LEM simulations of Doppler radar data from frontal cirrus: constraining vertical velocities and fallspeeds

Wind shear is an almost ubiquitous feature of the troposphere at cirrus altitudes, but there have been few studies focused on investigating its effects on cirrus clouds. In this study we consider a case of strongly sheared frontal ice cloud, which was observed by the Chilbolton radar. This was simulated using the Met Office large‐eddy model (LEM), which included a fully integrated radiation code. It is well known that there can be significant variations between results from different cirrus models, which makes a comparison of the simulations with observations important. So, radar observations of ice water content (IWC) were simulated directly within the framework of the LEM, and true and simulated IWC observations were compared, as well as their Fourier transforms and probability density functions (p.d.f.s). This showed that the LEM was capturing reasonably well the horizontally averaged IWC profile, and also at upper levels the variability in the IWC field, at scales of less than approximately 14 km. At lower levels the IWC field was too homogeneous.

Varying the shear within the LEM simulations showed that shear had little effect on the mean IWC profile, but it increased the mixing and so the homogeneity of the IWC field. P.d.f.s of IWC derived from the simulated radar data compared well with observed p.d.f.s, except that the largest IWCs were not captured and there was too little variance at lower levels. Comparing p.d.f.s of IWC and total water for runs with and without shear clearly showed the effects of shear‐induced mixing. LEM results also showed that the variation of the correlation of IWC with vertical separation is shear dependent and initially linear, with or without shear. In this case, shear had little effect on the top‐of‐atmosphere and surface fluxes, or the within‐cloud heating rate profile, but this would be significantly different for a more patchy cloud.

A modified case‐study, which allowed Kelvin–Helmholtz wave breaking, showed that wave breaking can significantly affect the microphysical processes, by increasing nucleation, deposition and sublimation rates. Wave breaking can occur in thin layers of limited horizontal extent, which would not be resolved by a global model. These effects were most significant when there was no large‐scale uplift and the vertical velocities from the wave breaking formed a cirrus cloud which did not otherwise occur.