The microphysical properties of a long-lasting heavy fog event are examined based on the results from a comprehensive field campaign conducted during the winter of 2006 at Pancheng （32.2°N, 118.7°E）, Jiang...The microphysical properties of a long-lasting heavy fog event are examined based on the results from a comprehensive field campaign conducted during the winter of 2006 at Pancheng （32.2°N, 118.7°E）, Jiangsu Province, China. It is demonstrated that the key microphysical properties （liquid water content, fog droplet concentration, mean radius and standard deviation） exhibited positive correlations with one another in general, and that the 5-min-average maximum value of fog liquid water content was sometimes greater than 0.5 g m-3. Further analysis shows that the unique combination of positive correlations likely arose from the simultaneous supply of moist air and fog condensation nuclei associated with the advection of warm air, which further led to high liquid water content. High values of liquid water content and droplet concentration conspired to cause low visibility （〈50 m） for a prolonged period of about 40 h. Examination of the microphysical relationships conditioned by the corresponding autoconversion threshold functions shows that the collision-coalescence process was sometimes likely to occur, weakening the positive correlations induced by droplet activation and condensational growth. Statistical analysis shows that the observed droplet size distribution can be described well by the Gamma distribution.展开更多
文摘利用2017—2021年的ERA5再分析资料和京津冀国家站地面资料,结合多种机器学习方法建立预报模型,开展轻雾、大雾客观预报。探讨了再分析资料、地形因素的影响,并结合多模型集成、统计消空进一步优化模型。结果表明:(1)XGBoost(eXtreme Gradient Boosting)、LightGBM(Light Gradient Boosting Machine)、随机森林等集成学习方法的预报效果均优于决策树方法;(2)在引入ERA5再分析资料、地形建模后,XGBoost、LightGBM模型的预报性能显著提高。相比仅使用地面要素建模,大雾预报的TS(Threat Score)提升了30%、32%,达到0.52、0.49,命中率分别为0.62、0.87。此外,经过多模型集成后,轻雾、大雾预报的TS提升到了0.51、0.54;(3)2022年秋季一次大雾过程中,本方法提前72 h准确预报了京津冀地区的大雾,其中以LightGBM模型表现最好。0~72 h轻雾预报和0~36 h逐小时大雾预报的TS均达到0.3,预报准确率、时效性均优于ECMWF(European Center for Medium Weather Forecasting)模式。
基金mainly provided by the National Natural Science Foundation of China (Grant Nos. 40537034 and 40775012)the Natural Science Fund for Universities in Jiangsu Province(Grant Nos. 06KJA17021 and 08KJA170002)+1 种基金the Meteorology Fund of the Ministry of Science and Technology [Grant No. GYHY (QX) 2007-6-26]the Qing-Lan Project for cloud-fog-precipitation-aerosol study in Jiangsu Province and the Graduate Student Innovation Plan in the Universities of Jiangsu Province (CX09B 226Z)
文摘The microphysical properties of a long-lasting heavy fog event are examined based on the results from a comprehensive field campaign conducted during the winter of 2006 at Pancheng （32.2°N, 118.7°E）, Jiangsu Province, China. It is demonstrated that the key microphysical properties （liquid water content, fog droplet concentration, mean radius and standard deviation） exhibited positive correlations with one another in general, and that the 5-min-average maximum value of fog liquid water content was sometimes greater than 0.5 g m-3. Further analysis shows that the unique combination of positive correlations likely arose from the simultaneous supply of moist air and fog condensation nuclei associated with the advection of warm air, which further led to high liquid water content. High values of liquid water content and droplet concentration conspired to cause low visibility （〈50 m） for a prolonged period of about 40 h. Examination of the microphysical relationships conditioned by the corresponding autoconversion threshold functions shows that the collision-coalescence process was sometimes likely to occur, weakening the positive correlations induced by droplet activation and condensational growth. Statistical analysis shows that the observed droplet size distribution can be described well by the Gamma distribution.
