Emily J. Judd, Jessica E. Tierney, Daniel J. Lunt, Isabel P. Montañez, Brian T. Huber, Scott L. Wing, Paul J. Valdes, Science. DOI: 10.1126/science.adk3705

An interesting approach to palaeo-temperature modelling – produced by assimilating observational data during the last 485 million years with climate model predictions.

This paper introduces a new statistical estimate of global mean surface temperature (GMST) throughout the Phanerozoic Eon, called PhanDA.  PhanDA was generated using a method known as paleoclimate data assimilation, which statistically integrates geological data with climate model simulations.  This has produced some obvious differences to previous widely circulated models which only use observational data.

Here are some key takeaways from the paper:

• PhanDA reveals that Earth’s temperature has varied between 11°C and 36°C over the past 485 million years, a larger range than previous reconstructions.
• Earth has spent more time in warmer than colder climate states during the Phanerozoic.
• CO₂ appears to be the dominant control on Phanerozoic climate. There is a strong correlation between PhanDA GMST and atmospheric CO₂ concentration during the Palaeozoic and the Cenozoic, with a consistent apparent Earth system sensitivity (the change in GMST for a doubling of CO₂ concentration) of about 8°C. This consistency is surprising because solar luminosity should also influence climate on these timescales, leading to a hypothesis that changes in planetary albedo and other greenhouse gases (e.g., methane) may have compensated for increasing solar luminosity over time.
• There is poor correlation between PhanDA GMST and CO₂ during the Mesozoic which the authors refer to as the “Mesozoic Conundrum“
• PhanDA supports the conclusion that there is no fixed upper limit on tropical warmth, suggesting that ancient life must have adapted to endure extreme heat.

While PhanDA generally agrees with other records of global change, there are some deviations. For example, the minimum GMST occurs earlier than the maxima in glacial occurrence data during the Late Paleozoic Ice Age. Additionally, PhanDA indicates relatively warm temperatures during the middle Miocene while benthic foraminiferal oxygen isotopes show comparatively enriched values.

CO2 -Temperature Relationship – The Mesozoic Conundrum

Upon review of this paper, it is evident that there is very poor correlation between the PhanDA GMST model and CO₂ during the Mesozoic.

The authors acknowledge this and comment that while PhanDA GMST closely tracks reconstructed CO₂ during both the Palaeozoic and the Cenozoic, there is no discernible relationship between the two during the Mesozoic (r= -0.37, p = 0.18). This is despite the fact that the Mesozoic CO₂ reconstruction is based on data from four different CO₂ proxy types.

The authors suggest the following possible explanations for this lack of correlation:

• Limited Range of Climate States: The Mesozoic Era, unlike the Paleozoic and Cenozoic, did not experience substantial coolhouse or coldhouse climate states. This compressed range of climate states might make it difficult to detect a statistically significant relationship between CO2 and GMST.
• Uncertainties in CO2 Reconstruction: Large uncertainties in the Mesozoic CO2 reconstruction could obscure the CO2-GMST relationship. The reconstructed CO2, for instance, does not increase across the mid-Cretaceous hothouse period, which contradicts other evidence of extreme warmth. This discrepancy might point to an incomplete understanding of how different proxies record past CO2 levels.
• The “Mesozoic Conundrum”: This term, coined in the authors, highlights the challenge of reconciling the seemingly low CO2 estimates for the Mesozoic with the abundant evidence for a hothouse climate during this era. This conundrum might stem from limitations in the available proxies or a need to re-evaluate our understanding of the factors influencing climate.

Cenozoic Cooling

There is a long cooling trend evident across the Cenozoic Era which spans the last 66 million years.

PhanDA, the reconstructed GMST record, closely follows the trajectory of the benthic foraminiferal d18O stack, demonstrating this cooling trend. This cooling culminated in the rapid expansion of ice sheets during the Quaternary Period, marking a shift to colder climate states.

Maximum Cenozoic Temperatures: The warmest temperatures during the Cenozoic occurred during the Paleocene-Eocene Thermal Maximum (PETM), an abrupt warming event around 56 million years ago. This warmth persisted into the early Eocene.

Drivers of Cooling: While the exact drivers of this long-term cooling trend are unclear, several factors likely played a role. These include:

• Decreasing CO2 Concentrations: The authors highlight a strong correlation between CO2 and GMST, particularly in the Cenozoic. As CO2 concentrations declined over this era, global temperatures also decreased.
• Tectonic Processes: Plate tectonics and continental drift can influence global climate patterns. For example, the uplift of the Himalayas and the opening of the Drake Passage are thought to have contributed to Cenozoic cooling.
• Changes in Ocean Circulation: Shifts in ocean currents can alter heat distribution across the globe. The formation of the Antarctic Circumpolar Current, for instance, likely played a role in isolating Antarctica and fostering ice sheet growth.

Middle Miocene Discrepancy: Interestingly, the authors point to a notable deviation between PhanDA and the benthic foraminiferal oxygen isotope record during the middle Miocene. While PhanDA indicates relatively warm temperatures during this period, the benthic oxygen isotope values suggest cooler conditions. This discrepancy could indicate that the rise in benthic d18O following the middle Miocene Climatic Optimum reflects ice sheet growth rather than a global cooling trend.