Monitoring of Compound Air Pollution by Remote Sensing in Lanzhou City in the Past 10 Years


 Based on satellite remote sensing data acquired by the Ozone Monitoring Instrument (OMI), this study used pixel space analysis, a coefficient of variation, stability analysis, and an atmospheric transmission model to determine the concentration of tropospheric ozone (O3), NO2, HCHO, and SO2 columns in Lanzhou from 2010 to 2019. A series of analyses were carried out on the temporal and spatial distribution of concentration, influencing factors and atmospheric transmission path. The results show that the air pollutants in this area present multi-dimensional characteristics and have a complex spatial distribution. In terms of inter-annual changes, in addition to the increase in the concentration of the HCHO column, the ozone, NO2, and SO2 column concentrations have all decreased over time. In terms of monthly average changes, these four pollutants reached their maximum values in April, December, June, and January, respectively. These four types of pollution had a strong spatial correlation, among which HCHO and SO2 had a significant positive correlation, with a correlation coefficient of 0.76. Many factors affect the Atmospheric Compound Pollution in Lanzhou. Among them, pollutants are closely related to urbanization and to the activities of coal-burning industries. Moreover, temperature, precipitation, and sunshine also have certain effects on air quality. The proliferation of pollutants in Gansu Province was one of the sources of pollutants in Lanzhou, while long-distance transportation in the atmosphere from outside the province (Qinghai, Sichuan, and Shaanxi) also exacerbated the pollution in Lanzhou.


9
Abstract: Based on satellite remote sensing data acquired by the Ozone Monitoring Instrument 10 (OMI), this study used pixel space analysis, a coefficient of variation, stability analysis, and an 11 atmospheric in the Lanzhou area can reach the medium level of SO 2 pollution in winter, which is closely related 69 to seasonal meteorological conditions (Peter et al., 2008). The Lanzhou area is one of the heavy 70 industrial bases and transportation hubs of western China; when coupled with the impact of special 71 topography, these conditions make this area one of the most severely polluted areas in the world 72 (Guo Y. T. et al., 2011). After years of strenuous effort, major breakthroughs have been made in 73 controlling air pollution in Lanzhou, achieving a transformation from a "famous polluted city" to 74 an ecologically civilized city that has successfully employed a model of air pollution control known 75 as the Lanzhou model. This paper systematically analyzes the temporal and spatial distribution of 76 four pollutants and factors that influence their concentrations-ozone, NO2, SO 2 , and HCHO-and 77 discusses the problems related to compound air pollution that have not been analyzed previously. 78 79   80  81  82  83  84  85  86  87  88  89  90  91  92  93  Figure 1 Overview of the study area showing eight counties and districts areas within Lanzhou  94 City with a digital elevation model overlay. An inset map shows the provinces and other administrative 95 area of mainland China and locations of their capital cities.-96

Overview of the study area
The Lanzhou area, located in northwestern China and at the geometric center of Chinese 97 territory, lies roughly between 35-37°N and 102-105°E. As the provincial capital of Gansu Province, 98 Lanzhou serves as the center of provincial politics, economy, and culture. It is also the center of silk 99 production. An important node city in the economic belt and road, Lanzhou also serves as an with precipitation mainly concentrated in summer and autumn, and an average annual rainfall of 105 about 327 mm (Wu et al., 2019). The terrain of this area is higher in the west and south, and lower 106 in the northeast (Figure 1), with the topography mainly composed of mountains and basins. 107 As one of China's important industrial bases, the Lanzhou area is also an important base for 108 petrochemical, biopharmaceutical, and equipment manufacturing industries in China. It is a frontier 109 and an important gateway for the country to implement the policy of opening to the west (Zhang et 110 al., 2014). As of the end of 2014, 156 proven mineral deposits and ore sites had been identified in 111 the area, which were mainly divided into nine categories including non-ferrous metals, rare earths, 112 energy minerals, and precious metals. Among them, the reserves of quartzite as the raw material 113 needed for the ferrosilicon industry were as high as 300 million tons. The development of the mining 114 industry has provided sufficient reserve resources for China, while the coal storage reserves total an 115 estimated 905 million tons, which has greatly promoted the development of the region's mining 116 industry. As of the end of 2019, the permanent population of the region was 3,790,900, an increase 117 of 37,300 over the previous year. The regional gross domestic product was 283.736 billion yuan, an 118 increase of 6% over the previous year; the ratio of the output structure of the above three industries The main task of this satellite is to observe and study the earth's ozone layer, air quality, and its 125 changing climate. In addition to socio-economic factors, the factors affecting the aforementioned 126 gaseous pollutants are also related to the natural factors in the area. Therefore, the data on the socio-127 economic factors selected in this article (regional production value, secondary production value, 128 raw coal consumption, urbanized area, urban construction land, and so on) are from the Gansu 129 Provincial Bureau of Statistics and the Lanzhou Regional Statistical Yearbook. Data related to 130 natural factors (e.g., precipitation, temperature, air pressure, relative humidity, and sunshine 131 duration) come from the Gansu and Lanzhou Regional Statistical Yearbooks; in addition, some 132 natural factor data come from ground-based weather monitoring stations in the Lanzhou area. 133

Data processing 134
The remote sensing data from the Ozone Monitoring Instrument (OMI) were verified by a large 135 number of aviation and ground experiments; the results show that the correlation between the 136 tropospheric and the near-ground pollutant concentrations can reach more than 0.8, which is a 137 significant positive correlation. However, because cloud coverage will have a certain effect on the 138 concentration of pollutants, this study uses daily concentration data over a ten-year period; the 139 amount of available data is relatively large. Therefore, data with a cloud cover greater than 0.2 were 140 removed during data processing and did not affect the final results. To improve the credibility of the 141 pollutant concentration data, the latitude range covered by each data point was expanded by 0.5° 142 when processing the data. The daily ozone, NO2, SO 2 , and HCHO column concentration data 143 downloaded from the NASA official website were processed in batches by Python software; next, 144 data were processed by ArcGIS software (ESRI, Redlands, CA, USA) for raster calculation, 145 interpolation, extraction, and analysis. Finally, the temporal and spatial distribution maps of the 146 aforementioned pollutants were obtained. 147 When analyzing the correlation between these four types of pollution, the year is the unit with 148 spatial correlation method used for analysis. The calculation process is shown in Eq. (1): 149 (1) 150 where r xy represents the correlation between the two gaseous pollutants, and its value is between −1 151 and 1. The closer r xy is to 1, the more significant a positive correlation between the two pollutants 152 will be; the closer r xy is to −1, the more significant a negative correlation between the two pollutants 153  The coefficient of variation was used to analyze and study the spatial stability of these four 159 pollutants using Eqs. (2) and (3) as follows: 160 The univariate linear analysis method was used to study the change trend of these four 169 pollutants in Lanzhou during the past 10 years; it was calculated using Eq. (4) as follows: 170 Finally, the atmospheric transmission model was used to simulate the transmission path of 177 pollutants in the four seasons in Lanzhou in the past 10 years. The specific process was as follows: 178 The position of the next point was obtained by the product of the average velocity of the To create an industrial transfer demonstration zone, together with the Lanzhou area, a major move 237 was made to transfer industries to the Lanzhou New District, including many heavy industry, 238 materials, heating, and power supply companies, thus making the ozone and SO2 column 239 concentrations higher than in other regions. Figure 1 shows that Yuzhong County has relatively high 240 elevation terrain with many mountains and hills, relatively lush vegetation, and a relatively sparse and 97.79 DU, respectively. Moreover, the overall change trend of the former is gentler than that of 354 the latter, with a decreasing trend that was not obvious. Among them, the SO 2 column concentration 355 began to decrease in January, falling to the lowest value in July and August, then rising from 356 September to a peak in December and January of the following year; that is, the SO 2 column 357 concentration reached its peak, especially in January. The ozone column concentration showed opposite. That is, the trough of the ozone column concentration occurred exactly at the peak of the 363 SO 2 column concentration, while the concentration of the former reaches its peak in March and 364 April, the ozone column concentration reaches the highest in the spring, and the latter drops to the 365 lowest in July and August; that is, a valley occurs in summer. However, the monthly average changes 366 of NO 2 and HCHO column concentrations were not very obvious when compared when those of 367 ozone and SO 2 concentrations. Among them, the NO 2 column concentration had two troughs in a 368 year. That is, the NO 2 column concentration reached a minimum in February and August, while  This study combined spatial pixel analysis by processing the concentration data of these four 508 pollutants in the study area to examine the mutual influences and contributions between the 509 pollutants in Lanzhou area. Figure 7

543
Origin software was used to analyze the correlation between natural factors and the 544 concentrations of pollutants in the Lanzhou area from 2010 to 2019 using statistics. Table 1 shows 545 the correlations between the concentrations of various pollutants and various natural factors in the 546 study area. Table 1 shows that, except for the increase in the HCHO column concentration over time, 547 the national binding indicators have been reduced. This shows that with the strengthening of air 548 pollution control in Lanzhou in recent years, the air quality has improved significantly. In addition 549 to the relatively weak relationship between temperature and HCHO, a strong correlation was 550 observed between temperature and the other three pollutants, specifically SO2, NO 2 , and ozone, with 551 correlation coefficients of −0.54, −0.89, and 0.67, respectively. Among them, ozone is a secondary 552 pollutant frequently formed as a product of photochemical reactions; higher temperature will 553 accelerate the production rate of ozone, thereby increasing the concentration of the ozone column 554 in the atmosphere. Studies have found that a significant correlation exists between the content of 555 nitrogen oxides in the air and the length of its life span and temperature (Ma et al., 2020). The 556 content of NO2 in the atmosphere in Lanzhou in winter was significantly higher than that in summer. 557 The high summer temperatures promote photochemical reactions involving nitrogen oxides in the 558 atmosphere and shortens its survival time, which is not conducive to the accumulation of NO 2 . Low 559 winter temperatures do not favor the conversion of NO 2 into other products, which causes it to 560 accumulate in the atmosphere. Therefore, the concentration of NO 2 in winter is normally 561 significantly higher than in the other three seasons. Precipitation and relative humidity have a greater 562 impact on the concentration of the SO 2 column, with correlation coefficients of −0.57 and −0.63, 563 respectively. With more precipitation, the humidity will increase. Rainwater will dilute some of the 564 pollutants in the atmosphere and cause them to be removed during precipitation. This will reduce 565 the content of some pollutants in the atmosphere, which may also be one of the reasons for the low 566 concentrations of SO 2 and NO 2 observed in summer. In addition, a significant correlation also exists 567 between the duration of sunshine and the ozone and HCHO column concentrations, with correlation 568 coefficients of 0.74 and −0.47, respectively. Specifically, the longer the duration of sunshine with 569 increased amounts of light radiation, catalyzes the formation of ozone from precursors such as VOC S , 570 NO X , and HCHO, as the precursors of ozone will also consume part of those precursors as ozone is 571 generated in the process; therefore, the number of sunshine hours also has a certain effect on the 572 concentration of the HCHO column. and built-up area all had obvious correlation with the SO 2 and NO 2 column concentrations, whereas 589 the correlations with the other two pollutants were relatively weak. A significant correlation was 590 observed between raw coal consumption and NO 2 , with a correlation coefficient of 0.79. Studies 591 have shown that a large amount of NO X will be produced during the combustion of coal, and under 592 certain conditions, NO x will be converted into a precursor of NO 2 , making it accumulate 593 continuously in the air, eventually causing the NO 2 content in the atmosphere to gradually rise. The  Province, followed by Qinghai Province. The pollutants in this area in summer are mainly imported 648 from southern Gansu Province, accounting for 67% of all air masses, followed by Shaanxi Province 649 which accounts for 23%. In autumn, polluted air mainly came from Qinghai Province (41%), 650 followed by Pingliang City and Sichuan Province (31% and 28%, respectively). The winter 651 pollutants primarily come from Qinghai Province (69%), followed by the southern cities of Gansu 652 Province (31%). In general, the air in Lanzhou area is mainly transmitted from Gansu Province, 653 followed by remote transmission from Qinghai Province, while the influence of Sichuan and 654 Shaanxi provinces cannot be ignored; this also shows that long-distance atmospheric transport, a 655 natural factor, has an increasingly prominent impact on regional pollutants. 656 temperature has a relatively weak relationship with HCHO, and has a greater impact on the other 686 three pollutants. Precipitation affects all four pollutants. Relative humidity has a greater impact on 687 SO2 and NO 2 ; meanwhile, the duration of sunshine has a greater impact on ozone while also having 688 a certain impact on HCHO and NO 2 . Among the economic factors influencing air quality, these four 689 pollutants were greatly affected by the speed of urban and economic development; in addition, a 690 certain connection existed between HCHO and the development of light industry. 691 6) Through the simulation of atmospheric movement trajectory, it is found that the pollutants in 692

Conclusions
Lanzhou area are mainly affected by Gansu Province. In addition, air pollutants from Qinghai, 693 Sichuan and Shanxi provinces have been transported remotely from the atmosphere, which has 694 exacerbated the pollution in Lanzhou. 695

696
The data sets used or analyzed during the current study are available from the corresponding author 697 on reasonable request. 698 Ethical approval and consent to participate Not applicable.

699
Consent for publication Not applicable.

Declaration of competing interest 701
The authors declare that they have no known competing financial interests or personal 702 relationships that could have appeared to influence the work reported in this paper. 703