An Assessment of Multivariate Bias Correction Effects and Model Consistency in CMIP6 Monthly Precipitation Trends

General Circulation Model (GCMs) are essential tools for projecting future precipitation trends; however, structural biases and shared errors across models raise concerns about whether the ensemble consensus genuinely reflects physical climate signals. While most bias correction (BC) studies focus on improving the statistical accuracy of individual models, the implications of BC on the structural uncertainty and collective consistency of multi-model ensembles remain underexplored. This study investigates how the Robust Multivariate Bias Correction (RoMBC) method, beyond reducing model-level errors, reconfigures the interpretation of precipitation trends and inter-model consensus within CMIP6 ensembles. We applied RoMBC and the conventional univariate quantile mapping (QM) to monthly precipitation outputs from ten CMIP6 GCMs and evaluated their performance and trend fidelity against the ERA5 reanalysis. RoMBC consistently outperformed QM across all statistical metrics—including Kling–Gupta Efficiency, root mean square error, and Pearson correlation—and better captured the spatial patterns and directions of long-term trends, as assessed via seasonal Mann–Kendall tests. More importantly, the Data Concurrence Index (DCI) revealed that RoMBC strengthened inter-model agreement in Europe while weakening it in Asia, suggesting that it removes spurious consensus caused by common biases and exposes underlying structural uncertainty. Additionally, ensemble agreement remained consistently low in Australia and Africa, regardless of the BC method, indicating inherently high uncertainty in those regions. These findings suggest that RoMBC does not simply reduce uncertainty but rather reshapes the ensemble structure to more faithfully represent the inter-model spread of projected signals. This work highlights the importance of expanding BC evaluation beyond individual model performance, offering a novel perspective on interpreting ensemble-based future precipitation projections.(This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (RS-2025-23523230))