CC BY
23
Интернет-журнал «Науковедение» ISSN 2223-5167 http ://naukovedenie.ru/ Том 7, №4 (2015) http://naukovedenie.ru/index.php?p=vol7-4 URL статьи: http ://naukovedenie.ru/PDF/86TangVN415 .pdf DOI: 10.15862/86angTVN415 (http://dx.doi.org/10.15862/86TangVN415)
Golubnichiy Artem Aleksandrovich
Katanov Khakass State University Russia, Abakan E-mail: artem@golubnichij.ru
Tuksina Elena Andreevna
Katanov Khakass State University Russia, Abakan E-mail: elena.tuksina@mail.ru
Autocorrelation analysis of time series of PM10 concentrations according to the data from the atmospheric monitoring subsystem of Krasnoyarsk Territory
Abstract. The article deals with the main factors influencing the concentration of suspended particles with diameter of 10 |im or less (PM10). It analyses the structure of pollutant time series and its constituents. The trends and cycles of pollution along with the correlation ratio are established by means of the autocorrelation method with regard to the different time shifts from the data sample including the concentration value of PM10 registered for the period of 8 weeks by four stations of the atmospheric monitoring subsystem of Krasnoyarsk Territory. On the basis of the analysis of the source data the authors determined the stations of the monitoring subsystem having similar values in dynamics of PM10 concentrations with the highest degree of the correlation dependence up to 0.6 achieved in the pair "Cheryomushki" - "Berezovka". By analyzing the spatial location of the stations with respect to the pollutant it is proved that this kind of dependence is quite viable. Anomalous values of the autocorrelation registered during the 7th week of measurement are explained by the comparative analysis of time series. The comparison of data on the concentrations of suspended particles in the impact area of the city of Krasnoyarsk with respect to the city of Achinsk considering the time deviations in proportion to nychthemeron allows for the conclusion about the existence of the conditional background of the pollutant explained by the smooth running of the graph and the cyclicity of concentrations within the seasonal cycle explained by a cosinusoidal shape of the autocorrelation graph. The presented results provide evidence of the presence of several constituents in the trend of PM10, as well as in the cyclical components of the pollutant time series.
Keywords: time series; autocorrelation; air pollution; PM10; seasonality of pollution; atmospheric monitoring; urban sources of pollution.
For citation:
Golubnichiy A.A., Tuksina E.A. Autocorrelation analysis of time series of PM10 concentrations according to the data from the atmospheric monitoring subsystem of Krasnoyarsk Territory // On-line journal "Naukovedenie" Vol 7, №4 (2015) http://naukovedenie.ru/PDF/86TangVN415.pdf (open access). DOI: 10.15862/86TangVN415
Introduction
Air pollution, particularly pollution of the urban air basins, is manifestly negative for the health of the population [1]. At the same time the local increase of concentrations on certain days, or even in certain hours can serve as a factor increasing the risk of mortality caused by certain diseases. The presence of correlation dependence between the mortality of population and the concentration of suspended particles with different diameter (PM2.5, PM10) has been proved by many investigators [24]. The forecast of elevated concentrations of certain pollutants using the different methods of analysis is an essential stage preceding the adoption of preventive measures to ensure the human safety in the technosphere. Existing principles and systems of monitoring of ground-level concentrations of pollutants are based on the systematic and periodic air sampling collected to determine the concentrations of the particular components. Consequently, the received data represent a typical time series and, that is why it makes sense to carry out the data analysis using the methods of time series analysis.
The purpose of study is to analyze the structure of the time series of suspended particles with diameter up to 10 |im. within a subsystem air monitoring of the Krasnoyarsk Territory.
Relevance of a study. A problem of atmospheric air monitoring through a system of monitoring posts is examined in the paper. Processing of bodies of data (time series) by methods of an autocorrelation with consideration is influence weather and other factors are insufficiently studied, which makes subjects of research relevant.
Novelty of a study. Data about concentrations of particulate matters on boundaries of a subsystem of atmospheric air monitoring of the Krasnoyarsk Territory are serving as an object of research. This facility was not studied previously from a perspective of an autocorrelation of time series, which points to newness of this topic.
Subject and methods of study
The initial data for the analysis was the information from the automated monitoring stations (AMS) of the atmospheric monitoring subsystem of the regional monitoring system over the environmental condition in the Krasnoyarsk Territory. [5] The data on PM10 the concentration was taken from four automated monitoring stations (AMS) of the subsystem: city of Achinsk (SouthEastern District), city of Krasnoyarsk (micro district Cheryomushki), urban type settlement Berezovka of Berezovsky District, and village Kubekovo of Emelyanovsky District. The data selection for automated monitoring stations is held intervals in 20 min. (maximum single values of concentrations). The analysis comprises the data registered within 8 weeks of monitoring (June - July, 2015).
The data for each week were analyzed and compared with the preceding results. The sampling size for correlation made up 480 values resulted from 504 records registered within a week. Decrease in the sampling size by 24 values as compared with the initial data was carried out for the purpose of the shift in the process of determining the autocorrelation with positive and negative margins up to 4 hours (12 values). To determine the autocorrelation associated with the seasonal changes in PM10 concentrations, we used the automated monitoring station "Achinsk" geographically remote from the other posts. All the selection values were used for the analysis and the time shift was actualized divisible by 72 values (diurnal cycle).
Autocorrelation of maximum single values of PM10 concentrations
Taking into account the specific features of the mass transfer associated with the spatial and climatic characteristics [6, 7], prevailing wind directions and speeds, the correlation dependence between the concentrations of particles at different stations is more likely to be different in time shift. However, the evidence of similar shifts will indicate the similarity of conditions for formation and distribution of pollutants and homogeneity of climate conditions.
For the analysis of time series there have been constructed the graphs of changes in Pearson's linear correlation coefficient for each of the studied stations in relation to three others, showing the change in the correlation coefficient for each of the eight weeks of monitoring, taking into account the time shift in the interval up to 12 values on each side (± 4 hours) (figures 1-4). The following intervals were selected as the criteria for assessing the correlation ratio:
(0 - 0.3] - weak correlation dependence
(0.3 - 0.6] - moderate correlation dependence
(0.6 - 1) - strong correlation dependence [8].
Figure 1. Change of linear correlation coefficient of PM10 concentrations between the stations "Kubekovo" ("Berezovka", "Cheryomushki", "Achinsk") taking into account the time shift
(compiled by the author)
Figure 2. Change of linear correlation coefficient of PM10 concentrations between the stations "Berezovka" ("Kubekovo", "Cheryomushki", "Achinsk") taking into account the time shift
(compiled by the author)
Figure 3. Change of linear correlation coefficient of PM100 concentrations between the stations "Cheryomushki" ("Kubekovo", "Achinsk", "Berezovka") taking into account the time shift
(compiled by the author)
-Kubekovo
-Berezovka
-CheremushM
Figure 4. Change of linear correlation coefficient of PM100 concentrations between the stations «Achinsk" ("Kubekovo", "Berezovka", "Cheryomushki") taking into account the time shift
(compiled by the author)
Anomalous values and false correlations
To start the analysis of time series of PM10 concentrations it is reasonable to exclude the values contradicting the common data, if there are clear reasons for the exclusion. As is seen from all the graphs of changes of the correlation coefficient, the values of the 7th week evidently stand out from the common trend concerning the determination of the coefficient in relation to the station "Cheryomushki". The graph of values of the initial concentrations for the 7th week taken from the station of "Cheryomushki" (Figure 5) shows the evident problems in the work of the automated monitoring station (AMS) for the period of more than 2.5 days, which significantly affects the time series.
Figure 5. Concentrations of PM10 according to the data from the automated monitoring station "Cheryomushki" (July, 13-19, 2015) (compiledby the author)
The existence of such gaps in the data lead to the fact that most of the correlation values for particles correlations over this period taken from the automated monitoring stations in relation to the station "Cheryomushki" gives the correlation value in the range from - 0.2 to 0.3. However, analyzing the data from the stations of the pair "Cheryomushki" - "Berezovka" it occurs that there is a moderate inverse dependence of the concentrations ranging from - 0.38 to - 0.52. The falsity of the correlation can be easily explained with the analysis of time series of PM10 concentrations for the station "Berezovka" (Figure 6).
Figure 6. Concentrations of PM10 according to the data of automated monitoring station "Berezovka" (July, 13-19, 2015) (compiled by the author)
The time series in the interval from July, 13 to July, 16 (up to 12:00) has values averagely 1.75-times exceeding the remainder of the series. This ratio is in good agreement with the data on the wind intensity within the observed period of time (Table 1).
Table 1
Prevailing wind speeds and directions according to the data from the automated monitoring station "Cheryomushki" (13.07.2015 -19.07.2015), %
Wind direction Speed, m/s date
13.07. 2015 14.07. 2015 15.07. 2015 16.07. 2015 17.07. 2015 18.07. 2015 19.07. 2015
N <0,5 2,84
0,52,0
NNE <0,5 4,26
0,52,0
NE <0,5 1,39 1,39
0,52,0
ENE <0,5 2,82 1,41 11,11 9,86 1,41
0,52,0 1,39 1,41
E <0,5 2,82 5,63 6,94 4,17 8,45 1,41
0,52,0 1,39 9,72 7,04 23,94
ESE <0,5 1,41 2,82 12,5 1,39 16,90 32,39
0,52,0 1,41 8,33 16,90
SE <0,5 9,72 37,5 28,17
0,52,0 1,41 1,41 4,17 11,27 4,23
Wind direction Speed, m/s date
13.07. 2015 14.07. 2015 15.07. 2015 16.07. 2015 17.07. 2015 18.07. 2015 19.07. 2015
SSE <0,5 1,41 2,78 5,56 1,41 1,41
0,52,0 4,23 1,39 2,82 2,82
S <0,5 5,64 52,10 7,03 4,17 2,78 4,23 4,23
0,52,0 4,23 1,39 1,41 1,41
SSW <0,5 2,82 12,67 5,56 1,41 5,63
0,52,0 1,41 5,64
SW <0,5 42,25 4,23 30,97 6,94 5,56 5,64 1,41
0,52,0 11,26 1,41 8,44 8,33
WSW <0,5 1,41 9,85 4,23 8,33 2,78
0,52,0 15,49 2,82 2,82 13,89
W <0,5 4,23 1,41 4,23 2,78 1,39
0,52,0 5,64 5,63 11,11
WNW <0,5 7,04 1,41 7,03 1,39 2,82
0,52,0 2,78
NW <0,5 2,82
0,52,0
NNW <0,5
0,52,0
During the period from July, 16 to July, 19 the intensity of the wind with the speed of 0.5-2.0 m/s has increased almost by 40 %, which explains the decrease of PM10 concentrations, at the same time the data on the concentrations of suspended particles with diameter up to 10 ^m within the period from July, 16 to July, 19 (up to 15:00) was selected with errors and presented as a set of values equal to or close to 0. The similarity of the peak concentrations of particles in the remaining period speaks for the similarity of conditions for formation of pollutants and (or) its spread, but these data values are small in comparison with the difference in the first part of the graph. That is the way how the false correlation is explained.
Time shifts and correlation dependence
The most frequent method used for determination of the correlation time series is the use of certain time shift (time correction). Most of the environmental factors, both artificial and natural acting upon the object causes its change after a lapse of time due to the properties of objects (tolerance to environmental factors) and also to the special characteristics of impact.
In the study of codependence of changes in the concentrations of suspended particles at the different stations of the monitoring system it is necessary to take into account the geographical
remoteness, climatic characteristics (wind speed and direction), as well as the other factors of temporary effect exerting a significant impact on pollution [9, 10].
The availability of the data concerning the changes in the concentration of suspended particles collected for the period of 8 weeks provides the opportunity to select similar or acceptable conditions for the determination of the impact produced by any given factors on the autocorrelation of PM10 pollution.
The presented graphs of changes in Pearson's correlation coefficient in dependence to the time shifts (Figures 1-4) makes it possible to distinguish the data sets of the same type.
Spatial autocorrelation features
On the territory of the city of Krasnoyarsk and its direct impact there are three automated monitoring stations (AMS) fixing the content of suspended particles with diameter up to 10 |im. The specific of the location of the stations allows to evaluate the impact of the city as a complex source of pollution. All the stations are located in the relatively ventilated area, or on the outskirts of the city (Figure 7), which conforms to the standards of the monitoring systems [11].
ZaDovednik "Stolbv"
Figure 7. Location of automated monitoring stations (AMS) for the air quality control in the city of
Krasnoyarsk (compiled by the author)
The distance between the posts is relatively small: "Berezovka" - "Cheryomushki" - 9.3 km, "Kubekovo" - "Berezovka" - 11 kilometers, "Kubekovo" - "Cheryomushki" - 17,3 km. For the sake of convenience the correlation dependence of PM10 concentrations is presented in Table 2 in general terms.
Table 2
Maximum values of autocorrelation rmax of PM10 concentrations between the automated
monitoring stations for time shift t
"Kubekovo" - " Cheryomushki"
Week 1 2 3 4 5 6 7 8
t 2 3 6 0 2 9 * 4
Imax 0,586 0,233 0,543 0,517 0,571 0,390 * 0,273
"Kubekovo" - - "Berezov ka"
t 2 8 12 9 3 8 0 2
Imax 0,435 0,437 0,454 0,352 0,491 0,619 0,196 0,257
"Berezovka" - " Cheryomushki"
T 1 1 5 1,2 -4 0 * 3
Imax 0,584 0,530 0,649 0,582 0,446 0,594 * 0,394
* - the data are disregarded due to the presence of gaps in the selection of values described above.
The strongest dependence of the change in PM10 concentrations of is observed in the pair of stations "Berezovka" - "Cheryomushki" (autocorrelation in all the cases under consideration has a value not less than moderate), while the correlating series of the concentrations in 5 out of 7 cases under study have a shift relatively to each other not exceeding one hour upward with the higher indicators of the data from the station "Cheryomushki". This character of the time shift completely agrees with the location of stations with regard to the sources of pollution. Consequently, a significant contribution to the content of PM10 in the air basins of the cities is made by the road transport. The automated monitoring station "Cheryomushki", situated on the territory of the city, records the concentration values which are 2-3 times higher than those of the station "Berezovka". However, the massif of the road infrastructure neighboring the station "Berezovka" is situated at the comparatively smaller distance compared to the massif of the automated monitoring station "Cheryomushki", and, accordingly, taking into account the prevailing wind directions for the city (SW, SSW) this correlation dependence can be fully explained by the location of the objects of the air condition monitoring.
The second pair in order of importance of the concentration ratio of suspended particles up to 10 |im, is the pair of the automated monitoring stations "Kubekovo" - "Cheryomushki". This pair stands out from all the other automated monitoring stations under study for the greatest distance between the stations which is equal to 17.3 km. The value of linear correlation coefficient in 5 out of 7 cases under the study has the level of moderate dependence while in the rest of the cases it is distinguished by the weak dependence. The greatest number of dependencies for this pair is also observed with a slight time shift of the data (5 out of 7 values have a shift no more than an hour) also the maximum correlation values of the concentrations occur in case of the minimal time shifts or their absence. This kind of dependence indicates the presence of a diurnal constituent in the time series.
The lowest correlation dependence of PM10 concentrations is observed in the pair "Kubekovo" - "Berezovka". The correlation values for most of the time series under study by weeks have the lowest indices compared to the other pairs of the automated monitoring stations, at the same time the distance between these stations is slightly different from the distance between the automated monitoring stations "Berezovka" - "Cheryomushki" with the highest correlation dependence. This fact can be attributed to the location of stations with respect to each other and with regard to the source of pollution. In view of the fact that the station "Kubekovo" is located to the north of the station "Berezovka", the pollutants transfer is highly improbable, which is confirmed by a significantly lower correlations of suspended particles concentrations. Still there is one case (the 6th week of measurement) with tome series shift in 2 hours 20 minutes where Pearson's coefficient takes value equal to 0.619, which corresponds to the strong correlation of pollutant. Such dependence is explained by the
predominance of SW and SSW wind directions along with the standard ones observed during the 6th week of measurement. Note that on some days there also were SSE and SE winds (Table 3).
Table 3
Prevailing wind speeds and directions according to the data from the automated monitoring
station "Berezovka" (06.07.2015 - 12.07.2015), %
Wind direction Speed, m/s date
06.07. 07.07. 08.07. 09.07. 10.07. 11.07. 12.07.
2015 2015 2015 2015 2015 2015 2015
N <0,5 1,39
0,5-2,0
NNE <0,5
0,5-2,0
NE <0,5 4,17
0,5-2,0 4,17
ENE <0,5 14,28 12,5 23,94
0,5-2,0 4,29 8,33
E <0,5 11,43 4,17 2,82
0,5-2,0 1,39
ESE <0,5 12,85 4,17 2,82 1,39
0,5-2,0 7,14
SE <0,5 12,05 17,14 7,04
0,5-2,0 15,71
2,0-4,0 4,29
SSE <0,5 14,04 5,56 1,43 18,06 9,86 1,39
0,5-2,0 11,11
2,0-4,0 1,39
S <0,5 8,77 4,17 1,43 8,45 15,28
0,5-2,0 2,78 1,39
SSW <0,5 6,94 4,29 1,39 1,39 12,68 12,05
0,5-2,0 8,77 8,33 2,82
2,0-4,0 7,02 2,78
4,0-6,0 5,26
SW <0,5 2,78 5,72 4,17 11,11 8,45 5,56
0,5-2,0 8,33 5,56 2,82 6,94
2,0-4,0 14,04 30,56 1,39
4,0-6,0 5,26 2,78
WSW <0,5 1,75 8,33 1,39 2,82 20,8
0,5-2,0 15,3 7,04 6,94
2,0-4,0 10,5 8,33
4,0-6,0 3,51
W <0,5 5,26 23,6 22,2 8,45 11,1
0,5-2,0 6,94 12,5
2,0-4,0 15,8
WNW <0,5 1,39 6,94
0,5-2,0 20,8 1,39
NW <0,5 2,78 2,78
0,5-2,0
NNW <0,5 1,39
0,5-2,0
Taking into account the distance between the stations (11 km) and the wind directions with intensity up to 0.5 and from 0.5 to 2 m/s it shall be assumed that the maximum correlation of the
concentrations will be observed in the range of the time shift from 6 to 10 values, which fully corresponds to the previously constructed graph (Figure 2).
Diurnal cycles of PM10 pollution and conditional background
To study the autocorrelation of PM10 concentrations we also took the automated monitoring station located outside the impact zone of the city of Krasnoyarsk. The autocorrelation of concentration values of suspended particles from the station "Achinsk" in relation to the other stations (Figure 4) provides an opportunity to trace certain dependence. Therefore, comparing the concentration data from the automated monitoring station "Achinsk"- "Cheryomushki" with the highest correlation relationship due to the site uniformity (density and spatial features) there can be seen a tendency to a significant decrease in the correlation in case of time shifts by 9 - 12 values. The maximum correlation relationship is observed either without or with a slight shift in time series (up to 4 values), which may be indicative of the uniformity of concentrations depending on the time of day i.e. the existence of some kind of diurnal cycles for concentrations.
The values of correlation coefficients in the individual monitoring's reach the level of the strong dependence. In case of shifts divisible by the value 72 of the index, however, this relationship has the properties to decrease (Figure 8).
0,6 0,5 0,4 0,3 0,2 0,1 0,0 -0,1 -0,2 -0,3
-Kubekovo -Berezovka -Cheremushki
Figure 8. Change of the linear correlation coefficient of PM10 concentrations between the stations "Achinsk" ("Kubekovo", "Berezovka", "Cheryomushki") taking into account the diurnal shift of the
time series (compiled by the author)
The presence of a significant linear correlation in case of the deviation of the time series in 12 values indicates the presence of seasonal factors in the formation and accumulation of suspended particles, and the nature of a smooth decrease of the correlation coefficient points to a certain conditional background of the pollutant, which in its turn is typical of the most pollutants different in
origin [12]. It should be noted that in general terms the graph has a cosinusoidal form. This dependence with reference to the period of the graph indicates the presence of objective laws depending on the day of the week.
Conclusion
The study of time series gave the opportunity to establish the trends and cycles of pollution and to determine the correlation ratio of the values in the dynamics of PM10 concentrations. This complexity of the formation of this pollutant contributes to the presence of different cycles, which partially reinforce and weaken each other.
In the further analysis, these obtained dependences with all climatic characteristics are most likely to serve as a basis for drawing up models of the formation and transfer of suspended particles with diameter up to 10 |im and other pollutants.
REFERENCES
1. 2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
Diane R. Gold and Jonathan M. Samet Air Pollution, Climate, and Heart Disease Circulation. 2013;128: e411-e414, DOI:10.1161/CIRCULATIONAHA.113.003988.
Yi O1, Hong YC, Kim H. Seasonal effect of PM(10) concentrations on mortality and morbidity in Seoul, Korea: a temperature-matched case-crossover analysis. Environ Res. 2010 Jan; 110(1):89-95. DOI: 10.1016/j.envres.2009.09.009.
Pope CA 3rd, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito K, Thurston GD (2002) Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 287(9):1132—1141.
Liuhua Shi, Antonella Zanobetti, Itai Kloog, Brent A. Coull, Petros Koutrakis, Steven J. Melly, and Joel D. Schwartz Low-Concentration PM2.5 and Mortality: Estimating Acute and Chronic Effects in a Population-Based Study Environ Health Perspect DOI: 10.1289/ehp.1409111.
Podsistema monitoringa zagrjaznenija atmosfernogo vozduha [Jelektronnyj resurs]: URL: http://krasecology.ru/Air (data obrashhenija: 01.08.2015).
Golubnichij A.A., Chajkina E.A. Dinamika zagrjaznenija atmosfernogo vozduha goroda Krasnojarska // Sovremennye nauchnye issledovanija i innovacii. 2015. №3 [Jelektronnyj resurs]. URL: http://web.snauka.ru/issues/2015/03/50474 (data obrashhenija: 01.08.2015).
Golubnichij A.A., Chajkina E.A. Analiz sezonnyh kolebanij urovnja zagrjaznenija atmosfernogo vozduha g. Krasnojarska po dannym avtomatizirovannyh postov nabljudenija // Sovremennye nauchnye issledovanija i innovacii. 2015. №5 [Jelektronnyj resurs]. URL: http://web.snauka.ru/issues/2015/05/53107 (data obrashhenija: 01.08.2015).
Ivanter Je.V., Korosov A.V. Jelementarnaja biometrija. - Petrozavodsk: Izd-vo PetrGU, 2010. -104 s.
Prashant Kumar, Paul Fennell, Rex Britter Effect of wind direction and speed on the dispersion of nucleation and accumulation mode particles in an urban street canyon Sci Total Environ. 2008 Aug 25;402(1):82-94. DOI: 10.1016/j.scitotenv.2008.04.032.
Kim J-J, Baik J-J. A numerical study of the effects of ambient wind direction on flow and dispersion in urban street canyons using the RNG k-[epsiv] turbulence model. Atmos Environ 2004; 38(19):3039-3048. DOI: 10.1016/j.atmosenv.2004.02.047.
OND 86, RD 52.04.212-86 Metodika rascheta koncentracij v atmosfernom vozduhe vrednyh veshhestv, soderzhashhihsja v vybrosah predprijatij
Golubnichij A.A., Chajkina E.A. Posezonnye izmenenija srednesutochnyh znachenij koncentracij monooksida ugleroda v g. Krasnojarske za odnoletnij period // Internet-zhurnal «NAUKOVEDENIE» Tom 7, №3 (2015) [Jelektronnyj resurs]: http://naukovedenie.ru/PDF/152TVN315.pdf (data obrashhenija: 01.08.2015). DOI: 10.15862/152TVN315.
УДК 504.3.054:519.246.85
Голубничий Артем Александрович
ФГБОУ ВПО «Хакасский государственный университет им. Н.Ф. Катанова»
Россия, Абакан1
Ассистент кафедры «Инженерной экологии и основ производства»
E-mail: artem@golubnichij.ru РИНЦ: http ://elibrary. ru/ author_items. asp?authori d=683836 ORCID: http://orcid.org/0000-0002-7559-7492
Туксина Елена Андреевна
ФГБОУ ВПО «Хакасский государственный университет им. Н.Ф. Катанова»
Россия, Абакан
Студентка кафедры «Инженерной экологии и основ производства»
E-mail: elena.tuksina@mail.ru
Автокорреляционный анализ временных рядов концентраций PM10 по данным подсистемы мониторинга атмосферного воздуха Красноярского края
Аннотация. В статье рассматриваются основные факторы, влияющие на концентрацию взвешенных частиц диаметром до 10 мкм (РМ10). Анализируется структура временного ряда загрязнителя и его составляющие. Методами автокорреляции с учетом разных временных сдвигов из выборки, включающей значение концентраций РМ10 за 8 недель четырех постов подсистемы мониторинга атмосферного воздуха Красноярского края, устанавливается наличие трендов и циклов загрязнения, а также определяется теснота связи. На основании анализа исходных данных были определены посты подсистемы мониторинга имеющие сходные значения динамики концентраций РМ10, наибольшая степень корреляционной зависимости до 0,6 достигнута в паре «Черемушки» - «Березовка». Посредством анализа пространственного расположения постов относительно загрязнителя доказывается состоятельность такого рода зависимости. Сравнительным методом анализа временного ряда объясняется аномалия значений автокорреляции за 7 неделю измерений. Сравнение данных о концентрациях взвешенных частиц в зоне воздействия г. Красноярска по отношению к г. Ачинску с временными отклонениями пропорционально суткам позволяет сделать вывод как о наличие условного фона загрязнителя, объясняемого плавным ходом графика, так и о цикличности концентраций внутри сезонного цикла, объясняемого косинусоидальной формой графика автокорреляциии. Представленные результаты свидетельствуют о наличие нескольких составляющих в тренде РМ10, и циклических компонентах временного ряда загрязнителя.
Ключевые слова: временные ряды; автокорреляция; загрязнение воздуха; РМ10; сезонность загрязнения; мониторинг атмосферного воздуха; городские источники загрязнения.
1 655017, Республика Хакасия, г. Абакан, ул. Ленина, д. 90
ЛИТЕРАТУРА
1. Diane R. Gold and Jonathan M. Samet Air Pollution, Climate, and Heart Disease Circulation. 2013;128: e411-e414, DOI:10.1161/CIRCULATIONAHA.113.003988.
2. Yi O1, Hong YC, Kim H. Seasonal effect of PM(10) concentrations on mortality and morbidity in Seoul, Korea: a temperature-matched case-crossover analysis. Environ Res. 2010 Jan;110(1):89-95. DOI: 10.1016/j.envres.2009.09.009.
3. Pope CA 3rd, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito K, Thurston GD (2002) Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 287(9):1132—1141.
4. Liuhua Shi, Antonella Zanobetti, Itai Kloog, Brent A. Coull, Petros Koutrakis, Steven J. Melly, and Joel D. Schwartz Low-Concentration PM2.5 and Mortality: Estimating Acute and Chronic Effects in a Population-Based Study Environ Health Perspect DOI: 10.1289/ehp.1409111.
5. Подсистема мониторинга загрязнения атмосферного воздуха [Электронный ресурс]: URL: http://krasecology.ru/Air (дата обращения: 01.08.2015).
6. Голубничий А.А., Чайкина Е.А. Динамика загрязнения атмосферного воздуха города Красноярска // Современные научные исследования и инновации. 2015. №3 [Электронный ресурс]. URL: http://web.snauka.ru/issues/2015/03/50474 (дата обращения: 01.08.2015).
7. Голубничий А.А., Чайкина Е.А. Анализ сезонных колебаний уровня загрязнения атмосферного воздуха г. Красноярска по данным автоматизированных постов наблюдения // Современные научные исследования и инновации. 2015. №5 [Электронный ресурс]. URL: http://web.snauka.ru/issues/2015/05/53107 (дата обращения: 01.08.2015).
8. Ивантер Э.В., Коросов А.В. Элементарная биометрия. - Петрозаводск: Изд-во ПетрГУ, 2010. -104 с.
9. Prashant Kumar, Paul Fennell, Rex Britter Effect of wind direction and speed on the dispersion of nucleation and accumulation mode particles in an urban street canyon Sci Total Environ. 2008 Aug 25;402(1):82-94. DOI: 10.1016/j.scitotenv.2008.04.032.
10. Kim J-J, Baik J-J. A numerical study of the effects of ambient wind direction on flow and dispersion in urban street canyons using the RNG k-[epsiv] turbulence model. Atmos Environ 2004; 38(19):3039-3048. DOI: 10.1016/j.atmosenv.2004.02.047.
11. ОНД 86, РД 52.04.212-86 Методика расчета концентраций в атмосферном воздухе вредных веществ, содержащихся в выбросах предприятий.
12. Голубничий А.А., Чайкина Е.А. Посезонные изменения среднесуточных значений концентраций монооксида углерода в г. Красноярске за однолетний период // Интернет-журнал «НАУКОВЕДЕНИЕ» Том 7, №3 (2015) [Электронный ресурс]: http://naukovedenie.ru/PDF/152TVN315.pdf (дата обращения: 01.08.2015). DOI: 10.15862/152TVN315.
Рецензент: Шанина Елена Владимировна, зав. кафедрой «Инженерной экологии и основ производства», кандидат технических наук, ФГБОУ ВПО «Хакасский государственный университет им. Н.Ф. Катанова».
