Filter Likelihood as an Observation-Based Verification Metric in Ensemble Forecasting

In numerical weather prediction (NWP), ensemble forecasting aims to quantify the flow-dependent forecast uncertainty. The focus here is on observation-based verification of reliability systems. particular, at short lead times, errors tend be relatively small compared observation and hence it very important that metric also accounts for observational uncertainties. This paper studies so-called filter likelihood score which deep-rooted in Bayesian estimation theory fits naturally filtering setup NWP. filter score considers errors, mean skill, spread one metric. Importantly, can made multivariate effortlessly expanded simultaneous against all types through operators contained parental data assimilation scheme. Here observations from global radiosonde network satellites (AMSU-A channel 5) are included OpenIFS-based forecasts using different initial state perturbations. Our results show sensitive system quality compares consistently with other metrics such as relationships between error, Dawid-Sebastiani score. conclusion provides a well-behaving metric, truly by including covariances, systems strong foundation theory.