Impact of Lateral Boundary Flows on Regional Convection‐Permitting Simulations Over the Tibetan Plateau: A Global‐Regional Integrated Modeling Study

Abstract

Abstract The GRIST model is used for the first time in a regional downscaling experiment based on the convection‐permitting third‐pole monsoon case. The simulations driven by external reanalysis data sets are assessed and compared with the global simulation through a rigorous global‐regional integrated modeling approach. Additional regional simulations with boundary data taken from the global simulation reveal the critical role of cross‐boundary flows in aligning regional model behaviors with global results. The study focuses on downscaling performance, global‐regional comparisons, and the impact of lateral boundary flow. The downscaling simulations using different reanalysis data sets produce overall comparable large‐scale circulation patterns and mean precipitation biases. Nudged lateral boundary conditions improve the circulation performance but result in mixed precipitation outcomes, including higher mean‐state biases and artificial rainfall around the Tibetan Plateau area. Some intrinsic model biases (e.g., diurnal cycle and excessive light rainfall frequency) are consistent across global and regional simulations. Using explicit convection can address these limitations. Intense rainfall events and topographic precipitation errors show high sensitivity to lateral boundary flow variations, underscoring the complexity of interactions between regional dynamics and boundary flows. Systematic topographic precipitation biases persist but varying lateral boundary flows can regulate the magnitude. The results underscore the uncertainties associated with kilometer‐scale downscaling simulations under strong lateral boundary flows particularly concerning small‐scale intense and/or topographic rainfall events.