Comparison of temperature reproducibility and extreme climate projections over South Korea by grid resolution (1 km vs 500 m)

This study evaluates the impact of grid resolution (1 km vs. 500 m) on temperature reproducibility and future extreme climate projections over South Korea under the shared socio-economic pathway (SSP) scenarios. First, a comprehensive performance evaluation for the historical observational period (2000 2019) revealed that the 500 m resolution improves the representation of local temperature variability compared to the 1 km dataset. By achieving higher topographic fidelity and accurately reflecting actual station altitudes, the 500 m grid effectively reduced the systematic cold bias (approximately -0.07 to -0.09°C) found in the 1 km data to near zero (±0.01°C), particularly in complex terrains. Future projections for the late 21st century (2071 2100) indicate a sharp contrast in extreme climate indices depending on emission forcing; however, the “Resolution Effect” defined as the systematic difference between 500 m and 1 km remained remarkably consistent regardless of the warming intensity of the scenarios. The 500 m resolution consistently adjusted the annual minimum temperature (TNn) upward (+0.20 to 0.22°C) and the annual maximum temperature (TXx) downward (over -1.0°C), successfully mitigating the overestimation of extreme values often observed in coarser grids. Furthermore, the inter-model spread is slightly smaller in the 500 m simulations. These findings demonstrate that ultra-high-resolution simulations are essential for capturing fine-scale topographic feedbacks and providing more reliable, physically consistent climate information necessary for establishing robust regional adaptation and disaster management strategies.