Abstract
Using the sub-seasonal to seasonal forecast model of Beijing Climate Center, several key physical parameters are perturbed by the Latin hypercube sampling method to find a better configuration for representation of Madden–Julian oscillation (MJO) in the free-run simulation. We find that although model simulation is especially sensitive to some parameters, there are overall no significant linear relationships between model skill and any one of the parameters, and the optimum performance can be obtained by combined perturbations of multiple parameters. By optimization, MJO’s spectrum, intensity, spatial structure and propagation, as well as the mean state and variance, are all improved to some extent, suggesting the correspondence and interrelation of model’s performances in simulating different characteristics of MJO. Further, several sets of initialized hindcasts using the optimized parameters are conducted, and their results are compared with the hindcasts using only improved initial conditions. We show that with an optimized model, the forecast of MJO beyond 3-week lead time is not improved, and the maximum useful skill is only slightly increased, implying that a decrease of model error does not always translate into an increase of forecast skill at all lead time. However, the skill is obviously enhanced during lead times of 2–3 weeks for forecasts in most seasons and initial phases except for a few cases. Particularly, the deficiency in forecasting MJO’s propagation from the Indian Ocean to the Pacific is relieved, further highlighting the positive contribution of reducing model error compared to previous work that only reduced initial condition error. In this study, we also show benefits of multi-scheme ensemble strategy in describing uncertainties of model error and initial condition error and thus improving MJO forecast.
Highlights
Madden–Julian oscillation (MJO; Madden and Julian 1971) is a well-known phenomenon that prevails in the tropics and exerts remarkable modulations on the tropical and extratropical atmospheric circulations
Using the method shown in Sect. 2.2.1, we compute the pattern correlation coefficient (PCC) of WFS_U850, WFS_OLR, empirical orthogonal functions (EOFs), and LC_PCs between simulations and observations for all the experiments with perturbed parameters, and rank the skillful and unskillful runs according to the magnitude of average PCC scores
The metrics for parameter optimization in this study are aimed at the improvement of MJO features other than that of climatology or variance, but the results shown in Figs. 4 and 5 partly confirm the correspondence of skill enhancement in climatology and that in MJO itself
Summary
Madden–Julian oscillation (MJO; Madden and Julian 1971) is a well-known phenomenon that prevails in the tropics and exerts remarkable modulations on the tropical and extratropical atmospheric circulations. 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 improving model performance become goals of research groups for numerical simulation of MJO. Details on similar versions of BCC_CSM and their use in climate change projection and short-term climate prediction have been documented in several studies (e.g., Wu et al 2013, 2014; Liu et al 2014, 2015)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
