The Impacts of Convective Parameterization and Moisture Triggering on AGCM-Simulated Convectively Coupled Equatorial Waves

Abstract

This study examines the impacts of convective parameterization and moisture convective trigger on convectively coupled equatorial waves simulated by the Seoul National University (SNU) atmospheric general circulation model (AGCM). Three different convection schemes are used, including the simplified Arakawa-Schubert (SAS) scheme, the Kuo (1974) scheme, and the moist convective adjustment (MCA) scheme, and a moisture convective trigger with variable strength is added to each scheme. The authors also conduct a "no convection" experiment with deep convection schemes turned off. Space-time spectral analysis is used to obtain the variance and phase speed of dominant convectively coupled equatorial waves, including the Madden-Julian oscillation (MJO), Kelvin, equatorial Rossby (ER), mixed Rossby-gravity (MRG), and eastward inertio-gravity (EIG) and westward inertio-gravity (WIG) waves.The results show that both convective parameterization and the moisture convective trigger have significant impacts on AGCM-simulated, convectively coupled equatorial waves. The MCA scheme generally produces larger variances of convectively coupled equatorial waves including the MJO, more coherent eastward propagation of the MJO, and a more prominent MJO spectral peak than the Kuo and SAS schemes. Increasing the strength of the moisture trigger significantly enhances the variances and slows down the phase speeds of all wave modes except the MJO, and usually improves the eastward propagation of the MJO for the Kuo and SAS schemes, but the effect for the MCA scheme is small. The no convection experiment always produces one of the best signals of convectively coupled equatorial waves and the MJO.

Validity of parameter optimization in improving MJO simulation and prediction using the sub-seasonal to seasonal forecast model of Beijing Climate Center

“…The adjustment time scale and precipitation efficiency for shallow convection (τ_shal and C 0 _shal), the Cai et al (2013). The sensitivities of RH_trig, C 0 _deep and K e _deep in deep convection parameterization for MJO simulation were shown by many studies (e.g., Wang and Schlesinger 1999;Maloney and Hartmann 2001;Zhang and Mu 2005;Lin et al 2008b;Boyle et al 2015). In addition, for the atmospheric component model of the BCC_CSM, Yang et al (2015) illustrated the significant sensitivity of precipitation and winds over the tropical Indian Ocean and western Pacific to the parameters of C 0 _deep, K e _deep, RH_low and RH_high.…”

“…Especially, convection parameterization scheme is a critical part in determining model's ability in simulating MJO (e.g., Lee et al 2003;Liu et al 2005;Kim and Kang 2012;Cai et al 2013;Zhu et al 2017). Even small changes of physical parameters related to convection and precipitation, such as closure assumption, convection trigger, evaporation of convective precipitation, entrainment rate, and diabatic heating, can result in marked differences in MJO's representation in a model (e.g., Wang and Schlesinger 1999;Maloney and Hartmann 2001;Zhang and Mu 2005;Lin et al 2008b;Li et al 2009;Boyle et al 2015;Del Genio et al 2015). In this context, refining physics parameterization, optimizing relevant key physical parameters and thus improving model performance become goals of research groups for numerical simulation of MJO.…”

MJO simulation in CMIP5 climate models: MJO skill metrics and process-oriented diagnosis

“…While GCM simulations of the MJO have largely remained at an unsatisfactory level in multi-model intercomparison studies, it has been continuously demonstrated that MJO simulation can be improved by making appropriate changes to parameterizations, especially that of cumulus convection (Tokioka et al 1988;Wang and Schlesinger 1999;Maloney and Hartmann 2001;Lin et al 2008;Zhang and Song 2009;Hannah and Maloney 2011;Kim and Kang 2012;, as summarized in Kim and Maloney (2017). However, the knowledge and experience from these studies have not been fully utilized in operational versions of climate and NWP models, possibly because the methods that improve the MJO often degrade other aspects of model simulation, such as the mean state (e.g., Kim et al 2011).…”

The parametric sensitivity of CAM5's MJO

“…However, even with conventional GCMs, there have been some encouraging results in individual models. In particular, improved simulation results from inhibiting the deep convection either with more restrictive closure assumptions [Sperber et al, 2005;Zhang and Mu, 2005;Lin et al, 2008;Benedict et al, 2013] or by increasing the lateral entrainment in convective updrafts which enhances the sensitivity of the convection scheme to free tropospheric moisture [Tokioka et al, 1988;Bechtold et al, 2008;Del Genio et al, 2012;Kim et al, 2011;Zhou et al, 2012;Hirons et al, 2012]. Given moisture's central role, it is not surprising that some improvements can also be found by altering the moistening tendencies of convection, for example, by increasing the amount of evaporation of convective rain [Maloney, 2009;Del Genio et al, 2012].…”