A set of experiments is described with the Community Atmosphere Model (CAM5) using a two‐plume convection scheme. To represent the differences of organized convection from General Circulation Model (GCM) assumptions of isolated plumes in uniform environments, a dimensionless prognostic “organization” tracer Ω is invoked to lend the second plume a buoyancy advantage relative to the first, as described in Mapes and Neale (2016). When low‐entrainment plumes are unconditionally available (Ω = 1 everywhere), deep convection occurs too easily, with consequences including premature (upstream) rainfall in inflows to the deep tropics, excessive convective versus large‐scale rainfall, poor relationships to the vapor field, stable bias in the mean state, weak and poor tropical variability, and midday peak in diurnal rainfall over land. Some of these are shown to also be characteristic of CAM4 with its separated deep and shallow convection schemes. When low‐entrainment plumes are forbidden by setting Ω = 0 everywhere, some opposite problems can be discerned. In between those extreme cases, an interactive Ω driven by the evaporation of precipitation acts as a local positive feedback loop, concentrating deep convection: In areas of little recent rain, only highly entraining plumes can occur, unfavorable for rain production. This tunable mechanism steadily increases precipitation variance in both space and time, as illustrated here with maps, time‐longitude series, and spectra, while avoiding some mean state biases as illustrated with process‐oriented diagnostics such as conserved variable profiles and vapor‐binned precipitation curves.

中文翻译：

---

简单的对流组织方案在两羽GCM中的作用

使用两水对流方案的“社区大气模型”（CAM5）描述了一组实验。为了表示统一环境中孤立羽流的一般对流模型（GCM）假设与有组织对流的差异，调用了无因次的“组织”示踪剂Ω，以使第二羽流相对于第一羽流具有浮力优势，如Mapes和尼尔（2016）。当无条件携带低夹流羽流时（各处Ω= 1），深对流太容易发生，其后果包括流入深热带地区的过早（上游）降雨，对流与大范围降雨过度，与蒸汽场的关系差，平均状态的稳定偏差，热带变异的弱弱性以及陆地日降水的正午峰值。这些中的一些也显示出其分离的深对流方案和CAM4的特征。当通过在各处设置Ω= 0来禁止低夹带羽流时，可以发现一些相反的问题。在这两种极端情况之间，由降水蒸发驱动的相互作用Ω充当局部正反馈回路，集中了深对流：在近期降雨很少的地区，只会出现夹带高度的羽流，不利于降雨。这种可调节的机制稳定地增加了时空上的降水方差，如此处的地图，时间经度序列和光谱所示，同时避免了某些平均状态偏差，如针对过程的诊断所说明的，例如，变量守恒分布和汽化降水曲线。