought during that same time period. We are perhaps the first to provide this context, that maybe it was partially caused by something external." 

Wang also spoke to another area of debate within the scientific community: whether the Angkor droughts were caused by internal climate variations or external forces like volcanoes or solar activity. The team’s model did not show significant drought periods during the 13th and 14th centuries, so the paper suggests that these droughts might have been caused by random, internal climate variations. However, not all scientists agree with that perspective, with some in the geological community believing that large external forcings were responsible for the droughts. “It's still an ongoing discussion between scientists like myself and scientists that are out there doing the fieldwork and collecting the data,” said Wang. “To my mind, it's always been an interesting conversation when two different fields try and address the same question. When they agree I think that's really awesome, and when they disagree I think it's also very interesting.” 

He added: “I'm always in awe of these scientists that go out in the field and do the hard work. Once they collect the data, then I get to play with it and analyze it and compare it with our climate models. I have it easy compared to all the hard work that they do.”

Wang emphasized that internal factors and external forces constantly interact to influence weather extremes. “It's usually never just one or purely the other, right? For example, if a volcano erupts while the Pacific Ocean is in a very warm state, then you're going to have a bigger drought than if the Pacific Ocean was in a colder state. These things matter, and how they interact with each other has big implications for rainfall.”

In another key finding, the study revealed asymmetries in the drivers of rainfall extremes: while both the Pacific and Indian Oceans influence enhanced rainfall, the Indian Ocean's impact on droughts is less significant due to counteracting external factors like volcanic eruptions, which can interact with ocean patterns to either intensify or dampen their effects on rainfall. Whether natural patterns and external forces amplify or counteract each other “depends on which ocean basin you're looking at and which external forcing you're looking at as well.”

Understanding these different influences, a feat made possible by CISL data science resources, is crucial for predicting future changes in rainfall and managing water resources in this region. Wang also underscored the importance of understanding climate patterns on decadal timescales (10 to 50 years) for informing long-term policy decisions related to infrastructure development and water resource management: 
 

“Ten to 50 years is the timescale in which people live their lives and build communities and build infrastructure. So I think that timescale is very important." 

"You usually think about weather and climate maybe ten days, maybe a year in the future. But if you're interested in building a dam or a reservoir, knowing what the water levels or rainfall is going to be for the next 50 years is more important than knowing what the rain's going to be over the next day.”

Wang also commented on the exciting, treasure hunt-like nature of exploring the CESM-LME data: “You never know what you're going to find. It looks overwhelming, like this giant collection of ones and zeros, but it's all about sifting through it and looking for the connections. NSF NCAR and CISL produced this dataset in 2016. We're still writing papers and discovering new things from it in 2025. It's been nine years, and there's still new discoveries being made. This just goes to show that even in a climate model, which is an imperfect representation of our true world, we're still mining new discoveries.” 

He talked further about why he is so passionate about statistical modeling and Earth system science research: “The coolest aspect about my research is that I get to think about the Earth and the climate system which is so complicated and so difficult to disentangle. I get to play around with data at my fingertips using a computer and screen in front of me. We can look at what's happening across the entire Pacific Ocean and these big patterns and how they affect rainfall and winds. We can make analyses and learn relationships in ways we couldn't do with our own two eyes. We're taking a bird's-eye view on our planet and getting to look at how it's connected. It's just really exciting.”

Finally, Wang overviewed his current research, which focuses on an emerging field: the Indonesian Throughflow, which connects two ocean basins. “It’s very interesting from a climate perspective, because it means that the Pacific and Indian oceans are actually not isolated from each other, that ocean signals can actually connect. That has sparked a lot of interest in our community. Can we use our understanding of how they're connected to better predict climate in these two basins?” 

Wang further indicated that he is interested in using the climate models produced by NSF NCAR and CISL to investigate some of these questions.

Wang, S., Ummenhofer, C.C., Murty, S.A. et al. Quantifying the internal and external drivers of Southeast Asian rainfall extremes on decadal timescales. Clim Dyn 62, 9821–9841 (2024). https://doi.org/10.1007/s00382-024-07412-x.

This study was supported by the U.S. National Science Foundation under AGS-2002083, AGS-2001949, OCE-2303513, and AGS-2302668.

 

Have a topic? Contact us. CISL News is seeking recently-published research using CISL resources. Email shiraf@ucar.edu.