Timing and Style of Final Closure of the Paleo-Asian Ocean: Perspectives From the Big Geodata Analysis and Machine Learning Model

Abstract

The Paleo-Asian Ocean (PAO), was located between the Siberian Craton and the Tarim-North China Craton, and its subduction-closure history are important for studying the evolution of the Central Asian Orogenic Belt that may have uncovered impacts on paleoclimate. However, when and how the PAO closed remains controversial. Additionally, the potential link between the Permian aridification in the northern North China Craton and the subduction-closure processes of the PAO also remains elusive. The closure locations of the PAO are at the South Tianshan Belt, Beishan Belt, Solonker Belt, and Changchun-Yanji Belt, respectively. Here, we apply a LightGBM model to estimate the crustal thickness of the four segments, combined with big geodata analysis methods and geological evidence, to address these scientific issues. The ∼300, ∼280, ∼260, and ∼240 Ma crustal thickening and corresponding zircon ε_Hf(t) isotopic pull-downs of these belts, coupled with flora distribution, sedimentary records, and paleomagnetic data, support that the PAO closed diachronously in a scissor style from west to east during the Late Carboniferous to the Middle Triassic. The new estimate regarding crustal thickness for northern North China by the LightGBM model was 47 ± 8 km during the Early Permian, which corresponded to an elevation of 2.2 ± 0.8 km. Furthermore, the profile of the new paleo-elevation of the northern North China suggests that the topography and elevation are remarkably different from those of the Andean Altiplano plateau, which argues against the interpretation of an Andean-type orogenic plateau resulting in the Permian aridification of the northern North China Craton.