Using Stochastically Perturbed Parameterizations to Represent Model Uncertainty. Part I: Implementation and Parameter Sensitivity

Abstract

Accurately representing model-based sources of uncertainty is essential for the development of reliable ensemble prediction systems for NWP applications. Uncertainties in discretizations, algorithmic approximations, and diabatic and unresolved processes combine to influence forecast s kill i n a fl ow-dependent wa y. An em erging ap proach de signed to pr ovide a process-level representation of these potential error sources, stochastically perturbed parameterizations (SPP), is introduced into the Canadian operational Global Ensemble Prediction System. This implementation extends the SPP technique beyond its typical application to free parameters in the physics suite by sampling uncertainty both within the dynamical core and at the formulation level using “error models” when multiple physical closures are available. Because SPP perturbs components within the model, internal consistency is ensured and conservation properties are not affected. The full SPP scheme is shown to increase ensemble spread to keep pace with error growth on a global scale. The sensitivity of the ensemble to each independently perturbed “element” is then assessed, with those responsible for the bulk of the response analyzed in more detail. Perturbations to surface exchange coefficients and the turbulent mixing length have a leading impact on near-surface statistics. Aloft, a tropically focused error model representing uncertainty in the advection scheme is found to initiate growing perturbations on the subtropical jet that lead to forecast improvements at higher latitudes. The results of Part I suggest that SPP has the potential to serve as a reliable representation of model uncertainty for ensemble NWP applications.

Formulation and use of 3D‐hybrid and 4D‐hybrid ensemble covariances in the Météo‐France global data assimilation system

“…In this context, the appeal of 4D ensemble information at high resolution is growing, since (i) ensembles rely on nonlinear integrations of the full model and (ii) ensembles allow random model errors to be represented through associated model perturbations provided, for example, by stochastic perturbations applied to model parameters (e.g., Ollinaho et al ., 2017). Indeed, the representation of random model errors is an active area of research, with recent progress achieved at several NWP centres (e.g., McTaggart‐Cowan et al ., 2022a; Wimmer et al ., 2022). This is all the more appealing as the representation of model errors is likely to be beneficial not only for the performance of ensemble predictions (e.g., McTaggart‐Cowan et al ., 2022b) but also, for example, for the reliability of EDA and for the realism of DA, through associated improved ensemble‐based background‐error covariances (e.g., Caron & Buehner, 2022).…”