| In China,air quality has always been one of the major challenges facing.In 2013,most areas of my country were attacked by haze,which was exposed by the media on a large scale and entered the public’s view.Scholars at home and abroad began to study the causes,sources and characteristics of pollutants of haze.Some scholars introduced perception and behavior,and deeply analyzed the public’s risk perception,response behavior,willingness to pay and willingness to green consumption.The public’s awareness of environmental protection is also constantly awakening.At the same time,the state and environmental protection departments have also issued various policies to deal with haze,but the governance of haze not only needs the strong control of the state,but also needs public’s participation.Although there have been a large number of literatures on the public perception and attitude to help the environmental protection department understand the public’s thoughts,the data obtained from the questionnaire survey method is very limited.Therefore,this paper used the network data to study the haze issue from the perspective of the public.Based on the monitoring data,Baidu Index and microblog data this study analyzed the temporal and spatial distribution characteristics of AQI index and other six pollutants（PM2.5,PM10,SO2,CO,NO2 and O3）concentrations in Xi’an from 2014 to 2019 by using descriptive statistics and spline spatial analysis method.Baidu search index was used to represent the public’s haze concern,and the relationship between public concern and AQI index of monitoring data was discussed.Finally,word frequency analysis,social semantic network analysis and sentiment analysis in text analysis were used to systematically study the public’s perception of haze and emotion expression.The purpose of this study is to help the government better understand the public concern and demand on haze,so as to formulate more appropriate policies and measures to control haze and improve public participation.The results of this paper are as follows.（1）The air quality index of Xi’an is on a downward trend,the pollution level is decreasing,and the air quality is improving.From the perspective of annual scale,AQI index,the annual average concentration of PM2.5,PM10,SO2,CO showed a downward trend,and the annual average concentration of NO2,O3 showed an upward trend.During the study period,the number of days with good air quality accounted for 54.67%,71.23%,52.73%,48.77%,62.14%and 61.64%of the total monitored days,respectively.The number of days with slightly polluted air quality or above accounted for 45.33%,28.77%,47.27%,51.23%,37.86%and 38.36%respectively.From the perspective of seasonal scale,the AQI in spring,summer and autumn are similar,which are far lower than those in winter.The seasonal variations of PM2.5,PM10,SO2,CO and NO2 showed a "U" shape,with the lowest values in summer and higher values in winter and spring.The seasonal variation diagram of O3 showed an"inverted U" shape,with the highest concentration in summer and lower in winter and spring.From the perspective of spatial scale,the 6-year average of AQI index increased from southwest to northeast and southeast.The low concentration areas of PM2.5 and PM10 were in the northwest and southwest,and the high concentration areas were in the northeast and southeast.The 6-year average concentration of SO2 increased with "N"shape from southeast to northwest.The 6-year average concentration of NO2 increased from northeast and southwest to the center with the shape of "aggregation".The low concentration center of CO was located in the central urban area.The variation trend of O3 concentration value was decreasing from northwest to northeast and southeast.（2）The public’s attention and perception of haze is more intense in the high incidence season of haze,but the degree of attention has declined in recent years.From 2014 to 2017,the public’s attention to haze in Xi’an increased year by year,and then decreased from 2017 to 2019.From the perspective of annual change,the total search volume in 2017 is the highest.The coefficient of variation and the concentration index of the rose show that the public’s haze attention distribution in 2016 is the most concentrated,and the difference between high value and low value is obvious.From the change of the year,the public’s search index shows a "U" shape,with two high peaks in spring and winter,and the lowest in summer.In 2014,the number of public micro blog posts showed a significant "double peak",and in 2019,the number of public micro blog posts showed a "U" shape,which was much more than that in 2019,indicating that compared with 2014,the public discussion on haze decreased.The number of microblogs posted by the public in spring,autumn and winter is significantly more than that in summer.（3）Public’s attention to haze is affected by air quality and other factors.The public’s attention to haze has a good fitting effect with the change trend of the AQI of Xi’an,and Pearson correlation test also shows that there is a significant positive correlation between the two,that is,with the aggravation of air pollution,the number of searches for haze by the public also increases.The lag analysis shows that the public is highly alert and increasingly sensitive to the haze.At the seasonal scale,the AQI in spring,autumn and winter are significantly positively correlated with the 6-year average change of haze search index,while the correlation is not high in summer,indicating that the public’s attention in summer will be affected by other factors.The Baidu index and AQI index after the de-trending treatment are significantly positively correlated,but the reduced correlation indicates that the public’s attention is affected by the short-term fluctuations of weather,but the influence is small.（4）The public’s perception of haze has clear concerns,but they lack the awareness of participation in haze governance.The content mining software ROST CM6 was used to carry out high-frequency word statistics,social semantic network analysis and sentiment analysis on the preprocessed microblog data.The results showed that the changes of high frequency words such as"haze","Xi’an","weather","pollution" and "serious" were small.In 2014,there were"high temperature","temperature","meteor shower","fireworks","firecrackers","maroons" and other unique high-frequency words.In 2019,words such as "sand","sand dust","willow catkins" and "poplar" appeared.It shows that the focus of public’s attention is mainly focused on the haze,but it will also change at different times.The social semantic network graphs of the four seasons in 2014 and 2019 allcentered on "haze" and"Xi’an",and haze connects almost all related words.However,the public seldom mentions their own participation in reducing pollutants in haze control.The success of haze control requires the joint efforts of the state,enterprises and the public,and publicity can be strengthened to call for more public participation.（5）The public’s sentiment on haze is mainly positive,with obvious seasonality.From the perspective of the public mood,the public has a positive mood towards the haze,with less negative mood,and the public with very negative mood towards the haze is relatively small,indicating that the public is not excessively panic due to the haze problem.Summer is the season with the least haze,while winter is the season with frequent haze,so the public mood will also change accordingly,which has obvious seasonality. |
PDF Full Text Request