Microphysical characterisation of west African MCS anvils

Deep convection is the ultimate source of tropical upper tropospheric extended clouds usually called tropical anvils. These anvils may produce precipitation (stratiform region of deep convective storms), and also cirrus shields persisting from several hours to several days. The microphysics of crystals in this type of clouds is an important parameter impacting radiation budget, the amount of water stored in ice phase (that may lead to precipitation) within the troposphere and chemical concentrations for both soluble species and species that adsorb onto ice.

During the AMMA field campaign a Special Observing Period was dedicated to the microphysical characterisation of the MCS anvils using aircraft instrumentation (cloud radar, lidar and a suite of microphysical in-situ measurements) in the Niamey (Niger) and Dakar (Senegal) regions. The measurements are combined in order to determine the most representative types of hydrometeors sampled in different parts of the anvils and their density as a function of their maximum diameter. Several analyses that will be detailed in the presentation, show that the predominant precipitation particles above the 0°C isotherm in the stratiform anvil are rimed aggregates. These rimed aggregates seem to get less dense and of smaller diameter when moving rearward of the system towards the cirriform region.

The particle size distribution are found highly variable with temperature, sampled region of the system or geographical regions. A normalization process is then applied producing a relatively invariant shape and implying that it can be represented by a single mathematical shape for all ice clouds, which is consistent with earlier studies.

The bulk microphysical and radiative parameters of the tropical ice anvils produced by deep convection (ice water content, effective radius and fall velocity) are derived from this data set. These parameters generally increase with temperature, in agreement with earlier studies and are found in average smaller in cirriform regions than in stratiform regions. These values are compared with statistical relationships often used in cloud resolving models and general circulation models. Large differences are found, the current parameterizations being unable to reproduce the large values of the considered microphysical parameters.