基于最优TS评分和频率匹配的江苏近海风速订正

近海风速预报在沿海防灾减灾、航运安全、风能评估等方面具有重要应用。本文基于欧洲中期天气预报中心数值预报 （ECMWF） 数据，将最优 TS 评分 （OTS） 与频率匹配 （FMM） 两种方法应用于江苏近海风速预报订正，对比了两种订正 方法的优缺点，分析了不同时间训练样本量以及空间站点观测资料数量对订正效果的影响。结果表明：在相同长度的训练期下，风速等级越小，时间训练期样本越多，预报订正值越稳定；对于 7 级以上高风速等级的预报，OTS 与 FMM 的 TS 评分差异不大， 但 FMM 订正更加稳定且频率偏差 （BIAS） 更接近 1；扩充站点观测资料可有效提升预报订正性能；对于 7 级以上灾害性大风，最佳的订正方案是采用 FMM 方法结合全年时间训练样本和扩充的站点观测资料。此外，本文还对比了不同风速情形下，订正前后 BIAS 和 TS 评分对于时间训练样本量、空间站点观测数量以及订正方法 （OTS 和 FMM） 的敏感性的相对大小。

Jiangsu offshore wind speed calibration based on optimal threat score and frequency matching

Offshore wind speed forecasts have wide applications in coastal disaster prevention and mitigation, shipping safety, and wind energy assessment. This study applies the optimal threat score (OTS) and frequency matching method (FMM) respectively to Jiangsu offshore wind speed calibration with data of the European Centre for Medium-Range Weather Forecasts (ECMWF). The strengths and weaknesses of the two calibration methods are compared, and the impact of using different number of temporal training samples and spatial station observations is analyzed. Results show that under the same length of training period, lower wind speed grades have more temporal training samples and therefore have more stable calibrated forecast values. For high wind speed larger than grade 7, there is little difference between the threat scores (TSs) of OTS and FMM, although FMM is more stable and has frequency bias(BIAS) closer to 1. Extending the number of station observations can improve forecast calibration quality effectively. The best choice to calibrate disastrously-strong wind speeds (grade,7) is the combination of FMM with annual training samples, and extending the number of station observations. In addition, the relative sensitivity of BIAS and TS before and after calibration to the number of temporal training samples, spatial station observations and calibration methods (OTS and FMM) are compared under different wind speeds.