Subgrid Scale Physics in 1-Month Forecasts. Part II: Systematic Error and Blocking Forecasts

Abstract The capability of blocking prediction is investigated with respect to four models of different subgrid scale parameterization packages, which were described in Part I. In order to assess the capability, blocking indices are defined, and threat and bias scores are set up for the predicted blocking index against the observation. Applying this evaluation scheme to the dataset of one-month forecasts for eight January cases, we conduct a study on the performance of blocking simulation. First, it is immediately disclosed that the systematic biases in this forecast set are overwhelmingly large, so that the blocking index has to be adjusted to this bias. One, of the major issues, suggested by Tibaldi and Molteni, is whether the systematic bias is generated by the failure of blocking forecasts. Overall, this study supports this assertion, despite the different definitions of blocking. The study also reveals that the A-model is inferior to the other three models, such as the E-model, with regard to blockin...