B1 - Microphysical uncertainties in deep convective clouds and their implications for data assimilation
Other researcher: Dr. Andrew Barrett (Postdoc)
Former researcher: Constanze Wellmann (PhD)
The liquid/ice phase partitioning in convective clouds and the generation of precipitation via the ice phase are currently poorly predicted in operational weather forecast models. Especially, supercooled cloud droplets and frozen hydrometeors are poorly represented although being important for the prediction of severe hailstorms with a high damage potential and for aviation forecasts. In addition, the realistic representation of clouds and hydrometeors is essential for the assimilation of remote-sensing observations related to these quantities.
Within our project, we will identify which microphysical processes have to be included or refined in numerical weather prediction (NWP) models to improve our ability to predict the convective cloud phase and resulting precipitation with a special focus on hail. Besides the direct forecast impact, this shall advance our capability to use the information provided by potential high-impact observations such as radar reflectivity and cloudy satellite radiances and reflectances in the visible and infrared spectrum in convective-scale data assimilation.
The work program will include:
- idealized simulations of deep convective clouds via the warm bubble method in which microphysical parameters (in particular related to heterogeneous ice formation) and initial conditions are systematically varied to evaluate the related sensitivity of the relevant micro- and macrophysical processes
- realistic simulations of severe hailstorms in the past for which comprehensive observational data (hail on the ground, satellite and ground-based remote sensing data on the vertical and temporal evolution of cloud phase, condensate amount and precipitation) are available with the objective to evaluate and improve the model parameterizations related to cloud phase and solid precipitation formation
- perturbed simulations of events with severe convective precipitation on the full COSMO-DE domain and the investigation of sources of cloud and precipitation errors using an ensemble sensitivity analysis to quantify the effects of initial condition errors on forecasts of severe convection.
The tool for this project will be the COSMO model, which will be employed in convection-resolving grid spacing and in different setups, including versions with the double-moment cloud scheme, parameterizations of aerosol-cloud interactions within COSMO-ART, and the KENDA/COSMO ensemble data assimilation system.
The project shall facilitate collaboration of experts in cloud and aerosol microphysics, meteorological extreme events, ensemble data assimilation and diagnostic tools to reach better understanding of errors sources and uncertainties of clouds and precipitation forecasts for the advancement of both convective-scale modeling and data assimilation.