Impact of Horizontal Resolution on the Simulation of Super Typhoon Lekima (2019) in the GRIST Variable-Resolution Model

This study assesses how horizontal resolution affects the simulation of Super Typhoon Lekima (2019) in the Global-Regional Integrated forecast SysTem (GRIST). Three global variable-resolution experiments, with finest grid spacings of 1 km, 3.5 km, and 10 km, are evaluated against the China Meteorological Administration-Shanghai Typhoon Institute (CMA-STI) best track and environmental diagnostics based on the fifth-generation European Centre for Medium-Range Weather Forecasts reanalysis (ERA5). All three experiments reproduce the general northwestward motion, but differences among the simulated tracks become more pronounced during the pre-landfall turning stage. The 1 km experiment yields the smallest track errors. At 36 h, the track errors are approximately 38, 63, and 92 km in the 1 km, 3.5 km, and 10 km experiments, respectively, increasing to approximately 140, 188, and 248 km at 72 h. At 0000 UTC on 8 August, the 500–850 hPa layer-mean steering direction in the 1 km experiment is closest to ERA5, whereas the 3.5 km and 10 km experiments exhibit stronger westward components. The intensity response is metric dependent. The 1 km experiment most closely reproduces the peak near-surface wind magnitude, whereas the 3.5 km experiment most closely reproduces the observed minimum pressure. All three simulations attain their intensity extrema earlier than the best track. Inner-core diagnostics show the strongest peak midlevel ascent in the 1 km experiment. However, the timing and mature-stage variability of the radius of maximum wind (RMW) do not vary monotonically with grid spacing. The radial alignment between the RMW and peak upward motion is also nonmonotonic. In addition, the 1 km experiment retains more kinetic energy at mesoscale and smaller wavelengths and delays spectral decay relative to the coarser experiments. Overall, the 1 km experiment produces the smallest track errors, most closely reproduces the peak-wind magnitude, and retains the most mesoscale kinetic energy. Its effects on other intensity and inner-core structural metrics are more complex and nonmonotonic.