ANALYSIS OF AQI CHANGE CHARACTERISTICS AND CORRELATION WITH PM2.5 AND PM10 IN BEIJING-TIANJIN-HEBEI REGION

The paper analyzed the variation characteristics of AQI and its correlation with PM2.5 and PM10 of in Beijing-Tianjin-Hebei region from July 2015 to July 2018 based on hours of pollutants in Beijing-Tianjin-Hebei region, using AQI calculation method and statistical correlation evaluation method. Results showed that:(1) The air quality compliance rate in Beijing-Tianjin-Hebei region was 67%, the average AQI was 97.6577, and the air quality was good. The distribution frequency of primary pollutants was PM2.5, followed by PM10, which accounts for 78.9% of the distribution frequency of the six major pollutants, indicated that PM2.5 and PM10 had a greater impact on the air quality of Beijing-Tianjin-Hebei. (2) The correlation between AQI and PM2.5 and PM10 was significantly positively correlated. R2 was 0.8225 and 0.7749, respectively, P<0.01, indicated that both showed a greater impact on air quality. (3) AQI and PM2.5 and PM10 showed a gradual decrease trend at 9h-16h, ie 9h highest and 16h lowest. The AQI fluctuated between 94.2816 and 103.3562, indicated that the air quality at 9h-16h was good or slightly polluted. (4) The spatial distribution of AQI, PM2.5 and PM10 was characterized by low northwest and high southeast, and the southeastern part was gradually decreasing from 9h-16h. AQI was negatively correlated with elevation. The higher the elevation, the better the air quality, and the worse the air quality.


INTRODUCTION
With the rapid development of the economy in the Beijing-Tianjin-Hebei region, the problem of air pollution became increasingly serious (Iii et al.,2002). Air pollution became an important environmental issue in the Beijing-Tianjin-Hebei region. In 2012, the national Air Quality Index (AQI) was used to evaluate the urban air quality level.
The AQI classification calculation refers to the new ambient air quality standard (GB3095-2012), mainly using sulfur dioxide (SO 2 ) and nitrogen dioxide ( NO 2 ), carbon monoxide (CO), ozone (O 3 ), and PM10, PM2.5 and other six pollutant concentration values to convert into corresponding indexes, which can be used for environmental status assessment, trend evaluation and retrospective evaluation, providing timely and accurate Air quality. Compared with the previous Air Pollution * Corresponding author:Juanli Jing, Email: jjlgut2008@163.com Index (API), AQI has stricter grading standards. More pollutants are monitored and the evaluation results are more objective (Kang et al.,2017).
In recent years, air quality issues have received increasing attention from scholars at home and abroad. Holland et al.(2004) studied the SO 2 concentration in the central and western regions, and the results showed that the SO 2 concentration decreased continuously in the past 10 years, and the air quality in the region was significantly improved. Du et al. (2017) used AQI to compare the spatial variation of 74 cities in China, analyzed the AQI grade distribution and spatial differences of air quality in the region, and constructed a model of the influencing factors of air quality index. Puustinen et al. (2007) studied the number of airborne particles in European cities. The results showed that urban air pollution was affected by particulate matter, but the average urban air pollution level could not be measured by the concentration of particulate matter at a certain location. Zhang et al.(2011) used GIS technology to study the spatial and temporal distribution characteristics of air quality in China and the Pearl River Delta.
The result showed that air pollution had spatial agglomeration characteristics and significant seasonal variation characteristics. Farzanegan et al.(2011)studied the air quality of 122 countries and combined the new air quality evaluation criteria to rank the air quality of each country in the past 20 years. The results showed that air quality was inseparable from urban industrialization and energy consumption. Xiao et al. (2018) discussed the influencing factors of air quality, and the results showed that population concentration, energy consumption and industrialization had a deteriorating effect on China's air quality.
The existing research mainly focused on the spatial and temporal evolution characteristics and influencing factors of air quality (Jiang et al.,2018;Xu et al.,2019;). Zhan et al. (2014) used air quality monitoring data combined with inverse distance weighted interpolation to analyze the spatial and temporal distribution of AQI in Wuhan and the correlation between AQI and other six major pollutants. The monthly AQI performance of Wuhan City was characterized by obvious spatial and temporal distribution characteristic, and AQI had the greatest correlation with PM2.5. Liu et al. (2016) used AQI and 6 major pollutants as research objects, and used Gini coefficient measurement and spatial statistical analysis method to analyze the characteristics of urban air pollution in China.
The results showed that China's urban air pollution presents significant autocorrelation characteristics, and air pollution was characterized by Geographical aggregation and spatial non-equilibrium characteristics. Chen et al. (2017) analyzed the temporal and spatial distribution characteristics of AQI and air content factors in the Yangtze River Delta urban agglomeration.
The results showed that AQI was the best in summer, the worst in winter, and the spatial distribution was characterized by low south and high north. There was significant correlation between AQI and PM2.5. Pei et al. (2018) used air quality monitoring data combined with spatial autocorrelation analysis to analyze the temporal and spatial distribution characteristics of AQI in Shenzhen.The results showed that AQI was positive spatial autocorrelation, and the primary pollutants were different every year. AQI was the best in summer and the worst in winter.Previous studies have mainly analyzed the temporal and spatial distribution of AQI from the daily, seasonal and annual scales, and rarely involved hourly scales.Therefore, this paper selected the Beijing-Tianjin-Hebei region with developed economy and high industrial population as the research area, and used AQI calculation method and statistical correlation evaluation method to study the characteristics of AQI and analyze the correlation between AQI and PM particle concentration and hourly temporal and spatial distribution.

Study Area Characteristics
The Beijing-Tianjin-Hebei region is located in the eastern

AQI Calculation Method
(1)The formula of AQI can be expressed as  (Jin et al.,2014). (see Table 1.) (2)The formula of IAQI can be expressed as:    (1) (2) X model,i = the true value The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-3/W10, 2020 International Conference on Geomatics in the Big Data Era (ICGBD), 15-17 November 2019, Guilin, Guangxi, China The root mean square error is used to measure the deviation between the observed value and the true value.

AQI Variation Characteristic
The average value of AQI in Beijing-Tianjin-Hebei region during July 2015 to July 2018 was 97.6577. It could be seen from  Figure 2 that the AQI was the second/good distribution frequency (40%), which was 2/5 of the total research scale days.The air quality was the first/optimal distribution frequency, which was 24% and the air quality was the sixth/severe pollution had the lowest frequency and its value was 2%. Overall, the air quality compliance rate was 64% (ie, air quality performance was excellent or good).  The AQI value in the study scale was from 0 to 50, the air quality was excellent, and there was no primary pollutant. It could be seen from Figure.2 that the most frequent distribution of primary pollutants in the Beijing-Tianjin-Hebei region was PM2.5, which was 41.6%, indicating that fine particulate matter dominated air pollution; followed by PM10, which had a value of 37.3%. The distribution frequency of PM2.5 and PM10 was 78.9%, which indicated that the air quality pollution was mainly polluted by fine and coarse particles, and other pollutants had little impact on air quality in Beijing-Tianjin-Hebei region.

Analysis of Correlation Between AQI and PM2.5, PM10
From the above-mentioned distribution characteristics of primary pollutants, the air pollution mainly came from PM2.5  MATLAB was used to program and draw a two-coordinate graph to analyze the hourly changes of AQI, PM2.5 and PM10 from July 2015 to July 2018 in Beijing-Tianjin-Hebei region.
As could be seen from Figure showed a decreasing trend, and the variation of PM10 was larger than that of AQI and PM2.5.The AQI had the lowest value at 16h, which was 94.2816, and the air quality was good.
The highest value was 103.3562 at 9h, and the air quality was mildly polluted.  In addition, AQI, PM2.5 and PM10 were consistent at high and low values, ie the highest overall value at 9h, especially in the southeast, with the lowest at 16h. This may be because the surface temperature dropped rapidly at night, and the temperature of the near-surface air layer dropped slowly, forming an inverse temperature phenomenon. It made the city's polluted air unable to spread upwards, resulting in greater pollution in the morning 9h. In addition, 9h was also the peak period of travel. At this time, the vehicles were dense, and the exhaust emissions were larger than other times.
In order to further study the influence of elevation on air quality, this paper analyzed the relationship between elevation and AQI, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-3/W10, 2020 International Conference on Geomatics in the Big Data Era (ICGBD), 15-17 November 2019, Guilin, Guangxi, China PM2.5 and PM10 based on DEM data.It could be seen from Plants can reduce the impact of dust on the atmosphere by covering the ground, and can also adsorb atmospheric particles to make the particles achieve sedimentation and thus reduce air pollution .Therefore, upgrading industrial technology, integrating urban green space and optimizing landscape patterns can minimize urban air pollutant emissions (Deng et al.,2013).