Probablistic and Deterministic Stratigraphy for Constructing Biostratigraphic Composite Sequence: A Casse Study of Carboniferous Biostratigraphy in South China

Constructing a global or regional biostratigraphic composite sequence as a reference of strati-graphic correlation is an important task for biostratigraphy researchers. The traditional method to construct the composite sequence mainly involves qualitative and empirical evaluation of biostratigraphic datums based on a few key sections. In contrast, quantitative stratigraphic methods provide a systematic approach to analyze and compile fossil records from a large number of stratigraphic sections, enabling the construction of more widely applicable composite sequences. Probabilistic and deterministic stratigraphy represent two distinct quantitative stratigraphic philosophies for inferring sequences of stratigraphic datums to support stratigraphic correlation. In this study, Ranking and Scaling (RASC) and Graph-Optimized Graphic Correlation (GOGC) are introduced as representative methods of quantitative probabilistic and deterministic stratigraphy, respectively. These methods are ap-plied to study the bioevent sequence within the Carboniferous stratigraphic framework of South China, aiming to clarify the differences in results between the two sequencing approaches. Major marine sections in South China dominated by stratigraphically indicative conodont and fusulinid fossils were compiled. Both methods were applied to this dataset to quantitatively evaluate the observed records of the first occurrence (FO) datums of all fossils, which are often used as indicators for biostratigraphic zonation. The resulting sequences were compared with each other, as well as with the empirical sequence of Carboniferous conodonts and fusulinids of South China. While the high similarity between the empirical sequence and the RASC sequence demonstrates largely consistent superposition relationship of the target fossil events across stratigraphic sections in South China, the divergences between the RASC and GOGC sequences reflect inherent discrepancies of the two quantitative methods, as well as uncertainties arising from empirical stratigraphic judgement.