Simulation of mid-latitude winter storms over the North Atlantic Ocean: impact of boundary layer parameterization schemes

Abstract

This study discusses the performance of various planetary boundary layer parameterization (PBL) schemes—the Quasi-Normal Scale Elimination (QNSE), the University of Washington Moist Turbulence (UWMT), and the Yonsei University (YSU)—for the simulation of rapidly developing North Atlantic (NA) mid-latitude winter storms. Sensitivity experiments with the three PBL schemes, YSU, QNSE, and UWMT, indicate that there are minor differences at the center of the storm while simulating the evolution of the three explosive storms Klaus (21–27 January 2009), Xynthia (25 February–03 March 2010), and Gong (16–20 January 2013). The differences are shown in terms of the central minimum pressure, 10-m wind, specific humidity, CAPE, transitional speed, boundary layer height and frictional velocity of these mid-latitude storms. One of the main result shows the capability of QNSE and UWMT PBL schemes to reproduce accurately both the cyclogenesis and explosive stage for these mid-latitude storms during the winter season, better than YSU scheme. Almost all PBL schemes show dry bias from middle to upper troposphere (600 hPa–250 hPa), while YSU scheme carries this bias at the surface boundary layer, for all simulations. Moreover, QNSE, UWMT and YMSU PBL schemes underestimate the tangential winds for these mid-latitude storms. The 24 h accumulated latent heat flux and precipitation from UWMT scheme show modified results as compared to YSU and QNSE PBL schemes. Overall results show the superiority of QNSE and UWMT PBL schemes for an accurate simulation of the explosive stage of these North Atlantic winter storms.