建立了一个基于深度学习的地面臭氧集合预报系统,以量化可能的天气形势范围内的污染风险(built a deep-learning surface ozone ensemble forecast system to quantify pollution risks given the range of possible weather outcomes)
深度学习模型强调天气的空间模式,有效地表示了臭氧与气象之间的关系(Deep-learning models accentuating the spatial patterns of weather effectively represented the ozone-meteorology relationship)
深圳 24 小时臭氧预报误差中,天气预报的不确定性贡献了 38%–54%(Weather forecast uncertainties contributed 38%–54% of the ozone forecast errors at 24-hr lead time in Shenzhen)
work flow
利用2D卷积神经网络构建深度学习模型
通过扰动具有广泛天气-中尺度气象变化的区域空气质量模式WRF-GC,生成了一个大型的训练数据集,并使用了集合预报的思想来量化由天气预报的不确定性导致的臭氧预报的不确定性(WRF-GC is an online coupling of the Weather Research and Forecasting (WRF) meteorological model (v3.9.1.1) (Skamarock et al., 2008, 2019) and the GEOS-Chem chemical transport model (v12.8.2) (Bey et al., 2001).)