A computer-driven quest for extinction causes
30.09.2023 posted by Admin
To settle the long-standing scientific dispute about whether dinosaurs and countless other species were wiped out by a colossal asteroid or volcanic eruptions 66 million years ago, Dartmouth researchers adopted a fresh strategy. They took scientists out of the equation and allowed computers to make the call.
Their findings, published on September 29th in the journal Science under the title "A Bayesian inversion for emissions and export productivity across the end-Cretaceous boundary," present a groundbreaking modeling method driven by interconnected processors that can sift through vast volumes of geological and climate data independently, devoid of human influence.
Nearly 130 processors were enlisted to retroactively analyze the fossil record, pinpointing the events and circumstances leading to the Cretaceous-Paleogene (K-Pg) extinction event. This event paved the way for the rise of mammals, including the ancestors of early humans.
The study's first author, Alex Cox, a graduate student in Dartmouth's Department of Earth Sciences, explained, "Our goal was to approach this question without any preconceived notions or biases. Most models typically move forward in time. We reversed a carbon-cycle model's direction to trace the cause from the effect using statistics, providing it with minimal prior information as it worked toward a specific outcome."
Ultimately, the model, using a machine learning technique called Markov Chain Monte Carlo, generated over 300,000 possible scenarios involving carbon dioxide emissions, sulfur dioxide release, and biological productivity in the million years surrounding the K-Pg extinction. These processors autonomously collaborated, compared, adjusted, and recalculated their results until they matched the outcome preserved in the fossil record.
The fossil record retains clear evidence of catastrophic conditions during the K-Pg extinction, marked by widespread extinctions as food chains collapsed under an unpredictable atmosphere laden with sun-blocking sulfur, airborne minerals, and heat-trapping carbon dioxide that swung between freezing and scorching conditions.
Although the effect of the event is evident, the cause remains elusive. Initially, theories pointed to volcanic eruptions, but the discovery of the Chicxulub impact crater in Mexico, caused by a massive asteroid, has shifted the narrative. Recent fossil evidence suggests a dual catastrophe scenario: an asteroid impact coinciding with the ongoing volcanic eruptions in India's Deccan Traps.
However, the extent to which each event contributed to the mass extinction remains a contentious issue among scientists. So, Cox and Brenhin Keller, an assistant professor of Earth Sciences at Dartmouth and a co-author of the study, decided to let the computer code decide.
Their model indicated that the emissions from the Deccan Traps alone could have been sufficient to trigger the global extinction. Over nearly a million years, these eruptions released up to 10.4 trillion tons of carbon dioxide and 9.3 trillion tons of sulfur into the atmosphere.
While historical evidence suggests that volcanoes can cause mass extinctions, this study provided an independent estimation of emissions based on environmental effects, confirming the Deccan Traps' emissions aligned with the disruptions seen in the geological record.
The model did show a sharp decline in the accumulation of organic carbon in the deep ocean around the time of the Chicxulub impact, likely due to the asteroid's effect on numerous plant and animal species. This period also exhibited a drop in temperature caused by the large sulfur release from the impact.
Despite the asteroid impact potentially emitting carbon and sulfur dioxide, the model found no significant spike in gas emissions at that time, indicating that the asteroid's role in the extinction may not have hinged on gas emissions.
To put it in perspective, the study's lead author, Cox, compared modern carbon dioxide emissions from 2000 to 2023 to the Deccan Traps' emissions, noting that the former was 100 times greater. However, the ancient volcanic emissions far exceeded the cumulative output of current fossil fuel burning.
What's encouraging is that the results align with known physical realities, considering the model had minimal initial constraints and could have theoretically gone off track.
By connecting the processors, the researchers significantly accelerated the data analysis process, reducing it from months or years to mere hours. This innovative method can be applied to other Earth system models, enabling the examination of geological events with known outcomes but unknown causal factors, minimizing human bias and expanding the solution space.
In conclusion, the approach employed by the researchers, allowing the computer model to independently decipher the cause of the K-Pg extinction, has opened new avenues for understanding Earth's past catastrophes and their impacts.
Their findings, published on September 29th in the journal Science under the title "A Bayesian inversion for emissions and export productivity across the end-Cretaceous boundary," present a groundbreaking modeling method driven by interconnected processors that can sift through vast volumes of geological and climate data independently, devoid of human influence.
Nearly 130 processors were enlisted to retroactively analyze the fossil record, pinpointing the events and circumstances leading to the Cretaceous-Paleogene (K-Pg) extinction event. This event paved the way for the rise of mammals, including the ancestors of early humans.
The study's first author, Alex Cox, a graduate student in Dartmouth's Department of Earth Sciences, explained, "Our goal was to approach this question without any preconceived notions or biases. Most models typically move forward in time. We reversed a carbon-cycle model's direction to trace the cause from the effect using statistics, providing it with minimal prior information as it worked toward a specific outcome."
Ultimately, the model, using a machine learning technique called Markov Chain Monte Carlo, generated over 300,000 possible scenarios involving carbon dioxide emissions, sulfur dioxide release, and biological productivity in the million years surrounding the K-Pg extinction. These processors autonomously collaborated, compared, adjusted, and recalculated their results until they matched the outcome preserved in the fossil record.
The fossil record retains clear evidence of catastrophic conditions during the K-Pg extinction, marked by widespread extinctions as food chains collapsed under an unpredictable atmosphere laden with sun-blocking sulfur, airborne minerals, and heat-trapping carbon dioxide that swung between freezing and scorching conditions.
Although the effect of the event is evident, the cause remains elusive. Initially, theories pointed to volcanic eruptions, but the discovery of the Chicxulub impact crater in Mexico, caused by a massive asteroid, has shifted the narrative. Recent fossil evidence suggests a dual catastrophe scenario: an asteroid impact coinciding with the ongoing volcanic eruptions in India's Deccan Traps.
However, the extent to which each event contributed to the mass extinction remains a contentious issue among scientists. So, Cox and Brenhin Keller, an assistant professor of Earth Sciences at Dartmouth and a co-author of the study, decided to let the computer code decide.
Their model indicated that the emissions from the Deccan Traps alone could have been sufficient to trigger the global extinction. Over nearly a million years, these eruptions released up to 10.4 trillion tons of carbon dioxide and 9.3 trillion tons of sulfur into the atmosphere.
While historical evidence suggests that volcanoes can cause mass extinctions, this study provided an independent estimation of emissions based on environmental effects, confirming the Deccan Traps' emissions aligned with the disruptions seen in the geological record.
The model did show a sharp decline in the accumulation of organic carbon in the deep ocean around the time of the Chicxulub impact, likely due to the asteroid's effect on numerous plant and animal species. This period also exhibited a drop in temperature caused by the large sulfur release from the impact.
Despite the asteroid impact potentially emitting carbon and sulfur dioxide, the model found no significant spike in gas emissions at that time, indicating that the asteroid's role in the extinction may not have hinged on gas emissions.
To put it in perspective, the study's lead author, Cox, compared modern carbon dioxide emissions from 2000 to 2023 to the Deccan Traps' emissions, noting that the former was 100 times greater. However, the ancient volcanic emissions far exceeded the cumulative output of current fossil fuel burning.
What's encouraging is that the results align with known physical realities, considering the model had minimal initial constraints and could have theoretically gone off track.
By connecting the processors, the researchers significantly accelerated the data analysis process, reducing it from months or years to mere hours. This innovative method can be applied to other Earth system models, enabling the examination of geological events with known outcomes but unknown causal factors, minimizing human bias and expanding the solution space.
In conclusion, the approach employed by the researchers, allowing the computer model to independently decipher the cause of the K-Pg extinction, has opened new avenues for understanding Earth's past catastrophes and their impacts.