Challenges in Enhancing Global Numerical Weather Prediction: Transitioning from 12 km to 8 km resolution in the Convective Gray-Zone

With the initiation of the Korea Institute of Atmospheric Prediction Systems (KIAPS) phase-II project (2020-2026), the 8-km version of the Korean Integrated Model (KIM) has been developed for operational medium-range prediction. This study presents a year-long semi-real-time validation comparing the performance of the 12 km and 8 km models. We aim to evaluate the benefits of increased resolution toward convective gray-zone scales and address critical issues necessary for optimizing gray-zone simulations. Specifically, we analyze physics sensitivity issues identified in the latest updates of KIM's horizontal and vertical resolutions. Additionally, we explore the impact of horizontal resolution on medium-range predictions across various configurations, from 100 km to 8 km, using a cold-run testbed without data assimilation. A particular emphasis is placed on performance improvements between the 12 km and 8 km configurations, where scale-aware parameterization is utilized, particularly in the deep convection scheme. Our study highlights KIM’s capability to predict high-impact weather events, including typhoons and heatwaves impacting South Korea. Furthermore, we will discuss the development strategies for km-scale predictions at KIAPS and provide an overview of our current km-scale NWP configurations, including fixed and variable-solution grids as well as limited area models.