=Paper=
{{Paper
|id=Vol-3006/57_short_paper
|storemode=property
|title=Assessment of the environmental situation in Krasnoyarsk using remote sensing data
|pdfUrl=https://ceur-ws.org/Vol-3006/57_short_paper.pdf
|volume=Vol-3006
|authors=Konstantin V. Krasnoshchekov,Oleg E. Yakubailik
}}
==Assessment of the environmental situation in Krasnoyarsk using remote sensing data==
https://ceur-ws.org/Vol-3006/57_short_paper.pdf
Assessment of the environmental situation in
Krasnoyarsk using remote sensing data
Konstantin V. Krasnoshchekov1 , Oleg E. Yakubailik2
1
Federal Research Center Krasnoyarsk Science Center SB RAS, Krasnoyarsk, Russia
2
Institute of Computational Modelling SB RAS, Krasnoyarsk, Russia
Abstract
The data on ground concentrations of aerosols and small gas components (particulate matter PM2.5 and
sulfur dioxide NO2 ) were compared with remote sensing data obtained over the territory of Krasnoyarsk
from June to August 2020. We use the air monitoring system of the Krasnoyarsk Scientific Center of the
Siberian Branch of the Russian Academy of Sciences (KSC SB RAS) to determine the concentration of
PM2.5 . NO2 concentrations were taken according to the data of the State departmental information and
analytical system of the Ministry of Ecology of the region. It is shown that the remote sensing data of the
MODIS MAIAC algorithm with a spatial resolution of 1 km can be used to determine the concentration of
PM2.5 as an addition to the data obtained by the ground-based air monitoring system of the KSC SB RAS.
The MAIAC data were calculated using two different models and are given to the measurement system
used in the KSC SB RAS monitoring network. A high coefficient of determination between satellite and
ground monitoring data was obtained. Determination coefficients were also obtained for NO2 , showing
how applicable the remote sensing data are for assessing the environmental situation in Krasnoyarsk.
Keywords
Environmental monitoring, MODIS MAIAC, PM25 , aerosol, AOD, Krasnoyarsk.
1. Introduction
The issue of air quality assessment is particularly acute in large industrial and developing cities.
The low quality of atmospheric air affects the health of the population and the state of the
environment in general. Air consists of a mixture of different gases. Air quality is determined by a
combination of various physical, chemical, and biological properties. One of the main parameters
for assessing atmospheric air quality is the concentration of small gas components (SGC). The
small gas components of the atmosphere include methane, carbon monoxide, nitrogen dioxide,
etc. The concentration of SGC has a significant impact on the absorption of optical radiation,
and increased concentrations of SGC harm the population’s health. In addition to SGC, aerosols
or particulate matter (PM) are present in the air. Particulate matter of natural or anthropogenic
origin significantly impacts the climate and environment and, like SGC, harms human health [1].
Various epidemiological studies link increased PM concentrations with an increase in the number
of deaths and an increase in respiratory diseases [2]. The results of toxicological studies indicate
SDM-2021: All-Russian conference, August 24–27, 2021, Novosibirsk, Russia
" krasko@icm.krasn.ru (K. V. Krasnoshchekov); oleg@icm.krasn.ru (O. E. Yakubailik)
0000-0002-2668-4776 (O. E. Yakubailik)
© 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
CEUR Workshop Proceedings (CEUR-WS.org)
484
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
the possibility of particulate matter with a size of 10 microns (PM10 ) or less entering the blood
through the lungs, thereby harming health [3].
After twenty years of epidemiological studies, scientists have found a significant correlation
between fine pollutants and respiratory morbidity [4]. It was shown in [5, 6] that an increased
PM concentration in the air could directly lead to an increase in morbidity and mortality of the
population. According to the results of work [7] in the countries of the European Union, the
presence of elevated PM2.5 values in the air (PM with diameters less than 2.5 microns) reduced
the average life expectancy by 8.6 months. After a study of 29 European countries, the authors
of [8] found that respiratory mortality increases by 0.58% for every 10 mg/m3 increase in PM10 .
It was reported in [9, 10] that the prevalence of respiratory diseases increased by 2.07%, and
the frequency of hospitalizations increased by 8% when the daily norm of PM2.5 increased by
10 mg/m3 . Therefore, the assessment of air quality, especially from the point of view of PM10
and PM2.5 , is an urgent problem at the moment.
Krasnoyarsk is an actively developing city. Currently, in the city, monitoring stations using
the optical registration method and the weight method are mainly used to determine PM
concentration. Ground-based observations from monitoring stations show important spatial
and temporal information about the concentration of PM in the atmosphere. Although the
network of monitoring stations in the city has dozens of observation posts, point measurements
do not provide information about PM’s spatial characteristics and distribution of PM in urban
areas of interest, nor do they show the influence of the city’s atmosphere on the surrounding
areas.
The influence of external factors contributing to pollution in the city (the private sector,
industrial facilities located outside the city) is also of interest. The time coverage of on-site
PM measurements also varies greatly depending on the device’s operation period and its
functionality. These reasons have led to ongoing efforts to assess PM using satellite remote
sensing techniques.
Nitrogen dioxide NO2 , like PM, harms the body. The main sources of NO2 are car exhaust
gases, CHP emissions, solid waste incineration, gas combustion. If a small concentration of
nitrogen dioxide enters the respiratory organs, a person experiences respiratory disorders,
coughing, an increase in concentration can lead to oxygen starvation and other negative
consequences. Nitrogen dioxide also negatively affects the environment, increasing the soil’s
acidity, adversely affecting vegetation, and plays an important role in the formation of urban
smog.
2. Data and measurements
2.1. Study area
The city of Krasnoyarsk is the regional center of the Krasnoyarsk Territory, with a population
of more than 1 million people. It is actively developing and has an area of about 350 km2 . The
coordinates of the city center are 56°00’ north latitude and 92∘ 52’ east longitude. The city and the
surrounding territories have a unique relief. From the south and west of the city, there are forests
and hilly terrain. From the north and east, the terrain is mostly flat. The Yenisei River, which
does not freeze in winter due to the nearby Krasnoyarsk hydroelectric power station, divides
485
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
the city approximately in half. Like all major cities, Krasnoyarsk is subject to a negative impact
on the environmental situation. Motor transport, the private sector, thermal power plants, large
enterprises of the metallurgical, machine-building, and chemical industries contribute many
emissions into the atmosphere and negatively affect the quality of the surrounding air. The
concentration of PM2.5 in Krasnoyarsk is 64% higher than the average in Russia [11].
2.2. Satellite data
The aerosol optical depth (AOD) parameter is usually used [12]. AOD is an integrated at-
mospheric scattering of radiation by aerosols in a vertical column of the atmosphere. This
parameter is proportional to the number of particles in the air and depends on their mass
concentration.
In our work, we used the data of the MAIAC product. This algorithm was developed for
processing MODIS data. MAIAC extracts aerosol parameters above the ground with a resolution
of 1 km [13]. The MCD19A2 product (MAIAC) contains data from the MODIS spectrophotome-
ter installed on the Terra and Aqua satellites. This product was published on May 30, 2018,
containing AOD data from February 1, 2000. Aerosol parameters include the optical depth at a
wavelength from 0.47 to 0.67 microns and the type of aerosol, including models of background,
smoke, and dust.
The increased accuracy of MAIAC results from using the explicit surface characterization
method instead of the empirical approach to surface parameterization, which is used in the
MOD04 and MYD04 algorithms. In addition, MAIAC includes a cloud mask algorithm based on
spatial-temporal analysis, which complements traditional methods of detecting clouds at the
pixel level [13]. MAIAC provides a uniform grid resolution of 1 km in the selected projection,
regardless of the scanning angle.
To determine how much AOD correlates with ground-based PM measurements, numerous
linear, chemical [14], transport [15], and neural [16] models have been developed. However, the
particle size distribution, composition, air humidity, and wind speed significantly reduce the
AOD-PM correlation [17]. The correlation between the measurements of the total AOD column
and the near-surface PM25 and these variables was investigated [18]. These studies have shown
a wide range of correlations between AOD and PM25 mass.
The method of obtaining data is based on the reflection of various wavelengths from the
Earth’s surface and the registration of reflected radiation on the device’s sensor. Because of this,
incorrect values are obtained on a highly reflective surface (snow, water, cloud cover). Therefore,
data on the territory of Krasnoyarsk becomes available after the snow cover disappears and
before it appears.
We used the OMNO2d product as satellite data on nitrogen dioxide. This product provides data
on NO2 in a vertical column of the atmosphere with a spatial resolution of 0.25×0.25 degrees. The
OMNO2d product contains daily NO2 values for the entire territory of the Earth from October
1, 2004, to the present [19]. This product is available throughout the year over Krasnoyarsk,
except if the cloud cover exceeds 30%. This product has a rough spatial resolution, which is not
suitable for assessing the environmental situation in the city at the district level. However, this
spatial resolution will allow us to assess the impact of the city on the nearest territories and the
contribution of external factors to the concentration of NO2 in the city.
486
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
2.3. Ground area
The data on PM25 of the air monitoring system of the Krasnoyarsk Scientific Center, consisting
of 26 monitoring posts, were used in work. A certified CityAir air monitoring station developed
by a group of companies from the Novosibirsk Technopark and the Skolkovo Innovation Center
is installed at each monitoring post. These stations provide information about the state of the
surrounding air, its temperature, pressure, and humidity. Additionally, an optical sensor is
installed inside the station to measure the concentrations of PM25 and PM10 .
Data on NO2 were obtained from the State regional environmental departmental information
and analytical data system of the Krasnoyarsk Territory. There are 7 automated observation posts
(AOP) installed on the territory of Krasnoyarsk, providing information on the concentrations of
carbon monoxide, sulfur dioxide, nitrogen dioxide, hydrogen sulfide, and other pollutants. For
our study, we used data from two AOPs since data from other AOP had gaps in the study period.
In Figure 1, a red circle is circled the monitoring stations used to determine the concentrations
of PM25 and NO2 .
2.4. Meteorological observations
To assess the compliance of the satellite sensing data with the ground monitoring data, the
satellite sensing data were adjusted according to meteorological parameters. We used data on
temperature, humidity, and atmospheric air pressure. All these parameters were obtained using
a network of monitoring posts of the KSC SB RAS. In addition to the parameters given above,
the planetary boundary layer height (PBLH) obtained from the atmospheric model of the Global
Forecasting System (GFS) was used.
Figure 1: Location of automated observation posts in Krasnoyarsk.
487
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
2.5. PM2.5 estimation based on satellite data
To assess the extent to which satellite monitoring data can be used to analyze the environmental
situation in the city according to the PM25 parameter, we calculated the correlation coefficient
between the PM25 data obtained from ground-based observations and the data on the AOD
parameter obtained from satellite monitoring data. Since the AOD data is available only in the
absence of snow cover, the period from June to August 2019 and 2020 was taken, a total of 184
days. However, the AOD data is also limited by the presence of cloud cover. As a result, the
final number of days used in the calculations was 80. For comparison, the AOD data located
above the ground monitoring station was taken, then they were averaged, getting an average
value over the city. The ground monitoring data was taken as the PM25 concentration value
measured for 12 hours of the day. The data between the posts was averaged, obtaining the
average PM2.5 value in the city.
To compare the AOD and PM25 data, the AOD data were recalculated to PM25 units of
measurement (mg/m3 ). We used two models.
The first model has a linear form, as shown in formula (1)
PM𝑐𝑎𝑙𝑐 = 𝑎 · AOD + 𝐵, (1)
where PM𝑐𝑎𝑙𝑐 is the calculated PM values in mg/m3 , 𝑎 and 𝐵 are linear regression coefficients
and are equal to 7.2 · 10−3 and 5.3 · 10−3 , respectively. The result obtained using this formula
is shown in Figure 2.
The next model considered has the form shown in formula (2). It includes meteorological
parameters and takes into account aerosol characteristics. This formula is widely used for
comparing AOD and PM data:
⧸︃ (︃ ]︂ )︃
1 − RH −𝛾
[︂
AOD
PM𝑐𝑎𝑙𝑐 = 𝐾· (2)
PBLH 1 − RH0
To obtain the calculated values according to the formula (2), it is necessary to take into
account, in addition to the AOD values, such parameters as PBLH — the height of the boundary
layer of the atmosphere, RH — air humidity, RH0 — the average value of air humidity for the
selected territory, the variables 𝛾 and 𝐾 are aerosol characteristics and are obtained from [11].
The results obtained by formula (2) are shown in Figure 3.
Figure 2: Comparison of ground data (y-axis) and calculated PM25 values obtained from the first model
(x-axis).
488
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
Figure 3: Comparison of ground data (y-axis) and calculated PM25 values obtained from the second
model (x-axis).
2.6. NO2 estimation based on satellite data
To assess the extent to which satellite monitoring data can be used to assess the environmental
situation in the city according to the NO2 parameter, we calculated the correlation coefficient
between NO2 data obtained from ground-based observations and NO2 data obtained from
satellite monitoring data. Since the data has a spatial resolution of 0.25 degrees, the territory of
Krasnoyarsk is covered by two pixels, which were averaged. The time period from June 1 to
August 31, 2020, was chosen to compare satellite and ground data. NO2 satellite monitoring
data is provided in moles/cm3 units in the vertical column of the atmosphere, data from ground
monitoring posts are provided in mg/m3 units. To calculate the correlation coefficient between
Figure 4: Comparison of NO2 ground data (y-axis) and NO2 values calculated using formula (3) (x-axis).
489
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
the two data series, we converted the satellite monitoring data series using formula (3).
NO2 calc = 𝑎 · NO2 sat + 𝑏 · 𝑅 + 𝑐 · 𝑇 + 𝑑 · 𝑃 + 𝑒 · PBLH (3)
where NO2 sat is the NO2 data obtained using the OMNO2d product, 𝑅 is the air humidity, 𝑇
is the air temperature, 𝑃 is the air pressure, PBLH is the height of the atmospheric boundary
layer. The results obtained using formula (3) are shown in Figure 4.
3. Results and discussion
Satellite AOD measurements, ground-based PM25 data, and meteorological parameters were
used to compare the PM25 data. Applying formula (1) for the satellite data series and comparing
the calculated data series with ground measurements, we obtain the graph shown in Figure 2.
Satellite monitoring data were converted to the units of ground monitoring stations data
and then compared with them. The coefficient of determination between these data sets was
calculated (𝑅2 = 0.71). Such a high value of the coefficient of determination suggests that the
data for PM, calculated using satellite monitoring, have high compliance with the data measured
by ground monitoring stations.
The next stage of the work is to obtain the calculated values of PM𝑐𝑎𝑙𝑐 using the formula (2).
The result is shown in Figure 3.
When adding meteorological parameters and aerosol characteristics of PM25 to the model, it
was possible to increase the correlation between the two data series significantly. Therefore,
satellite monitoring data can be used in addition to data from ground observation posts to
improve the assessment of the environmental situation in Krasnoyarsk. However, the data from
the first model can also be used to assess air quality since they have a strong correlation with
ground data. Data from the linear model can be used in places where weather parameters are
unknown.
To compare the NO2 data, we also used data obtained using satellite measurements and
ground measurements at monitoring posts. Weather parameters were also used to increase the
correlation between these two data series.
Applying formula (3) to a series of data obtained from satellite measurements and comparing
the calculated data series with ground monitoring data, we obtain the graph shown in Figure 4.
The value of the determination coefficient for NO2 is worse than the value for PM25 . Most
likely, this is due to the coarser spatial resolution of satellite data. However, this accuracy is
sufficient for assessing the impact of the urban environmental situation on the surrounding
suburban territory; this accuracy is sufficient.
4. Conclusions
The joint processing of data from ground-based monitoring networks with remote sensing
data will improve the assessment of Krasnoyarsk’s environmental situation. It will be possible
to obtain data on the spatial distribution of pollution in the city, which will strengthen the
understanding of the influence of the city’s atmosphere on the surrounding areas and the
influence of the surrounding areas on the environmental situation in the city.
490
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
Using satellite monitoring data, it is possible to significantly supplement the data obtained
from ground monitoring posts, strengthening the understanding of atmospheric processes
occurring in the city and beyond.
References
[1] Y. Kaufman, D. Tanré, O. Boucher, A satellite view of aerosols in the climate system,
Nature (2002) 215–223. doi:10.1038/nature01091.
[2] J. Schwartz, F. Laden, A. Zanobetti, The concentration-response relation between
PM(2.5) and daily deaths, Environmental Health Perspectives (2002). doi:10.1289/ehp.
021101025.
[3] A. Seaton, D. Godden, W. MacNee, K. Donaldson, Particulate air pollution and acute health
effects, The Lancet 345 (1995) 176–178. doi:10.1016/S0140-6736(95)90173-6.
[4] B. Brunekreef, S. Holgate, Air pollution and health, Lancet 360 (2002) 1233–1242. doi:10.
1016/S0140-6736(02)11274-8.
[5] B. Nemery, P. Hoet, A. Nemmar, The Meuse Valley fog of 1930: An air pollution disaster,
Lancet 357 (2001) 704–708. doi:10.1016/S0140-6736(00)04135-0.
[6] W. Helfand, J. Lazarus, P. Theerman, Donora, Pennsylvania: An environmental disaster of
the 20th century, American Journal of Public Health 91 (2001) 553. doi:10.2105/AJPH.
91.4.553.
[7] H. Orru, M. Maasikmets, T. Lai, T. Tamm, M. Kaasik, V. Kimmel, K. Orru, E. Merisalu,
B. Forsberg, Health impacts of particulate matter in five major Estonian towns: Main
sources of exposure and local differences, Air Quality, Atmosphere and Health 4 (2011)
247–258. doi:10.1007/s11869-010-0075-6.
[8] A. Analitis, K. Katsouyanni, K. Dimakopoulou, E. Samoli, A. Nikoloulopoulos, Y. Petasakis,
G. Touloumi, J. Schwartz, H. Anderson, K. Cambra, F. Forastiere, D. Zmirou, J. Vonk,
L. Clancy, B. Kriz, J. Bobvos, J. Pekkanen, Short-term effects of ambient particles on
cardiovascular and respiratory mortality, Epidemiology 17 (2006) 230–233. doi:10.1097/
01.ede.0000199439.57655.6b.
[9] A. Zanobetti, M. Franklin, P. Koutrakis, J. Schwartz, Fine particulate air pollution and its
components in association with cause-specific emergency admissions, Environmental
Health: A Global Access Science Source 8 (2009). doi:10.1186/1476-069X-8-58.
[10] F. Dominici, R. Peng, M. Bell, L. Pham, A. McDermott, S. Zeger, J. Samet, Fine particulate air
pollution and hospital admission for cardiovascular and respiratory diseases, Journal of the
American Medical Association 295 (2006) 1127–1134. doi:10.1001/jama.295.10.1127.
[11] C. Lin, L. Labzovskii, H. Mak, J. Fung, A. Lau, S. Kenea, M. Bilal, J. Vande Hey, X. Lu,
J. Ma, Observation of PM2.5 using a combination of satellite remote sensing and low-cost
sensor network in Siberian urban areas with limited reference monitoring, Atmospheric
Environment 227 (2020). doi:10.1016/j.atmosenv.2020.117410.
[12] K. Donaldson, X. Li, W. MacNee, Ultrafine (nanometre) particle mediated lung injury,
Journal of Aerosol Science 29 (1998) 553–560. doi:10.1016/S0021-8502(97)00464-3.
[13] A. Lyapustin, Y. Wang, S. Korkin, D. Huang, MODIS Collection 6 MAIAC algo-
rithm, Atmospheric Measurement Techniques 11 (2018) 5741–5765. doi:10.5194/
491
Konstantin V. Krasnoshchekov et al. CEUR Workshop Proceedings 484–492
amt-11-5741-2018.
[14] X. Yap, M. Hashim, A robust calibration approach for PM10 prediction from MODIS
aerosol optical depth, Atmospheric Chemistry and Physics 13 (2013) 3517–3526. doi:10.
5194/acp-13-3517-2013.
[15] I. Kloog, F. Nordio, B. Coull, J. Schwartz, Incorporating local land use regression and
satellite aerosol optical depth in a hybrid model of spatiotemporal PM2.5 exposures in
the mid-atlantic states, Environmental Science and Technology 46 (2012) 11913–11921.
doi:10.1021/es302673e.
[16] P. Gupta, S. Christopher, Particulate matter air quality assessment using integrated
surface, satellite, and meteorological products: Multiple regression approach, Journal of
Geophysical Research Atmospheres 114 (2009). doi:10.1029/2008JD011496.
[17] J. Wang, S. Christopher, Intercomparison between satellite-derived aerosol optical thick-
ness and PM2.5 mass: Implications for air quality studies, Geophysical Research Letters
30 (2003) ASC 4–1 – ASC 4–4. doi:10.1029/2003GL018174.
[18] R. Hoff, S. Christopher, Remote sensing of particulate pollution from space: Have we
reached the promised land?, Journal of the Air and Waste Management Association 59
(2009) 645–675. doi:10.3155/1047-3289.59.6.645.
[19] OMNO2d File Specification, 2013. URL: https://is.gd/Tc0Xoz.
492