The Impact of Implicit-Explicit Vertical Advection Schemes on Operational Weather Forecasting: A Summer-Period Study with the Regional Prediction System by the KMA

Joonsung Park, Sang-Hun Park, Uju Shin, Se-Hyun Lee, Eunhee Lee, Jeong-Soon Lee

October 2025

Abstract

Operating convection-allowing models pose challenges because of the increased computational demands associated with higher horizontal and vertical resolutions, especially for forecasts involving strong vertical motions, such as convective systems. The primary stability concern in these models arises from the high vertical velocities associated with convection-resolving simulations rather than being limited by horizontal wind conditions. Minimizing the computational costs while maintaining stability is crucial for operational forecasting systems. This study investigates the performance of an implicit-explicit vertical advection scheme (IEVA) integrated into the Weather Research and Forecasting (WRF) version 4.3 and later, when applied to the Korean Meteorological Administration’s Regional Prediction System (KRPS). As designed, IEVA enables stable model operation even with a 28.125-second time step, exceeding the recommended 18 s for the WRF models. The integration time is reduced to 74.3%, although this reduction is slightly less than anticipated. Our validation, based on a one-month historical dataset from the 2020 Changma season, confirmed the consistency of the IEVA in simulating meteorological and precipitation fields at the meso-β scale, particularly emphasizing the challenges posed by high vertical velocities associated with convection-resolving simulations. Statistical analysis, using skill scores, indicates that the variability introduced by IEVA was similar in magnitude to that resulting from differences in the time step.

Type

Journal article

Publication

Weather and Forecasting, 40(12)