Observing and explaining variations in deep convective cloud top height

Deep convective cloud top height, important for cloud radiative forcing, varies significantly over the world for reasons that are not fully understood. We have analyzed morning cloud heights over tropical oceans using Terra/MISR, focusing on the seasonal cycle. Three peaks in frequency of occurrence appear, similar to those shown previously from space-borne lidars, but are better resolved here owing to better sampling.

The upper peak (near 13 km) shifts upward by 600 m in DJF over JJA. This shift is quantitatively reproduced by the WRF (Weather Research and Forecasting) model run in cloud-resolving mode, which furthermore shows it to be entirely due to the different temperature structure above 13 km. The shift is not predicted by standard parcel theory, but may be explained by an overshooting+mixing process. This indicates that turbulent mixing above the usual anvil cloud level must be considered to predict where anvil tops will be, a result of potential relevance to cloud longwave radiation feedbacks on global climate. Shifts are also observed in the mid-tropospheric peak and will be briefly discussed.