Prediction skill and practical predictability depending on the initial atmospheric states in S2S forecasts2023 · Abstract The hypothesis that predictability depends on the atmospheric state in the planetary-scale low-frequency variability in boreal winter was examined. We first computed six typical weather patterns from 500-hPa geopotential height anomalies in the Northern Hemisphere using self-organising map (SOM) and k-clustering analysis. Next, using 11 models from the subseasonal-to-seasonal (S2S) operational and reforecast archive, we computed each model’s climatology as a function of lead time to evaluate model bias. Although the forecast bias depends on the model, it is consistently the largest when the forecast begins from the atmospheric state with a blocking-like pattern in the eastern North Pacific. Moreover, the ensemble-forecast spread based on S2S multi-model forecast data was compared with empirically estimated Fokker-Planck equation (FPE) parameters based on reanalysis data. The multi-model mean ensemble-forecast spread was correlated with the diffusion tensor norm; they are large for the cases when the atmospheric state started from a cluster with a blocking-like pattern. As the multi-model mean is expected to substantially reduce model biases and may approximate the predictability inherent in nature, we can summarise that the atmospheric state corresponding to the cluster was less predictable than others.
季節アンサンブル予報を用いた東京のサクラ開花日予測2018 · <p> This study assessed the flowering-date forecast skill of cherry blossom in Tokyo from 2018 to 2023 using seasonal ensemble forecasts from three numerical weather prediction centers: the Deutscher Wetterdienst, the European Centre for Medium-Range Weather Forecasts, and the Météo-France. First, the optimal seven parameters used in the flowering-date estimation model, developed by Maruoka and Itoh (2009), were determined for Tokyo, based on the period from 1994 to 2017, during which the estimation bias was ±1.91 days. Then, flowering dates were predicted using bias-corrected seasonal ensemble forecast of 2 m temperature as a model input. The root-mean-square errors for the flowering-date forecasts initialized on 1st January, February, and March, averaged over all ensemble members, were about ±8.0 days, ±6.2 days, and ±2.3 days, respectively. The best- or worst-performing center is dependent on the specific cases. The grand ensemble forecast, comprising all forecasts from all single-center ensembles, showed better performance in predicting flowering dates of cherry blossoms than the single-center ensemble forecasts alone. These results suggest that the grand ensemble approach at seasonal timescales holds potential for predicting of the growth of flowers and fruits.</p>
