基于随机物理过程扰动的BMJ积云参数化方案对降水集合预报的影响

基于ARPS模式和随机物理过程参数化扰动（stochastically perturbed parameterization）方法，通过10个2018年6—7月间的降水个例，讨论了针对BMJ积云降水参数化方案下不同参数化扰动方式对降水预报的影响。扰动方式包括扰动BMJ方案的温湿倾向和扰动BMJ方案的温湿参考廓线。试验的结果表明BMJ方案在华东区域的降水预报中存在湿偏差，即预报的降水事件多于相应的观测事件。这一偏差在系统性增加参考廓线湿度后仍然存在。BMJ方案对不同扰动方式的响应存在较大差异。扰动BMJ方案的温湿倾向对降水预报的影响较小，且集合离散度低。扰动BMJ方案的温湿参考廓线对降水预报影响显著，能够大幅增加集合离散度，其中对称的BMJ参考廓线扰动对预报技巧评分改进有限，原因是小雨的湿偏差有所增加，而非对称的BMJ参考廓线扰动（扰动均值大于1.0）能够有效提高预报技巧评分并降低湿偏差。此外，非对称扰动大幅改善了BMJ降水预报初期（0～3 h）的空间分布形态，并且改进了夜间降水预报的强度。非对称扰动评分较高的原因是减少了原BMJ方案在降水预报初期的的大范围虚假预报，避免了大气湿度的大范围下降，保障了预报后期的强降水预报能力。而BMJ方案温湿倾向量级较小则是造成倾向扰动方法效果不明显的重要原因。

Impacts of the stochastically perturbed BMJ cumulus parameterization scheme on the convective-permitted ensemble forecasts

In this study, the impacts of stochastic tendency perturbations and stochastic parameter perturbations on the precipitation ensemble forecasts using the Betts-Miller-Janjic'' (BMJ) scheme were investigated.Next, both stochastic perturbation approaches were implemented on the advanced regional prediction system (ARPS).Then, precipitation ensemble forecasts were evaluated using 10 cases that occurred in eastern China in June 2018 and July 2018.The tendency perturbations involved the temperature and specific humidity tendencies from the BMJ scheme, while the parameter perturbations were applied in the temperature and specific humidity reference profiles of the BMJ scheme.The results showed that the precipitation forecasts when using the BMJ scheme were characterized by wet bias at examined thresholds, thus signifying a greater number of forecast precipitation events than the corresponding observations.In addition, the bias issue remained after using wetter reference profiles, while the responses of the BMJ scheme to stochastic approaches differed substantially.The tendency perturbations bore little impact on the BMJ precipitation forecasts as the forecast precipitation frequency did not significantly change in comparison with the unperturbed BMJ scheme.In the case of using tendency perturbations, the ensemble spread is low.In contrast, perturbing the reference profiles bore great impacts on the precipitation ensemble forecasts.Symmetric perturbations of reference profiles produced a large ensemble spread, yet this approach also increased the wet bias at light rain thresholds, and yielded small improvement on the forecast skill scores.Compared to tendency perturbation approach, asymmetric perturbations (perturbation mean> 1.0) of the reference profiles were conducive to larger ensemble spread, higher forecast skill scores, and smaller wet bias, although the wet bias at large precipitation thresholds increased.Moreover, the asymmetric perturbations substantially improved the precipitation spatial distribution in early forecast stage (0-3 h) and the nocturnal precipitation intensity.The large area of spurious precipitation yielded by the BMJ scheme in the early stage of the forecast substantially dried the air, which in turn suppressed the precipitation intensity in subsequent forecast.Finally, the asymmetric perturbations effectively reduced the spurious precipitation and improved the precipitation intensity forecast.The quantities of temperature and specific humidity tendencies were both small, and this was likely the cause of small impact of tendency perturbations on the BMJ scheme.