The Response of Shallow-to-Deep Convective Transitions to Increased Shallow Cloud Populations in the Amazon

While much scientific attention has been given to mature deep convection, there remains a large gap in the understanding of the transition from shallow to deep convection. In South America, climate models struggle to accurately represent the variability and extent of cloud populations, resulting in significant convective and large-scale precipitation biases. To improve the representation of non-local mixing in the convective boundary layer, an Eddy-Diffusivity Mass-Flux (EDMF) scheme was implemented in the Mellor-Yamada-Nakanishi-Niino (MYNN) planetary boundary layer (PBL) scheme in the Weather Research and Forecasting (WRF) model. Subgrid-scale convective plumes of varying diameters are produced by the EDMF scheme and develop into shallow cumulus clouds if the plumes surpass the lifted condensation level (LCL). The objective of this study is to determine how subgrid-scale shallow clouds produced by the MYNN-EDMF PBL scheme impact the total cloud population in the Amazonian rainforest, with emphasis on shallow-to-deep convection transitions. A pair of convective-permitting (2-km grid spacing), month-long, wet season WRF simulations with and without the EDMF scheme are performed for a portion of the GoAmazon2014/5 field campaign period. Measurements from the Atmospheric Radiation Measurement (ARM) user-facility and satellite observations are used to quantify the biases in the simulations. Convection from the control and EDMF simulations is spatially tracked throughout the simulations to investigate the convective response to shallow clouds. For each convective track, environmental characteristics prior to deep convective initiation are examined to better understand how shallow-to-deep transition is influenced by the ambient shallow cloud population. Results indicate that the large-scale synoptic conditions in simulations with and without the EDMF scheme turned on are nearly identical. However, mesoscale rainfall patterns between the simulations diverge drastically, especially overnight. The EDMF scheme significantly increases the frequency of shallow clouds. However, the frequencies of deep clouds are similar between the control and EDMF simulations. This suggests that more frequent shallow cumulus may not enhance deep convective development and precipitation.