Size-Resolved Evaluation of Simulated Deep Tropical Convection

Abstract

Deep moist convection is an inherently multiscale phenomenon with organization processes coupling convective elements to larger-scale structures. A realistic representation of the tropical dynamics demands a simulation framework that is capable of representing physical processes across a wide range of scales. Therefore, storm-resolving numerical simulations at 2.4 km have been performed covering the tropical Atlantic and neighboring parts for 2 months. The simulated cloud fields are combined with infrared geostationary satellite observations, and their realism is assessed with the help of object-based evaluation methods. It is shown that the simulations are able to develop a well-defined intertropical convergence zone. However, marine convective activity measured by the cold cloud coverage is considerably underestimated, especially for the winter season and the western Atlantic. The spatial coupling across the resolved scales leads to simulated cloud number size distributions that follow power laws similar to the observations, with slopes steeper in winter than summer and slopes steeper over ocean than over land. The simulated slopes are, however, too steep, indicating too many small and too few large tropical cloud cells. It is also discussed that the number of larger cells is less influenced by multiday variability of environmental conditions. Despite the identified deficits, the analyzed simulations highlight the great potential of this modeling framework for process-based studies of tropical deep convection.

How organized is deep convection over Germany?

“…; Senf et al . ; White et al . ): underestimation of rainfall, overestimation of cold cloud coverage and a too large number of small cloud tops and precipitation cores.…”

“…More details are provided in Senf et al . (), including a list of parameters also chosen for the current study. For the reflectivity fields, the approach is slightly modified.…”

“…; Senf et al . ) and the resulting model‐derived BTs are ideally suited for a direct comparison with real satellite observations. The derived synthetic BTs typically have uncertainties in the order of several kelvins due to uncertainties in the specification of radiative properties of cloud particles (e.g.…”

Mesoscale marine tropical precipitation varies independently from the spatial arrangement of its convective cells

“…Our segmentation procedure follows Senf et al . (), but is applied to precipitation and not to brightness temperature. The segmentation procedure is summarized below and illustrated for four distinct convective situations in Figure .…”

Different Representation of Mesoscale Convective Systems in Convection-Permitting and Convection-Parameterizing NWP Models and Its Implications for Large-Scale Forecast Evolution

“…The model top was at 30 km, and 75 vertical levels were used. More details on the simulation setup and an evaluation of some aspects of the simulations can be found in [12,31]. An evaluation of bulk atmospheric quantities, which were represented surprisingly well in ICON-HiRes despite being an untuned model version, as well as specific aspects like the diurnal cycle are given in [16].…”