Is there added value of convection-permitting regional climate model simulations for storms over the German Bight and Northern Germany?

Abstract

Abstract. This study tackles the question: Do very high-resolution convective-permitting regional climate model (RCM) simulations add value compared to coarser RCM runs for certain extreme weather conditions, namely strong wind and storm situations? Ten strong storm cases of the last two decades were selected and dynamically downscaled with the RCM COSMO-CLM (24 and 2.8 km grid point distance). These cyclones crossed the high-resolution model domain, which encompasses the German Bight, Northern Germany, and parts of the Baltic Sea. The multiple storm analysis revealed added value for the high-resolution regional climate simulation for 10 m wind speed, mean sea level pressure, and total cloud cover for most storms which were examined, but the improvements are small. Wind direction and precipitation were already well simulated by the coarser RCM and the higher resolution could often not add any value for these variables. The analysis showed that the added value is more distinct for the synoptic comparisons than for the multiple storm study analyzed with statistical measures like the Brier Skill Score.

Long-Term Atmospheric Changes in a Convection-Permitting Regional Climate Model Hindcast Simulation over Northern Germany and the German Bight

“…The similarity in the long-term evaluation is due to the fact coastDat II was used as forcing data for GB0028. The advantage of the higher resolution is evident in the small-scale synoptic comparison of single weather situations [13]. From all data sets, coastDat II shows the lowest warming (2 m temperature increase of 0.6 • C) occurring for the area mean in the metropolitan region of Hamburg (Figure 1) between the two time intervals 1986-2015 and 1961-1990 (Table 1).…”

“…This is presumably a consequence of the relatively small model domain, as spacious weather patterns which deviate from observed large weather systems have no time to develop inside this domain [31]. The data set was validated regarding added value of this high resolution of 2.8 km compared to a simulation with 24 km grid distance and observations for ten storm events [13]. It was shown that the simulation can reproduce storm events realistically and the added value of the convection-permitting resolution is mainly in synoptic comparisons of single storm cases.…”

“…Presently, very high-resolution regional climate model studies are used to simulate climate at convection-permitting resolution [11]. Such simulations can add value in comparison to coarser regional climate model simulations (e.g., [12,13]), but are generally limited to shorter time scales and smaller regions due to computational constraints. The added value of a convection-permitting simulation in comparison to its forcing reanalysis was described in [13] for ten storm cases.…”

“…Such simulations can add value in comparison to coarser regional climate model simulations (e.g., [12,13]), but are generally limited to shorter time scales and smaller regions due to computational constraints. The added value of a convection-permitting simulation in comparison to its forcing reanalysis was described in [13] for ten storm cases. The added value mainly showed in dynamical processes such as convective precipitation or post-frontal cloud cover and for multiple storm events in 10 m wind speed, mean sea level pressure, and total cloud cover.…”

“…In this study, 2 m temperature, precipitation amount, Convective Available Potential Energy (CAPE), 10 m wind speed, and wind gusts are analyzed. As we did evaluate the simulation in a former study [13] the added value of this study is not the main emphasis of this article, but we here focus on long-term variability, multi-decadal trends, and their regional features. Next, the long-term variability of potentially damaging extreme events is examined.…”