Earth system box models are essential tools for reconstructing long-term climatic and environmental evolution and uncovering Earth system mechanisms. To overcome the spatiotemporal resolution limitations of current deep-time models, a research team has developed CESM-SCION, a new-generation high-resolution climate-biogeochemistry coupled model. This model advances the spatiotemporal resolution of long-term Earth system simulations to a new level and identifies marine regression as a key driver of the Late Paleozoic Ice Age onset.

The findings were recently published in Geophysical Research Letters. Led by Prof. ZHU Maoyan from the Nanjing Institute of Geology and Palaeontology of the Chinese Academy of Sciences (NIGPAS), the team collaborated with researchers from Peking University, the University of Leeds, and the University of Adelaide.

Since Berner and colleagues established the classic GEOCARB carbon cycle model in the 1980s, Earth system box models have evolved from early zero-dimensional biogeochemical models (e.g., GEOCARB, COPSE) to climate-biogeochemistry coupled models such as SCION. These advanced models can dynamically represent terrestrial climate and silicate weathering across space and time. While SCION achieved three-dimensional dynamic expression of terrestrial silicate weathering, its built-in climate emulator (or climate datasets) was constrained by low spatial resolution. This limitation hindered the accurate capture of complex, heterogeneous weathering processes on the deep-time Earth surface, often overlooking small-scale weathering "hotspots."

To address this gap, the team leveraged high-resolution climate datasets from the Community Earth System Model (CESM). By increasing the spatiotemporal resolution to 3.75°×3.75° and 10 Myr, they replaced the original low-resolution climate emulator to develop CESM-SCION. This advancement enables the high-resolution dynamic capture of terrestrial silicate weathering processes.

To verify the new model's reliability, the team selected the Late Paleozoic Ice Age—the longest-lasting icehouse event of the Phanerozoic—as a test case. Previous studies using low-resolution versions of SCION, despite accounting for mechanisms such as vascular plant expansion and Pangea assembly, failed to reproduce the greenhouse-to-icehouse climate transition during the mid-to-late Devonian. Using CESM-SCION, the team successfully identified a previously overlooked key driving mechanism: the critical control of tectonically driven land area changes on Earth's "weatherability" and surface albedo.

Modeling results indicate that during the mid-to-late Devonian, Pangea's assembly was accompanied by long-term global sea-level fall (marine regression). With enhanced resolution, the CESM-SCION model quantified the "dual cooling effect" of this process. Increased global continental exposure from marine regression acted in two ways: first, it induced physical cooling by raising global surface albedo; second, it greatly expanded regions available for silicate weathering, boosting global silicate weathering flux. This accelerated atmospheric CO₂ consumption, driving global cooling. The result not only reproduces the critical mid-to-late Devonian greenhouse-to-icehouse transition but also proposes a new perspective: "Tectonically driven marine regression preceded and triggered the Late Paleozoic Ice Age."

This study underscores the importance of improving the resolution of deep-time Earth system simulations. It resolves long-standing debates about the initiation of the Late Paleozoic Ice Age and highlights the potential of high-resolution climate-biogeochemistry coupled models in deep-time paleoclimate research.

The work was funded by the National Key R&D Program of China, the National Natural Science Foundation of China (NSFC), and other sources.