Skip to main content
NCBI home page
Springer Nature - PMC COVID-19 Collection logoLink to Springer Nature - PMC COVID-19 Collection
. 2022 Jun 1;15(9):1635–1646. doi: 10.1007/s11869-022-01209-9

Statistical analysis of the temporal change of PM10 levels in the city of Sivas (Turkey)

Zinnur Yılmaz 1,✉, Mustafa Bünyamin Karagӧzoğlu 1
PMCID: PMC9155192  PMID: 35668745

Abstract

The objective of this study is to statistically examine the variation of PM10 values measured at three stations in the center of Sivas between the years 2016 and 2020. Hourly PM10 measurement values were taken from three different stations (İstasyon Kavşağı, Meteoroloji, and Başöğretmen AQMSs) in the city center. Then the mean values of the measurements obtained between 2016 and 2020 were compared according to the years and the stations, as well as with the limit values given in the Regulation on Air Quality Assessment and Management(RAQAM). Analyses of variance were conducted to determine any differences between PM10 levels and 24-h limit values of PM10 for Turkey and between PM10 values of stations over the years. Considering the 5-year mean values, the mean value of all PM10 concentrations measured in the city center was calculated as 56.36 µg/m3. No statistical difference was found between the PM10 values measured in 2017 and 2018 at the İstasyon Kavşağı AQMS, and the comparisons of PM10 between stations over the years showed no difference between the Meteoroloji AQMS and the Başöğretmen AQMS in 2019 and 2020. The Spearman’s rank-order correlation results of PM10 over the years among the stations in the city showed that the strongest relationship was a moderate one between the years 2019 and 2020 with regard to the İstasyon Kavşağı AQMS. Probable dust transports were examined for the days when PM10 was at its highest, and the conclusion was that desert dust coming from the continent of Africa (south) to the center of Sivas had been effective.

Keywords: Air pollution, PM10, SPSS, Turkey

Introduction

Air pollution is the presence of foreign substances in the atmosphere in solid, liquid, and gaseous forms with concentrations and durations that can harm human health, life, and ecological balance (Müezzinoğlu 2004). Air quality in big cities is inversely proportional to heating, uncontrolled traffic increase, urban sprawl, growth in urban population, decrease in urban forests, and increase in traffic emissions, and all these conditions significantly contribute to increasing air pollution worldwide (Samet et al. 2000; Jacob and Winner 2009; Kulshrestha et al. 2009; Shen et al. 2010; Dall’Osto et al. 2012; Rashki et al. 2013; Pascal et al. 2014; Kim et al. 2015; Von Schneidemesser et al. 2015; Sharma et al. 2016). Air pollution negatively affects the nervous system (Genc et al. 2012), digestive system (Kaplan et al. 2013), and even urinary system (Calderon-Garciduenas et al. 2004) in humans. Air pollution is the cause and aggravating factor of many respiratory diseases such as chronic obstructive pulmonary disease (COPD) (Calderón-Garcidueñas et al. 2008; Oberdörster et al. 2009), asthma (Peters et al. 2006; Oberdörster et al. 2009), and lung cancer (Watson 2006; Craig et al. 2008; MohanKumar et al. 2008; Gold and Samet 2013).

Particulate matter (PM), an air pollutant characterized by the wide physicochemical diversity of atmospheric aerosol components and sources, is the main pollutant in the air (Moreno et al. 2006; Gieré and Querol 2010; Calvo et al. 2013). PM consists of micro-sized particles suspended in the air. Inhalable PM is categorized according to its size (diameter) as coarse (smaller than 10 mm, called PM10), fine (smaller than 2.5 mm, called PM2.5), and very fine (smaller than 0.1 mm) (Mukherjee and Agrawal 2017). PM shows variation in terms of physical properties and chemical compositions depending on meteorological conditions and sources of emission (Seinfeld and Pandis 1998). Particulate matter produces different effects on health and climate depending on the concentration of its various components. The main components of particulate matter make up a few percent or even more of the total mass of PM. Depending on location and sources of emission, carbonaceous particles (organic carbon—OC, elemental carbon—EC, and carbonate carbon—CC) make up 50% of the total mass of PM10 and 80% of the total mass of PM2.5, which shows a great biogenic effect. Additionally, inorganic ions contribute 40% of the mass of particulate matter, and trace elements generally represent less than 1% of PM (Szidat et al. 2004; Styszko et al. 2015).

Atmospheric PM10 originates from cars, factories, and fossil fuels and is emitted into the atmosphere (Chen et al. 2013; Kim et al. 2019; Yun and Yoo 2019). Previous studies have shown that particulate matter in the air has an impact on climate change, biogeochemical cycles in ecosystems, cloud formation, visibility, and human health (Nriagu 1988; Nriagu and Pacyna 1988; Dockery et al. 1993; Dockery and Pope 1996; Husar et al. 1997; Broecker 2000; Ramgolam et al. 2009).

Today, most of the world’s population is concentrated in urban areas, generally located close to industrial zones. Therefore, determining the share of industrial sources is a great concern as emissions of air pollutants may negatively affect human health by adding toxic elements and compounds to atmospheric pollution (Pope III 2007).

The first objective of this study is to evaluate the temporal nature of PM10 in Sivas between 2016 and 2020 using surface measurements at three locations. It is examined whether the places where these measurements are made comply with the air standards in Turkey or not. The second objective is to statistically examine the variation of PM10 values measured at three stations (according to sources: industry, urban, and traffic) in the center of Sivas between the years 2016 and 2020; examine the similarities, differences, and relationships between and over the years; and determine where the PM10/dust transport to the city center originates from.

Materials and methods

In this study, the objective of which is to figure out the differences or similarities of hourly PM10 concentrations between 2016 and 2020 at three air quality monitoring stations in Sivas; hourly data was converted into daily mean data in order to examine PM10 changes on a yearly basis. Analyses of variance (The Wilcoxon signed-rank test and the Kruskal–Wallis test) were conducted to determine any differences between PM10 levels and 24-h limit values (µg/m3) (Table 1) of PM10 for Turkey and between PM10 values of stations over the years.

Table 1.

PM10 limit values for Turkey by years (µg/m3) (Zeydan 2021)

2015 2016 2017 2018 2019 2020
24 h 90 80 70 60 50 50
Year 56 52 48 44 40 40
Open in a new tab

Along with analysis of variance, post hoc multiple comparison tests were used to analyze changes (differences) in pollutant levels in terms of air pollution between 2016 and 2020 and when this pollutant was at their highest and lowest levels over the past 5 years.

PM10 in the regulation on air quality

The current limit values for the PM10 parameter in Turkey are determined according to the provisions of the Regulation on Air Quality Assessment and Management (RAQAM). In Turkey, limit values for air pollutants have been reduced starting from 2009; the limit values for PM10, which were 300 µg/m3 per day and 150 µg/m3 per year in 2009 and previous years, were reduced in equal measure until 2014 (Zeydan 2021). In 2019, the PM10 limit value was made equal to those used in European countries. Since 2019, PM10 limit values in Turkey have been 50 µg/m3 and 40 µg/m3 for daily and annual periods, respectively (Official Gazette of the Republic of Turkey 2008). Table 1 shows the change in the limit values for PM10 by years. The Regulation on Air Quality Assessment and Management states that the daily limit value for the PM10 parameter cannot be exceeded more than 35 times a year. As the limit values for PM10, the World Health Organization has recommended 50 µg/m3 as the daily mean and 20 µg/m3 as the annual mean (European Environment Agency 2020).

Location of the study

With a population of 635.889, according to data from the Turkish Statistical Institute for 2020 (Turkish Statistical Institute (TURKSTAT) (2021)), and a surface area of 28.488 km2, Sivas is the second largest city of Turkey in terms of land. The city experiences a harsh, continental climate. Winters are cold and severe with a lot of snowfall, and the city is covered in snow for a mean of 4–5 months. Summers in the city are hot, dry, and short, and the spring and autumn months see lots of rainfall (Lale 2008).

There are three air quality measurement stations in Sivas. Table 2 shows the coordinates of the three air quality monitoring stations in Sivas, and Fig. 1 shows the locations of the stations on the map.

Table 2.

Geographical coordinates of the air quality monitoring stations (T.R. Ministry of Environment and Urbanization 2021)

Station name Latitude Longitude
Meteoroloji 39° 44′ 37″ 37° 00′ 06″
Başöğretmen 39° 44′ 55″ 37° 01′ 32″
İstasyon Kavşağı 39° 44′ 50″ 37° 00′ 47″
Open in a new tab

Fig. 1.

Fig. 1

Open in a new tab

Locations of the air quality monitoring stations on the map (Google Earth Image© 2019)

Meteoroloji AQMS is located in a residential area, İstasyon Kavşağı AQMS is located on the street that sees the most intense traffic and human crowds in the city, and Başöğretmen AQMS is located close to an area known as the small industrial site of Sivas. İstasyon Kavşağı AQMS is approximately 1 km away from both Meteoroloji and Başöğretmen AQMS, while Meteoroloji AQMS and Başöğretmen AQMS are 2.09 km away from each other.

The stations were categorized according to their locations with Meteoroloji AQMS being categorized as urban, İstasyon Kavşağı AQMS as traffic, and Başöğretmen AQMS as (small) industrial and then statistically compared with each other. Additionally, the three monitoring stations are in different directions that represent various dominant urban traffic characteristics associated with high, medium, and low densities.

Methodology

Data were categorized in various datasets using spreadsheets (Excel) and statistical analysis software (SPSS V.22) to examine the changes in PM10 concentration over the years.

In the first stage of the study, the general descriptive statistical parameters (means, confidence intervals for the mean at 95% significance level, standard deviations, min/max values) of the air pollutant (PM10) that make up the large data set were calculated.

Determining how the measured PM10 parameter changes annually may be important in terms of understanding the characteristic features of the region where the air quality is analyzed. For this reason, in the second stage of the study, variance analysis tests were conducted on air pollutants and to examine whether there was a difference between the annual changes in pollutants and the changes according to stations. For the statistical tools, namely SPSS (Statistical Package for the Social Sciences), Wilcoxon signed rank, and Kruskal–Wallis tests, pvalue = 0.05 for probability and its reliability is 95%. pvalue < 0.05 means there is a difference between the mean of the two global variables and the PM10 concentration, and pvalue ≥ 0.05 means there is a similarity between the mean of the two global variables.

In the third stage of the study, bivariate correlation analyses were performed (Spearman’s rank-order correlation) to examine the years of each measured hourly variable and their statistical correlations with each other. In the last stage of the study, the transport of particulate matter and dust to the city center was examined in detail. Because of this, the model on the Barcelona Supercomputing Center software, showing possible dust transports to the center of Sivas on the days with the maximum PM10 value, was used to examine the dust transports to Sivas between 2016 and 2020, and even the dust loads coming from other countries or continents.

Results and discussion

General statistics of PM10 parameter

Table 3 shows the general descriptive statistical parameter means, confidence intervals for the mean at 95% significance level, standard deviations, and min/max values of the air pollutant PM10 that make up the large data set obtained from the three air quality monitoring stations.

Table 3.

General descriptive statistics of PM10 pollutant by stations

Başöğretmen AQMS Meteoroloji AQMS İstasyon Kavşağı AQMS
Mean (µg/m3) 53.55 49.16 66.22
Confidence intervals for the mean at 95% significance level Lower limit (µg/m3) 53.01 48.75 65.67
Upper limit (µg/m3) 54.10 49.56 66.77
Standard deviation (µg/m3) 55.72 41.97 57.17
Minimum (µg/m3) 0.24 0.02 0.00
Maximum (µg/m3) 3554.02 779.14 4766.95
Interval (µg/m3) 3553.78 779.12 4766.95
Number of data 40,202 41,041 41,371
Open in a new tab

Hourly measurement values were taken for each parameter in this study. In order to represent the period covering the last 5 years (2016–2020) with daily and monthly means, 43,848 data required for the measured parameter for each station and a total of 122,614 data obtained from the stations were included in the analysis. Taking into account the 5-year mean value, the mean values of all PM10 concentrations measured were calculated as 53.55, 49.16, and 66.22 µg/m3 at Başöğretmen, Meteoroloji, and İstasyon Kavşağı AQMS, respectively. Estimations show the mean PM10 at a statistical 95% significance level is between 53.01 and 54.10 µg/m3 at Başöğretmen AQMS, 48.75 and 49.56 µg/m3 at Meteoroloji AQMS, and 65.67 and 66.77 µg/m3 at İstasyon Kavşağı AQMS.

Table 4 shows the highest hourly and daily PM10 levels measured at the three stations in µg/m3.

Table 4.

The highest hourly and daily PM10 levels measured at the three stations (µg/m3)

The highest hourly PM10 level
Date of measurement Time of measurement Station PM10max (µg/m3)
16.10.2018 01:00 Başöğretmen AQMS 3554.02
15.10.2018 21:00 Meteoroloji AQMS 779.14
16.10.2018 01:00 İstasyon Kavşağı AQMS 4766.95
The highest daily PM10 level
Date of Measurement Day Station PM10max (µg/m3)
16.10.2018 Tuesday Başöğretmen AQMS 723.08
16.10.2018 Tuesday Meteoroloji AQMS 365.78
16.10.2018 Tuesday İstasyon Kavşağı AQMS 927.52
Open in a new tab

All 5-year maximum daily PM10 values in October 2018 were measured, and according to RAQAM, these values for PM10, which has a 24-h limit value of 60 µg/m3, are approximately 6 to 15 times higher than the legal limit (Official Gazette of the Republic of Turkey 2008). Table 5 shows the maximum daily PM10 values measured at stations over the years and the number of days exceeding the PM10 limit value permitted in RAQAM.

Table 5.

The maximum daily PM10 values (µg/m3) measured at stations and the number of days exceeding the permitted limit value

Year Station PM10max (µg/m3) Date The number of days exceeding the PM10 limit value permitted in RAQAM
2016 Başöğretmen AQMS 316.82 26.11.2016 66
Meteoroloji AQMS 194.23 7.11.2016 38
İstasyon Kavşağı AQMS 198.27 26.11.2016 118
2017 Başöğretmen AQMS 417.96 23.04.2017 122
Meteoroloji AQMS 240.15 21.12.2017 80
İstasyon Kavşağı AQMS 401.16 25.10.2017 158
2018 Başöğretmen AQMS 723.08 16.10.2018 83
Meteoroloji AQMS 365.78 16.10.2018 108
İstasyon Kavşağı AQMS 927.52 16.10.2018 153
2019 Başöğretmen AQMS 170.34 28.01.2019 121
Meteoroloji AQMS 177.06 23.11.2019 111
İstasyon Kavşağı AQMS 260.08 28.01.2019 186
2020 Başöğretmen AQMS 153.05 19.02.2020 89
Meteoroloji AQMS 129.55 28.11.2020 91
İstasyon Kavşağı AQMS 174.99 19.02.2020 156
Open in a new tab

It is important to examine what parameters air pollution in city centers consists of and how those parameters change over the years. Figure 2 consists of the graphs showing the daily PM10 distributions of the three stations in the city center where measurements were taken for the 5-year period (2016–2020).

Fig. 2.

Fig. 2

Open in a new tab

Graphs showing the daily PM10 distributions of the three stations in the city center where measurements were taken for the 5-year period (2016–2020)

The months with the highest mean PM10 concentrations in the 5-year measurement period at the air quality monitoring stations were determined as November 2016 with 126.01 µg/m3 at Başöğretmen AQMS, November 2019 with 96.86 µg/m3 at Meteoroloji AQMS, and November 2016 with 115.74 µg/m3 at İstasyon Kavşağı AQMS. Looking at graph 2, PM10 values vary from day to day. In general, it is seen that PM10 levels in Sivas are lower in hot months than in cold months.

Distribution of PM10 by years and analysis of variance

At this stage of the study, hourly data were converted to daily mean data to examine PM10 changes over the years. Analyses of variance (the Wilcoxon signed-rank test and the Kruskal–Wallis test) were conducted to determine any differences between PM10 levels and 24-h limit values (µg/m3) (Table 1) of PM10 for Turkey and between PM10 values of stations over the years.

Table 6 presents the results of variance analyses and descriptive statistics showing the changes in PM10 over years based on evaluations.

Table 6.

Analysis of variance of PM10 by stations and time

Year Station N Mean Confidence interval for 95% mean Standard deviation Min Max Wilcoxon signed-rank test sig
Lower limit Upper limit
2016 Başöğretmen 8111 58.686 57.485 59.886 55.147 1.31 1645.48 0.000
Meteoroloji 8287 46.950 46.193 47.707 35.144 0.02 756.62 0.000
İstasyon K 8400 71.713 70.723 72.702 46.252 2.87 461.46 0.000
Kruskal–Wallis test pvalue = 0.000
2017 Başöğretmen 8667 64.293 63.005 65.581 61.184 1.81 1658.15 0.000
Meteoroloji 8559 53.177 52.106 54.248 50.539 1.16 612.38 0.000
İstasyon K 8349 78.966 77.701 80.231 58.970 0.74 920.79 0.061*
Kruskal–Wallis test pvalue = 0.000
2018 Başöğretmen 8687 51.483 49.923 53.043 74.167 0.49 3554.02 0.000
Meteoroloji 8293 55.771 54.804 56.737 44.918 4.00 779.14 0.000
İstasyon K 8714 62.747 60.974 64.521 84.439 0.00 4766.95 0.084*
Kruskal–Wallis test pvalue = 0.000
2019 Başöğretmen 8355 46.451 45.711 47.191 34.504 0.82 315.28 0.000
Meteoroloji 7832 47.467 46.593 48.342 39.473 0.03 495.54 0.000
İstasyon K 8598 59.822 58.986 60.658 39.528 1.13 334.43 0.000
Kruskal–Wallis test pvalue = 0.000
2020 Başöğretmen 6382 44.549 43.691 45.407 34.967 0.24 341.47 0.000
Meteoroloji 8070 42.013 41.234 42.792 35.706 0.17 570.38 0.000
İstasyon K 7310 57.027 56.180 57.874 36.932 2.67 286.70 0.014
Kruskal–Wallis test pvalue = 0.000
Open in a new tab

N, number of data. * The difference is not statistically significant at the 95% significance level

Wilcoxon signed-rank test revealed that the comparison between the PM10 pollutant measured at stations in three different locations and the 24-h limit values in Table 1 showed that there were differences between stations in the relevant years (pvalue < 0.05), except for the years 2017 (p = 0.06) and 2018 (p = 0.08) at İstasyon Kavşağı AQMS. The measurement results of the stations for the same years were also compared with the Kruskal–Wallis test. As a result of the statistical test, the station values for all years were determined to be different from one another and the difference between them was statistically significant at 95% significance level (p = 0.00).

In addition to the results of variance analyses and descriptive statistics showing the changes in PM10 over years, the PM10 changes of previous years were also analyzed. Along with analysis of variance, post hoc multiple comparison tests were used to analyze changes (differences) in pollutant levels between 2016 and 2020 and when those pollutants were at their highest and lowest levels over the past 5 years (Table 7).

Table 7.

Post hoc/Kruskal–Wallis test results showing comparisons of PM10 by year

Station Year 2019 2018 2017 2016
Sig.* Sig.* Sig.* Sig.*
Başöğretmen 2020 .000 .000 .000 .000
2019 .000 .000 .000
2018 .000 .000
2017 .000
Meteoroloji 2020 .000 .000 .000 .000
2019 .000 .006 1.000
2018 .000 .000
2017 .023
İstasyon K 2020 .000 .000 .000 .000
2019 .218 .000 .000
2018 .000 .000
2017 .000
Open in a new tab

*Significance values were corrected with the Bonferroni correction (pBonferroni = 0.05/10 = 0.005) for replicate tests

According to the PM10 statistics for the years presented in Tables 6 and 7 and the Kruskal–Wallis analysis selected for the post hoc multiple comparison test, the highest annual mean of PM10 levels was measured in 2017 and the lowest annual mean in 2020 at Başöğretmen AQMS, and Wilcoxon signed-rank test results showed that the annual mean values were not equal to each other (pvalue = 0.000). According to the post hoc/Kruskal–Wallis comparison test, a difference was observed between all years and the difference between them was determined to be statistically significant at 95% significance level (pBonferroni = 0.000).

The highest annual mean of PM10 levels was measured in 2018 and the lowest annual mean in 2020 at Meteoroloji AQMS, and Wilcoxon signed-rank test results showed that the annual mean values were not equal to each other (pvalue = 0.000). According to the post hoc multiple comparison test, a difference was observed between at least 2 years and the difference between them was significant. The results of the post hoc bivariate comparison test that was conducted to understand from which years the difference originated revealed that in terms of PM10 values, there was not a statistically strong difference between 2016 and 2019 (pBonferroni > 0.005) only, and although there were numerical differences between 2016–2017 and 2017–2019, there was no statistical difference (pBonferroni > 0.005).

The highest annual mean of PM10 levels was measured in 2017 and the lowest annual mean in 2020 at İstasyon Kavşağı AQMS, and Wilcoxon signed-rank test results showed that the annual mean values were not equal to each other (pvalue < 0.05). According to the post hoc multiple comparison test, a difference was observed between at least 2 years and the difference between them was significant. The results of the post hoc bivariate comparison test that was conducted to understand from which years the difference originated revealed that in terms of PM10 values, there was not a statistical difference between 2018 and 2019 (pBonferroni > 0.005) only.

The values measured at the stations over the years were not equal to each other (pvalue = 0.000) and the differences between them were determined to be significant. According to the post hoc bivariate comparison test, station measurements showed that there was no difference between Meteoroloji AQMS and Başöğretmen AQMS in 2019 and 2020 (p = 0.556 and p = 1.00, respectively), and that there were differences between PM10 values of the other stations by years (p = 0.00 and p = 0.01) (Table 8).

Table 8.

Post hoc/Kruskal–Wallis test results showing comparisons of PM10 between stations over years

2016 2017 2018 2019 2020
Sig.* Sig.* Sig.* Sig.* Sig.*
Meteoroloji-Başöğretmen .001 .000 .000 .556 1.000
Meteoroloji-İstasyon K .000 .000 .001 .000 .000
Başöğretmen-İstasyon K .000 .000 .000 000 .000
Open in a new tab

*Significance values were corrected with the Bonferroni correction (pBonferroni = 0.05/3 = 0.017) for replicate tests

Correlations of PM10 by years

The PM10 values measured at stations over the years were evaluated with the Kolmogorov–Smirnov test for normality. As a result of the normality tests performed for all measured years and all stations, the data in all series did not comply with the normal distribution (p < 0.05). Spearman rank-order correlation analysis was performed on these series that did not conform to the normal distribution (Özdamar 2013), and Table 9 shows the respective coefficients.

Table 9.

Results of the Spearman’s rank-order correlation

Station Year 2016 2017 2018 2019 2020
Başöğretmen AQMS 2016 Correlation coefficient (r) 1.000 0.213* 0.199* 0.273* 0.242*
pvalue 0.000 0.000 0.000 0.000
2017 Correlation coefficient (r) 0.213* 1.000 0.192* 0.235* 0.218*
pvalue 0.000 0.000 0.000 0.000
2018 Correlation coefficient (r) 0.199* 0.192* 1.000 0.265* 0.170*
pvalue 0.000 0.000 0.000 0.000
2019 Correlation coefficient (r) 0.273* 0.235* 0.265* 1.000 0.258*
pvalue 0.000 0.000 0.000 0.000
2020 Correlation coefficient (r) 0.242* 0.218* 0.170* 0.258* 1.000
pvalue 0.000 0.000 0.000 0.000
Meteoroloji AQMS 2016 Correlation coefficient (r) 1.000 0.148* 0.073* 0.154* 0.195*
pvalue 0.000 0.000 0.000 0.000
2017 Correlation coefficient (r) 0.148* 1.000 0.098* 0.171* 0.130*
pvalue 0.000 0.000 0.000 0.000
2018 Correlation coefficient (r) 0.073* 0.098* 1.000 0.184* 0.102*
pvalue 0.000 0.000 0.000 0.000
2019 Correlation coefficient (r) 0.154* 0.171* 0.183* 1.000 0.174*
pvalue 0.000 0.000 0.000 0.000
2020 Correlation coefficient (r) 0.195* 0.130* 0.102* 0.174* 1.000
pvalue 0.000 0.000 0.000 0.000
İstasyon Kavşağı AQMS 2016 Correlation coefficient (r) 1.000 0.199* 0.249* 0.311* 0.260*
pvalue 0.000 0.000 0.000 0.000
2017 Correlation coefficient (r) 0.199* 1.000 0.223* 0.290* 0.310*
pvalue 0.000 0.000 0.000 0.000
2018 Correlation coefficient (r) 0.249* 0.223* 1.000 0.255* 0.238*
pvalue 0.000 0.000 0.000 0.000
2019 Correlation coefficient (r) 0.311* 0.290* 0.255* 1.000 0.352*
pvalue 0.000 0.000 0.000 0.000
2020 Correlation coefficient (r) 0.260* 0.310* 0.238* 0.352* 1.000
pvalue 0.000 0.000 0.000 0.000
Open in a new tab

*The correlation coefficient at the significance level is pvalue < 0.01 (the difference is statistically significant at 99% significance level)

The correlation coefficient is a ratio and a value between − 1 and + 1. When the correlation coefficient is − 0.3 < r < 0.3, the relationship between factors is considered to be “weak association.” When this value is 0.3 < r < 0.7 or between − 0.7 r − 0.3, the relationship between factors is considered to be “moderate association.” When the correlation coefficient is greater than 0.7 or less than − 0.7, the relationship between factors is considered to be “strong association.” The Spearman’s rank-order correlation results of PM10 over the years among the stations in the city showed that at İstasyon Kavşağı AQMS, there was a moderate association between 2016–2019 (r = 0.311, p = 0.00), 2017–2020 (r = 0.310, p = 0.00), and 2019–2020 (r = 0.352, p = 0.00). However, a weak association was found between PM10 values measured at Başöğretmen AQMS and Meteoroloji AQMS over all years.

PM10 and dust transport modeling

Dust transport event is an important meteorological disaster. Presence of dust transport adversely affects the environment, living and human health, and the effects of climate change. In order to mitigate the effects of all these, analysis and early warning of dust transport are crucial for risk reduction and resilience (Oğuz and Pekin 2019). According to the highest PM10 concentrations measured in these three stations in Sivas in the last 5 years (Table 5), dust transports from the surrounding area are important for monitoring the extraordinary circumstances of the days when PM10 is highest, and they seem to be impactful at times (Mutlu 2019).

The main objective at this stage of the study is to examine and analyze the effects of dust particles in the city with satellite and ground-based observations, in addition to the “Barcelona Supercomputing Center Dust Regional” dust transport prediction model, while keeping in mind the sample dust transport events that have been identified. Dust transport forecasts were created for North Africa, the Middle East, and Europe with the Barcelona Supercomputing Center prediction model and the software developed based on it. Figure 3 shows the dust transport patterns created by the Barcelona Supercomputing Center (WMO 2020) for the days between 2016 and 2020 with the maximum PM10 value.

Fig. 3.

Fig. 3

Open in a new tab

Dust transport patterns for the days between 2016 and 2020 with the maximum PM10 value

To be able to expertly analyze the dust and particulate matter transport, the model developed for Eastern Europe and the Middle East region should be put into effect at least 2 to 3 days before the first day of study and the dust transport should be examined in detail. Accordingly, the dust transport models developed for the city of Sivas were started 2 days in advance of the dates in the study, and dust loads from other countries or continents were examined using the data from the software.

The possible dust transports to the center of Sivas for the days when PM10 was the highest were examined using the software created by Barcelona Supercomputing Center, and on 26 November 2016, the particulate matter came from the provinces to the west and south-west of Sivas; to the cities where the transport to the center of Sivas originated, the particulate matter is estimated to have come from Syria, Lebanon, Israel, and Egypt. Examinations using the dust transport model showed that on 23 April 2017, dust transport originated from cities in the west, south-west, south, and east and came to those cities from the African continent (Table 10).

Table 10.

Possible dust transports to the center of Sivas on the days between 2016 and 2020 with the maximum PM10 value according to the Barcelona Supercomputing Center software

Years Regions where dust transport may come from Countries where dust transport to the center is possible
26 November 2016 From cities to the west and south-west of Sivas Syria, Lebanon, Israel, and Egypt
23 April 2017 From cities to the south-west, south, and east of Sivas From Syria, Iraq, and Iran
16 October 2018 From cities to the south-east and east of Sivas From Syria, Iraq, and Iran
28 January 2019 From cities to the west of Sivas From Libya and Egypt
19 February 2020 From cities to the east of Sivas From Syria and Iraq
Open in a new tab

Results and discussion

This study was conducted to figure out the differences or similarities of hourly PM10 concentrations between 2016 and 2020 at three air quality monitoring stations in Sivas. The hourly PM10 values measured at the stations in the center of the city of Sivas were observed to have increased for all three stations between 2016 and 2017, and decreased in the following years. One of the reasons for the observed decrease in the PM10 values may be the decline in fossil fuel consumption and the increase in the use of natural gas in the city (Çetin and Demirci 2016). Additionally, in 2020, as part of the COVID-19 restrictions, the Turkish government imposed temporary, complete, and partial curfews, industrial production was halted, domestic and international travel was restricted, and use of private vehicles and public transport decreased. The extraordinary circumstances may be noted as another reason for the possible reduction in air pollution and the decrease in PM10 levels in 2020.

An examination of the number of days exceeding the limit value permitted in RAQAM reveals that the PM10 24-h limit value was gradually reduced from 80 µg/m3 in 2016 to 50 µg/m3 in 2019. The data show that PM10 pollution increased between 2016 and 2017, and the number of days that exceeded the permitted limit value for PM10 in 2017 increased due to the reduction of the permitted limit value in the current regulation. However, there was no proportional course between the PM10 values measured in other years and the number of days that exceeded the permitted limit value after the PM10 limit values in the regulation was reduced. The city of Sivas seems to exceed the PM10 limit values specified in the regulation. This could suggest that the PM10 parameter plays a significant role in Sivas in terms of air pollution.

The evaluations made in this section have been made considering the RAQAM limit values for PM10 and SO2. It is clear that this situation aims to protect human health in the most effective way. In a similar study, it was calculated that the limit value given according to RAQAM exceeded 31.9% of the year for PM10 between 2010 and 2019 in Nevşehir (Oğuz 2020). In a similar study, it was reported that PM10 exceeded 47.2% days of the year, according to the limit values recommended by RAQAM between 2015 and 2017 in Siirt province (Alkan 2018). Considering the changing conditions over the years, according to RAQAM, the PM10 values at the three stations in Sivas, where the measurements were obtained, exceeded the limit values on a mean of 23.9% days out of the year in Meteoroloji AQMS, 28.2% in Başöğretmen AQMS, and 43.1% in İstasyon Kavşağı AQMS. The calculations show that according to EU standards, the PM10 values at the three stations in Sivas, where the measurements were obtained, exceeded the 24-h limit value of 50 µg/m3, 37.6% of the year at Meteoroloji AQMS, 40.8% at Başöğretmen AQMS, and 63.9% at İstasyon Kavşağı AQMS. Additionally, a comparison between years showed that 2020 was the lowest year for all three stations in terms of mean PM10 concentrations. Despite this decrease in PM10 parameter in view of the standards in Turkey in 2020, the annual permitted limit value (annual mean 40 µg/m3) appears to have been exceeded and the daily permitted limit value (mean 50 µg/m3 and maximum 35 times a year) appears to have been exceeded 2.5–4.5 times more.

In a study conducted in Balıkesir between 2014 and 2018, no difference was observed between the PM10 levels between 2014 and 2018 (p > 0.05) at the measuring station. Between the other years, PM10 levels were found to be different from each other (p < 0.05) (Mutlu 2019). A study conducted in Los Angeles showed a very good year-round correlation (r2 = 0.80) in the coastal region (LDS and GRD) in West Los Angeles. Also, a relatively weak correlation (r2 = 0.46) was found between the two stations, although the distance between the 2 measurement stations in the semi-rural area (VBR and GRA sites) in Los Angeles was less than 3 km (Pakbin et al. 2010).

As a result of statistical estimations, the values measured at all three stations were found to have differed (pvalue < 0.05) within their respective years. Post hoc/Kruskal–Wallis test showed that the measurements were different between all years at Başöğretmen AQMS (p < 0.05), that there was no difference for 2016–2017 and 2016–2019 at Meteoroloji AQMS (pBonferron > 0.005), and that there was no difference for 2018–2019 at İstasyon Kavşağı AQMS (pBonferroni > 0.005), while the other years showed difference among themselves (pBonferroni < 0.005). Statistical estimations of station measurements within the years have shown that there was no difference between the measurements of Meteoroloji AQMS and Başöğretmen AQMS in 2019 and 2020 (pBonferroni > 0.017), but that there was a difference between the measurements of the stations in other years (pBonferroni < 0,017). After the discovery of differences between the years within the stations, a correlation analysis was performed and a significantly weak association was found, but could no strong associations were revealed. The strongest association (r = 0.352) was estimated to have been between the years 2019 and 2020. For variance statistics estimations, the low level of association revealed by the correlation analysis due to the differences in the values at all stations among years indicates compatibility.

The model on the Barcelona Supercomputing Center software, showing possible dust transports to the center of Sivas on the days with the maximum PM10 value, was used to examine the dust transports to Sivas between 2016 and 2020, and even the dust loads coming from other countries or continents. There seems to be need for improvement of the current dust transport models in order to be able to acquire more information about particulate matter and dust transport in Sivas.

Since Turkey is a country that may be exposed to desert dust due to its location, this study is important in order to determine the effect of transport. In a study, it was aimed to evaluate the desert dust transport that affected Hatay and its surroundings in September 2015. It has been concluded that the deserts in Iraq and Syria at an altitude of 1500 m to Turkey, and the deserts of Saudi Arabia, Jordan, and Syria at an altitude of 3000 m form the source area as a result of long-distance transport (Topuz and Karabulut 2017). Within the scope of the examined model, estimations suggest that dust transport mostly originates from the neighboring cities to Sivas, and that to those neighboring cities, desert dusts are transported especially from neighboring countries of Syria, Iraq, and Iran as well as African countries such as Israel, Egypt, and Libya. With the HYSPLIT model analysis, it was concluded that the desert dusts were transported from the center of the source area to the surrounding areas at a long distance.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Alkan A. An example of cıtıes where aır pollutıon has reached serıous dımensıons: Siirt. J Bitlis Eren Univ Inst Soc Sci. 2018;7(2):641–666. [Google Scholar]
  2. Broecker WS. Abrupt climate change: causal constraints provided by the paleoclimate record. Earth-Sci Rev. 2000;51:137–154. doi: 10.1016/S0012-8252(00)00019-2. [DOI] [Google Scholar]
  3. Calderon-Garciduenas L, Reed W, Maronpot RR, et al. Brain inflammation and Alzheimer’s-like pathology in individuals exposed to severe air pollution. Toxicol Pathol. 2004;32:650–658. doi: 10.1080/01926230490520232. [DOI] [PubMed] [Google Scholar]
  4. Calderón-Garcidueñas L, Mora-Tiscareño A, Ontiveros E, et al. Air pollution, cognitive deficits and brain abnormalities: a pilot study with children and dogs. Brain Cogn. 2008;68:117–127. doi: 10.1016/j.bandc.2008.04.008. [DOI] [PubMed] [Google Scholar]
  5. Calvo AI, Alves C, Castro A, et al. Research on aerosol sources and chemical composition: past, current and emerging issues. Atmospheric Res. 2013;120:1–28. doi: 10.1016/j.atmosres.2012.09.021. [DOI] [Google Scholar]
  6. Çetin M, Demirci OK. The effect on air quality the usage of natural gas in Erzincan. Erzincan Univ J Sci Technol. 2016;9:8–18. doi: 10.18185/eufbed.61841. [DOI] [Google Scholar]
  7. Chen C, Su M, Liu G, Yang Z. Evaluation of economic loss from energy-related environmental pollution: a case study of Beijing. Front Earth Sci. 2013;7:320–330. doi: 10.1007/s11707-013-0360-4. [DOI] [Google Scholar]
  8. Craig L, Brook JR, Chiotti Q, et al. Air pollution and public health: a guidance document for risk managers. J Toxicol Environ Health A. 2008;71:588–698. doi: 10.1080/15287390801997732. [DOI] [PubMed] [Google Scholar]
  9. Dall’Osto M, Beddows DCS, Pey J, et al. Urban aerosol size distributions over the Mediterranean city of Barcelona, NE Spain. Atmospheric Chem Phys. 2012;12:10693–10707. doi: 10.5194/acpd-12-16457-2012. [DOI] [Google Scholar]
  10. Dockery D, Pope A. Epidemiology of acute health effects: summary of time-series studies. Cambridge: Harvard University Press; 1996. [Google Scholar]
  11. Dockery DW, Pope CA, Xu X, et al. An association between air pollution and mortality in six US cities. N Engl J Med. 1993;329:1753–1759. doi: 10.1056/NEJM199312093292401. [DOI] [PubMed] [Google Scholar]
  12. European Environment Agency (2020) Air quality in Europe — 2020 report. Publications Office of the European Union, Luxembourg, p 157. 10.2800/786656
  13. Genc S, Zadeoglulari Z, Fuss SH, Genc K (2012) The adverse effects of air pollution on the nervous system. J Toxicol 2012.10.1155/2012/782462 [DOI] [PMC free article] [PubMed]
  14. Gieré R, Querol X. Solid particulate matter in the atmosphere. Elements. 2010;6:215–222. doi: 10.2113/gselements.6.4.215. [DOI] [Google Scholar]
  15. Gold DR, Samet JM. Air pollution, climate, and heart disease. Circulation. 2013;128:e411–e414. doi: 10.1161/CIRCULATIONAHA.113.003988. [DOI] [PubMed] [Google Scholar]
  16. Husar RB, Prospero JM, Stowe LL. Characterization of tropospheric aerosols over the oceans with the NOAA advanced very high resolution radiometer optical thickness operational product. J Geophys Res Atmospheres. 1997;102:16889–16909. doi: 10.1029/96JD04009. [DOI] [Google Scholar]
  17. Jacob DJ, Winner DA. Effect of climate change on air quality. Atmos Environ. 2009;43:51–63. doi: 10.1016/j.atmosenv.2008.09.051. [DOI] [Google Scholar]
  18. Kaplan GG, Tanyingoh D, Dixon E, et al. Ambient ozone concentrations and the risk of perforated and nonperforated appendicitis: a multicity case-crossover study. Environ Health Perspect. 2013;121:939–943. doi: 10.1289/ehp.1206085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kim BM, Park J-S, Kim S-W, et al. Source apportionment of PM10 mass and particulate carbon in the Kathmandu Valley. Nepal Atmos Environ. 2015;123:190–199. doi: 10.1016/j.atmosenv.2015.10.082. [DOI] [Google Scholar]
  20. Kim J-H, Kim M-K, Ho C-H, et al. Possible link between arctic sea ice and January PM10 concentrations in South Korea. Atmosphere. 2019;10:619. doi: 10.3390/atmos10100619. [DOI] [Google Scholar]
  21. Kulshrestha A, Satsangi PG, Masih J, Taneja A. Metal concentration of PM2. 5 and PM10 particles and seasonal variations in urban and rural environment of Agra. India Sci Total Environ. 2009;407:6196–6204. doi: 10.1016/j.scitotenv.2009.08.050. [DOI] [PubMed] [Google Scholar]
  22. Lale N (2008) Forest remnants in the mountainous area in the north of Sivas, Master Thesis. Istanbul University, Institute of Social Sciences, İstanbul
  23. MohanKumar SM, Campbell A, Block M, Veronesi B. Particulate matter, oxidative stress and neurotoxicity. Neurotoxicology. 2008;29:479–488. doi: 10.1016/j.neuro.2007.12.004. [DOI] [PubMed] [Google Scholar]
  24. Moreno T, Querol X, Alastuey A, et al. Variations in atmospheric PM trace metal content in Spanish towns: illustrating the chemical complexity of the inorganic urban aerosol cocktail. Atmos Environ. 2006;40:6791–6803. doi: 10.1016/j.atmosenv.2006.05.074. [DOI] [Google Scholar]
  25. Müezzinoğlu A. Hava Kirliliğinin ve Kontrolunun Esasları. İzmir: Dokuz Eylul University Publications; 2004. [Google Scholar]
  26. Mukherjee A, Agrawal M. World air particulate matter: sources, distribution and health effects. Environ Chem Lett. 2017;15:283–309. doi: 10.1007/s10311-017-0611-9. [DOI] [Google Scholar]
  27. Mutlu A. Air quality and meteorology: analysis of correlations, trends and episodes. Gumushane Univ J Sci Inst. 2019;9:724–741. doi: 10.17714/gumusfenbil.563848. [DOI] [Google Scholar]
  28. Nriagu JO. A silent epidemic of environmental metal poisoning? Environ Pollut. 1988;50:139–161. doi: 10.1016/0269-7491(88)90189-3. [DOI] [PubMed] [Google Scholar]
  29. Nriagu JO, Pacyna JM. Quantitative assessment of worldwide contamination of air, water and soils by trace metals. Nature. 1988;333:134–139. doi: 10.1038/333134a0. [DOI] [PubMed] [Google Scholar]
  30. Oberdörster G, Elder A, Rinderknecht A. Nanoparticles and the brain: cause for concern? J Nanosci Nanotechnol. 2009;9:4996–5007. doi: 10.1166/jnn.2009.GR02. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Official Gazette of the Republic of Turkey (2008) Air quality assessment and management regulation. T.R. Ministry of Environment and Urbanization.
  32. Oğuz K. Investigation of air quality and the effects of meteorological factors on air pollution in Nevşehir Province Artvin Çoruh University Natural Hazards Application and Research Center. J Nat Hazards Environ. 2020;6(2):391–404. [Google Scholar]
  33. Oğuz K, Pekin MA (2019) Evaluation of dust event in Turkey dated on 23–26 April 2019. International Disaster and Resilience Congress (idRc2019) 419–423. ISBN: 978–605–031–286–7
  34. Özdamar K (2013) Paket programlar ile istatistiksel veri analizi (Cilt 1). Ank Nisan Kitapevi
  35. Pakbin P, Hudda N, Cheung KL, Moore KF, Sioutas C. Spatial and temporal variability of coarse (PM10−2.5) particulate matter concentrations in the Los Angeles area. Aerosol Sci Technol. 2010;44(7):514–525. doi: 10.1080/02786821003749509. [DOI] [Google Scholar]
  36. Pascal M, Falq G, Wagner V, et al. Short-term impacts of particulate matter (PM10, PM10–2.5, PM2. 5) on mortality in nine French cities. Atmos Environ. 2014;95:175–184. doi: 10.1016/j.atmosenv.2014.06.030. [DOI] [Google Scholar]
  37. Peters A, Veronesi B, Calderón-Garcidueñas L, et al. Translocation and potential neurological effects of fine and ultrafine particles a critical update. Part Fibre Toxicol. 2006;3:1–13. doi: 10.1186/1743-8977-3-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Pope CA., III Mortality effects of longer term exposures to fine particulate air pollution: review of recent epidemiological evidence. Inhal Toxicol. 2007;19:33–38. doi: 10.1080/08958370701492961. [DOI] [PubMed] [Google Scholar]
  39. Ramgolam K, Favez O, Cachier H, et al. Size-partitioning of an urban aerosol to identify particle determinants involved in the proinflammatory response induced in airway epithelial cells. Part Fibre Toxicol. 2009;6:1–12. doi: 10.1186/1743-8977-6-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Rashki A, Rautenbach C, Eriksson PG et al (2013) Temporal changes of particulate concentration in the ambient air over the city of Zahedan, Iran. Air Qual Atmos Health 6:123–135
  41. Samet JM, Dominici F, Curriero F, et al. Particulate air pollution and mortality in 20 US cities: 1987–1994. New Engl J Med. 2000;343:1742–1757. doi: 10.1056/NEJM200012143432401. [DOI] [PubMed] [Google Scholar]
  42. Seinfeld JH, Pandis SN (1998) Atmospheric chemistry and physics. In: Air pollution to climate change. John Wiley&Sons, New York, p 1925
  43. Sharma SK, Mandal TK, Srivastava MK, et al. Spatio-temporal variation in chemical characteristics of PM 10 over Indo Gangetic Plain of India. Environ Sci Pollut Res. 2016;23:18809–18822. doi: 10.1007/s11356-016-7025-2. [DOI] [PubMed] [Google Scholar]
  44. Shen Z, Han Y, Cao J, et al. Characteristics of traffic-related emissions: a case study in roadside ambient air over Xi’an, China. Aerosol Air Qual Res. 2010;10:292–300. doi: 10.4209/aaqr.2009.10.0061. [DOI] [Google Scholar]
  45. Styszko K, Szramowiat K, Kistler M, et al. Mercury in atmospheric aerosols: a preliminary case study for the city of Krakow, Poland. Comptes Rendus Chim. 2015;18:1183–1191. doi: 10.1016/j.crci.2015.05.016. [DOI] [Google Scholar]
  46. Szidat S, Jenk TM, Gäggeler HW, et al. Radiocarbon (14C)-deduced biogenic and anthropogenic contributions to organic carbon (OC) of urban aerosols from Zürich, Switzerland. Atmos Environ. 2004;38:4035–4044. doi: 10.1016/j.atmosenv.2004.03.066. [DOI] [Google Scholar]
  47. T.R. Ministry of Environment and Urbanization, (2021) T.R. Ministry of Environment and Urbanization, National Air Quality Monitoring Network. www.havaizleme.gov.tr/. Accessed 15 Aug 2021
  48. Topuz M, Karabulut M. Desert dust transport assessment in Hatay in September 2015. Turk Stud. 2017;12(3):565–580. doi: 10.7827/TurkishStudies. [DOI] [Google Scholar]
  49. Turkish Statistical Institute (TURKSTAT) (2021) Address based population registration system results. In: Address Based Popul. Regist. Syst. Results 2020. data.tuik.gov.tr/Bulten/Index?p=Adrese-Dayali-Nufus-Kayit-Sistemi-Sonuclari-2020–37210. Accessed 12 Oct 2021
  50. Von Schneidemesser E, Monks PS, Allan JD, et al. Chemistry and the linkages between air quality and climate change. Chem Rev. 2015;115:3856–3897. doi: 10.1021/acs.chemrev.5b00089. [DOI] [PubMed] [Google Scholar]
  51. Watson KE. Air pollution and heart disease. Rev Cardiovasc Med. 2006;7:44–44. [PubMed] [Google Scholar]
  52. WMO (2020) URL-1. In: Dust forecasts. https://www.sds-was.aemet.es/forecast-products/dust-forecasts. Accessed 12 Oct 2021
  53. Yun S, Yoo C. The effects of spring and winter blocking on PM10 concentration in Korea. Atmosphere. 2019;10:410. doi: 10.3390/atmos10070410. [DOI] [Google Scholar]
  54. Zeydan Ö. Assessment of particulate matter (PM10) pollution in Turkey in 2019. J Inst Sci Technol. 2021;11:106–118. doi: 10.21597/jist.745539. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Articles from Air Quality, Atmosphere, & Health are provided here courtesy of Nature Publishing Group

RESOURCES