斜剖面图的设计及在PM2.5中期潜势预报中的应用

本文将强降水潜势预报的“配料法”原理引入到环境预报中，选取ECMWF数值预报产品中的700hPa相对湿度、1000hPa风向风速和850hPa温度作为气象要素“配料”，沿占比最多的冷空气输送路径穿过华北重污染区和上海，设计制作了未来10d气象要素“配料”时空斜剖面图，并具体定义了业务中常见的典型污染过程（冷空气输送、干静稳和湿静稳累积）和清洁过程（冷空气清洁、湿沉降清洁）共5种图形的识别特征.该斜剖面图由于融合了单站时序图和单时次空间面图的优点，无需一张张翻阅天气形势图，显著缩短了看图分析时间；斜剖面图还建立了气象要素图形特征、污染气象条件和PM_2.5污染潜势三者的对应关系，通过图形特征可以快速掌握区域内天气形势演变情况，识别出上述典型污染和清洁过程，最终得出PM_2.5中期污染潜势，便于在业务中应用和推广.一次典型污染个例应用显示，斜剖面图和污染实况的图形型态特征非常相似，斜剖面图中静稳累积和冷空气输送过程图形特征辨识度高，区域和单站提前5d的预报结论和实况较为一致；进一步评估显示，2018年冬季期间预报时效5d以内的TS评分维持在0.32~0.50区间，未来10d的预报准确率维持在74.4%~86.7%区间.

Design of oblique section and its application to PM2.5 mid-term potential forecast

“Ingredients-based forecast methodology” in intensive precipitation potential forecast was adapted to environmental meteorological forecast. The ingredient-based oblique section for the next 10days was produced with RH at 700 hPa, wind direction and wind speed at 1000 hPa and temperature at 850 hPa from ECMWF numerical forecast products as ingredients. The oblique section was designed to be along the predominant cold air transport pathways which passed through the heavily polluted area in North China and Shanghai. 5recognition features were defined specifically for cold air transport type, dry stagnant type, wet stagnant type (3polluted types), cold air clean type and wet deposition clean type (2clean types). This oblique section combined the advantages of a single station’s time series diagrams and spatial graphs at a single time, thus significantly reduced the time of chart analysis. The oblique section also produced the relationship among meteorological pattern characteristics, pollution meteorological conditions and PM_2.5 pollution potentials. The evolution of synoptic patterns was easily recognized as typical polluted type or clean type, which led to the PM_2.5 mid-term potential forecast. Thus, this ingredient-based oblique section could be applied and promoted to operational environmental forecast. The model was applied to a typical pollution case. The characteristics of oblique section and actual pollution observation of this case were very similar. The stagnant pattern and cold air transport pattern were both identified. The 5-day forecast for the entire region and single stations were consistent with the observations. Further assessment showed that the TS score within 5 days of the forecast duration in the winter of 2018 was maintained in the range of 0.32~0.50, and the forecast accuracy for the next 10days was maintained in the range of 74.4%~86.7%.