Sensitivity of bow-echo forecasts to ensemble and model configuration

Bow-echo structures, both embedded within quasi-linear convective systems and as singular systems, are often poorly forecast within deterministic numerical-weather-prediction model simulations. However, the spread of bow-echo structures within an ensemble of model runs is not known. This study assesses the inter-ensemble-member sensitivity of the structures' simulated reflectivity and radius of curvature to the following: perturbations in initial conditions using a global dataset; different microphysical schemes; and in response to the use of a stochastic kinetic-energy backscatter (SKEB) scheme. It is found that the ensemble spread decreases, respectively, through the three aforementioned ensemble types. Interestingly, a poor deterministic forecast using a given microphysical scheme and a given set of initial conditions can be somewhat improved in some SKEB ensemble members, suggesting that model error needs to be better accounted for when forecasting these mesoscale phenomena, and that one-shot evaluations of parameterization schemes are dangerous.