Using an innovative machine model, researchers at Dartmouth postulated that the mass extinction marking the end of the dinosaur era was predominantly triggered by volcanic activity rather than an asteroid impact. This pioneering approach opens up new avenues for investigating various geological phenomena.
By employing forward-thinking laptops, scientists reverse-engineered the fossil record to unravel the causes of a catastrophe.
Addressing the longstanding debate over whether the demise of dinosaurs and numerous other species was instigated by a massive asteroid collision or geological processes.
A species constitutes a group of organisms capable of reproduction and producing viable offspring, sharing a set of distinct characteristics. In scientific terms, the concept of species plays a vital role in organizing and classifying the diversity of life. While there are various definitions of a species, the prevailing natural species theory defines it as a population of animals capable of interbreeding and producing fertile offspring in their natural habitat. This concept serves as a fundamental tool in evolutionary biology and biodiversity for the identification and categorization of living organisms.
A team at Dartmouth College took a bold step by removing human bias from the equation and allowing computers to autonomously determine the outcome. A cutting-edge model powered by interconnected processors, as detailed in the journal Science, was utilized to investigate the factors and conditions leading up to the Cretaceous period. Nearly 130 computers were tasked with analyzing changes in the fossil record.
The Cretaceous epoch, spanning from approximately 145 to 66 million years ago, represents the fourth and final phase of the Mesozoic Era. The Cretaceous-Paleogene extinction event marked its culmination.
The Cretaceous-Paleogene (K-Pg) extinction event, which paved the way for the rise of mammals, including the ancestors of early humans, was characterized as a pivotal moment in history.
A Fresh Perspective on Historical Events
According to Alex Cox, the lead author of the study and a graduate student at Dartmouth’s Department of Earth Sciences, the objective was to approach the issue without preconceived notions or biases. By modifying a carbon-cycle model to work in reverse, the researchers allowed the data to dictate the outcome through statistical analysis, minimizing prior information input.
Ultimately, the model elucidates the processes leading to the observed geological phenomena, irrespective of prior assumptions or beliefs.
The model processed over 300,000 instances of carbon emissions, biological productivity, and carbon monoxide output in the million years surrounding the K-Pg extinction event using a specialized form of machine learning.
Machine learning, a subset of artificial intelligence, focuses on developing algorithms and statistical models that enable computers to learn from data, identify patterns, categorize information, and make predictions without explicit programming. The collaborative nature of the processors allowed them to refine their conclusions iteratively until aligning with the fossil record, a process known as Markov Chain Monte Carlo, akin to predictive text on smartphones.
Unraveling the Root Causes of Extinction
The harsh environmental conditions during the K-Pg extinction event, named for the geological periods bracketing the disaster, are evident in geological and biological remnants. Ecosystems collapsed under volatile climatic shifts, ranging from extreme cold to scorching temperatures, accompanied by sulfur-laden atmospheres, airborne minerals, and carbon dioxide, resulting in widespread extinctions.
While the exact cause remains uncertain, evidence suggests a combination of factors, with the Chicxulub impact crater in Mexico, previously attributed to the extinction event, now challenged by emerging theories linking volcanic eruptions in the Deccan Traps region of India. The asteroid impact likely compounded the effects of pre-existing volcanic activity, leading to unprecedented global consequences.
The extent to which each event contributed to the mass extinction remains a subject of ongoing scientific inquiry, prompting Cox and Brenhin Keller, a co-author and assistant professor of earth science at Dartmouth, to adopt an unbiased computational approach.
Implications of Modeling and Volcanic Influence
The model suggests that emissions from the Deccan Traps volcanoes alone could have precipitated the mass extinction event. Prior to the Chicxulub impact, the Deccan Traps had been erupting for approximately 300,000 years, releasing significant quantities of carbon and carbon dioxide into the atmosphere.
Keller, who previously linked four of Earth’s mass extinctions to volcanic activity, highlights the significance of this independent assessment, shedding light on the substantial environmental impact of volcanic emissions.
By utilizing unbiased data, the model determined the requisite levels of carbon dioxide and carbon emissions responsible for the observed disruptions in the climate and carbon cycle. These findings align with existing research on the connection between Deccan volcanism and the K-Pg extinction event.
Asteroids and Contemporary Context
The model indicates a sharp decline in deep-sea carbon formation following the Chicxulub impact, likely due to the extinction of numerous species by the meteorite. The substantial release of sulfur, a short-term cooling agent, into the atmosphere upon impact contributed to a temporary drop in temperatures.
While the meteor impact may have released carbon dioxide and carbon monoxide, the model suggests that gas emissions did not significantly influence the extinction event, as indicated by the absence of corresponding spikes in emissions.
Conclusion: Potential Applications and Methodological Innovation
Cox notes that annual carbon dioxide emissions from fossil fuel combustion between 2000 and 2023 totaled 16 billion tons, significantly surpassing the highest annual emissions from the Deccan Traps. Despite the alarming scale of volcanic emissions, it would take thousands of years for contemporary carbon dioxide emissions to match the cumulative output of the ancient lava flows.
The model’s robustness in generating plausible outcomes underscores its effectiveness, particularly in scenarios where prior constraints are limited. By leveraging parallel processing, the computational time required to analyze vast datasets has been significantly reduced, offering a promising approach for investigating complex geological events.
The novel methodology pioneered by Cox and Keller holds promise for exploring diverse Earth systems models, such as those related to climate dynamics or the carbon cycle, by circumventing individual biases and maximizing solution space exploration.
The research by Alexander A. Cox and C. Brenhin Keller, titled “A Bayesian Approach to Pollutants and Trade Performance across the End-Cretaceous Boundary,” published in Science on 28th September 2023, delves into the innovative computational techniques employed in unraveling historical mysteries.
DOI: 10.126/science.adh3875
