Comparison of Simulated Precipitation over East Asia in Two Regional Models with Hydrostatic and Nonhydrostatic Dynamical Cores

Abstract

This study examines the characteristics of a nonhydrostatic dynamical core compared to a corresponding hydrostatic dynamical core in the Regional Model Program (RMP) of the Global/Regional Integrated Model system (GRIMs), a spectral model for regional forecasts, focusing on simulated precipitation over Korea. This kind of comparison is also executed in the Weather Research and Forecasting (WRF) finite-difference model with the same physics package used in the RMP. Overall, it is found that the nonhydrostatic dynamical core experiment accurately reproduces the heavy rainfall near Seoul, South Korea, on a 3-km grid, relative to the results from the hydrostatic dynamical core in both models. However, the characteristics of nonhydrostatic effects on the simulated precipitation differ between the RMP and WRF Model. The RMP with the nonhydrostatic dynamical core improves the local maximum, which is exaggerated in the hydrostatic simulation. The hydrostatic simulation of the WRF Model displaces the major precipitation area toward the mountainous region along the east coast of the peninsula, which is shifted into the observed area in the nonhydrostatic simulation. In the simulation of a summer monsoonal rainfall, these nonhydrostatic effects are negligible in the RMP, but the simulated monsoonal rainfall is still influenced by the dynamical core in the WRF Model even at a 27-km grid spacing. One of the reasons for the smaller dynamical core effect in the RMP seems to be the relatively strong horizontal diffusion, resulting in a smaller grid size of the hydrostatic limit.

Exploring the effects of a nonhydrostatic dynamical core in high‐resolution aquaplanet simulations

“…These are then amplified by feedbacks associated with moist processes and their parameterizations. Comparing simulations with hydrostatic and nonhydrostatic solvers in two models, WRF and the Regional Model Program (RMP), applied to the mountainous region of Korea, Jang and Hong [] also found notable H and NH differences in the WRF simulations even at 27 km grid spacing. They attributed the larger differences in WRF compared to the results from RMP to the weaker horizontal diffusion in WRF.…”

Comparison of Simulated Tropical Cyclone Intensity and Structures Using the WRF with Hydrostatic and Nonhydrostatic Dynamical Cores

“…It has been validated for meso-α, synoptic, and global-scale circulations, and thus it could be considered a reliable and accurate assumption [3]. As one of the fundamental assumptions adopted in atmospheric and oceanic models [3][4][5], the hydrostatic approximation has been incorporated in numerical weather prediction models for decades to allow longer time steps by filtering out vertically propagating sound waves [6].…”

“…Following the rapid improvement of computing resources in recent years, the use of nonhydrostatic numerical models has become widespread. With the aid of new numerical treatments designed to handle sound waves, e.g., semi-implicit [7] and split-explicit techniques [8], nonhydrostatic mesoscale models have developed rapidly, and they have been used actively in research and operational communities during the previous two decades [6]. However, global climate models remain mostly based on hydrostatic atmospheric dynamical cores, even though the increase in computational cost associated with the use of a nonhydrostatic dynamical core is not particularly large [9].…”

“…Based on analytical solutions, the characteristics of hydrostatic and nonhydrostatic flows have been studied theoretically [10,11]. Furthermore, idealized experiments have been conducted to validate nonhydrostatic mesoscale models in situations that include mountain waves over a bell-shaped mountain, density currents produced by a cold bubble, and rising thermal bubble cases [6,12,13]. These previous studies found that two-dimensional idealized cases were reproduced well only in nonhydrostatic experiments.…”

“…Besides this, the vertical motion in the eyewall of TCs is weaker than in other severe storms such as tornadoes and mid-latitude thunderstorms [20] Zhang et al [21] suggested that hydrostatic models might be used to simulate hurricanes using a grid size of <6 km, providing that some cloud microphysics processes and water loading were incorporated properly. Although several studies have compared the effects of using both hydrostatic and nonhydrostatic dynamical cores in simulations of severe weather phenomena [6,9], an examination of the influence of both cores in simulations of tropical cyclone (TC) intensity and structure, especially at high resolution, is still lacking. The objective of this study was to examine the differences between hydrostatic and nonhydrostatic simulations of TC intensity and structure using the weather research and forecasting [22] model as a tool at cloud-resolving resolutions.…”