On modeled and observed warm rainfall occurrence and its relationships with cloud macrophysical properties

Rainfall from low-level, liquid-phase ("warm") clouds over the global oceans is ubiquitous and contributes non-negligibly to the total amount of precipitation that falls to the globe. In this study, modeled and observed warm rainfall occurrence and its bulk statistical relationships with cloud macrophysical properties are analyzed independently and directly compared with one another. Rain is found to fall from ~25% of the warm, maritime clouds observed from space by CloudSat and from ~27% of the warm clouds simulated within a large-scale, fine-resolution radiative convective equilibrium experiment performed with the Regional Atmospheric Modeling System (RAMS). Within both the model and the observations, the fractional occurrence of warm rainfall is found to increase with both column-integrated liquid water mass and cloud geometric depth, two cloud-scale properties that are shown to be directly related to one another. However, warm rain within RAMS is more likely with lower amounts of column water mass than observations indicate, suggesting that the parameterized cloud-to-rain conversion processes within RAMS produce rainfall too efficiently. To gain insight into the relationships between warm rainfall production and the concentration of liquid water within a cloud layer, warm rainfall occurrence is subsequently investigated as a joint, simultaneous function of both cloud depth and column-integrated water mass. While rainfall production within RAMS is largely governed by the availability of liquid water within the cloud volume, rain from observed warm clouds with relatively little column water mass is actually more likely to fall from deeper clouds with lower cloud-mean water contents. The latter, CloudSat-derived trend is shown to be robust across different seasons and environmental conditions; it varies little when the warm cloud distribution is stratified into ascending (day) and descending (night) CloudSat overpass groups. Using temperature differences between RAMS cloud tops and their immediate, surrounding environments as a proxy for cloud-top buoyancy, an attempt is then made to quantitatively investigate simulated warm rain occurrence within the broader context of cloud life cycle. It is found that rainfall likelihoods from RAMS-simulated warm clouds with cloud top temperatures warmer than their surrounding environments more closely resemble the overall CloudSat-derived rainfall occurrence trends. This result suggests that the CloudSat-observed warm cloud distribution is characterized by increased numbers of positively buoyant, developing clouds.