The Impact of Surface Synoptic Circulation Types on Air Quality and the PM2.5 Concentration Forecast in Fen-Wei Plains

Abstract

Based on the sea level pressure data from 2014 to 2018， Lamb-Jenkinson（L-J） atmospheric circulation classification scheme was used to classify the surface synoptic circulation types during air pollution seasons in Fen-Wei Plains and their relationships with the PM2.5 concentration and surface meteorological elements were analyzed.Combining the observed PM2.5 mass concentration and mixing layer height， relative humidity， wind speed， retention index， a PM2.5 concentration forecast model was established in each city under different surface synoptic circulation types.The results show： （1） Occurrence frequencies of different surface circulation types vary， under which the distributions of PM2.5 concentration and meteorological elements differentiate also.Under controls of circulation types South （S）， South East （SE）， East （E） and cyclone （C） the ratio of PM2.5 polluted days proportion to PM2.5 polluted days proportion under all synoptic circulation types were more than 1， corresponding to a larger product of the occurrence frequency and the pollution days， which will lead to PM2.5 pollution events easily.（2） In the case of mild or above pollution in Xi 'an， the mixed layer height ranged from 483.5 m to 601.38 m， the relative humidity ranged from 42.94% to 64.03%， the wind speed ranged from 1.45 m∙s-1 to 2.61 m∙s-1 and the retention index ranged from 0.13 to 0.53.Under controls of circulation types S， SE， E and C the average mixed layer height was 488.98 m， the average wind speed was 1.89 m∙s-1 and the average retention index was 0.46.Except for under type C， the mixed layer height was below 500 m on more than 60% of the sites， and the relative humidity over 60% on more than 50% of the sites.The wind speed was below 2 m∙s-1 with a retention index more than 0.4 on more than 60% of the sites under all four circulation types.（3） According to the PM2.5 concentration forecast models for different circulation patterns in each city， there exists a 0~1 level difference between predicted level and measured level about 82% of all samples， where about 37% of all samples were perfectly consistent.The predicted PM2.5 values from 2019 to 2020 were larger compared with the measured PM2.5 values， and the average absolute deviation values were close to the values of 2014 -2018 years， indicating that the forecast accuracy and stability of the forecast models were good.From 2019 to 2020， there exists a 0~1 level difference between predicted level and measured level among 80% of all samples， where about 33% of all samples were perfectly consistent.

Keywords

• synoptic circulation types
• pm2.5 mass concentration
• meteorological factors
• forecast model