Entrainment rates at varying horizontal grid spacing were diagnosed and analyzed based on Large Eddy Model (LES) and Cloud Resolving Model (CRM) simulations of shallow and deep convection. Results show the estimated entrainment rates increase with increasing resolution, and the growth rate is roughly power-law related to resolution. However, all commonly used parameterizations except for the buoyancy sorting scheme fail to reproduce the monotonic increase of entrainment with resolution, but rather a slight decrease instead. For these schemes, it is suggested the power-law fitting formula derived from LES/CRM simulations can be used as a scaling function for high-resolution entrainment correction. Preliminary tests show that the application of entrainment scaling largely reduces the parcel buoyancy and thus the convective available potential energy, favoring the reduction of parameterized convection at high resolution.

中文翻译：

---

参数化夹带率是否与通过云解析模型模拟估计的尺度相关？


基于浅层和深层对流的大涡模型（LES）和云解析模型（CRM）模拟，诊断和分析了不同水平网格间距下的夹带率。结果表明，估计的夹带率随着分辨率的增加而增加，并且增长率与分辨率大致呈幂律关系。然而，除了浮力排序方案之外，所有常用的参数化都无法重现夹带随分辨率的单调增加，而是略有下降。对于这些方案，建议使用 LES/CRM 模拟导出的幂律拟合公式作为高分辨率夹带校正的缩放函数。初步测试表明，夹带缩放的应用很大程度上降低了包裹浮力，从而降低了对流可用势能，有利于高分辨率下参数化对流的减少。