Evaluation of NO2 column variations over the atmosphere of Kazakhstan using satellite data

Abstract. Tropospheric NO2 concentrations obtained from the measurements of the Ozone Monitoring Instrument on board the NASA Aura satellite from 2005 to 2016 were studied to identify major NO2 emission hot spots, trends, and seasonal variations over Kazakhstan. Emission hot spots are observed over the locations of thermal power plants (Ekibastuz) and major urban and industrial regions (Almaty and Shymkent), as well as the capital city (Astana). Some decreasing trends have been observed for NO2 over Ekibastuz, whereas the regions of Almaty and Shymkent showed increasing trends due to industrial growth. The seasonal pattern of the NO2 concentration shows a difference between three industrial cities of Almaty, Shymkent, and Ekibastuz versus the rest of Kazakhstan. In these three cities, a NO2 maximum is found during wintertime, which we attribute to seasonality of emissions associated with electricity production and the longer chemical lifetime of NO2 in winter. In contrast, in Astana and the rest of Kazakhstan, the NO2 concentration reaches a maximum in the summer.


Introduction
Nitrogen dioxide (NO 2 ) and its precursor nitric acid (NO) together are known as nitrogen oxides (NO x ), which play a vital role in the tropospheric chemistry.NO x is a precursor of tropospheric ozone and nitrate particulate matter, both of which affect air quality and climate. 1 NO x concentrations in the troposphere are highly variable in both space and time, reflecting its short chemical lifetime in the atmosphere and its nonuniform distribution. 2The lifetime of NO 2 was reported to be ∼6 h in summer, ∼6 to 12 h in spring and fall, and 12 20 h in winter, due to the variability in the photolysis rate and water vapor content in the atmosphere. 3missions of NO x come from both anthropogenic and natural sources.Elevated temperatures associated with combustion combine ambient N 2 and O 2 to form NO, which oxidizes to NO 2 .Major contributors to global NO x emissions include industrial activities, electricity generation, transportation, and prescribed burnings. 4Natural phenomena, such as lightning, wildfires, and soil microbial activities, contribute to NO x emissions. 5his study focuses on NO x emissions in Kazakhstan, which is the ninth largest country in the world at 2.72 million km 2 , one of the top coal and oil producing countries in the world, 6 while maintaining one of the lowest population densities of any country at 7 per km 2 . 7Over 50% of the Kazakhstan's energy is supplied by coal, and the oil sector is the largest determinant of the country's economic growth (ibid).Between 2002 and 2016, the gross domestic product of Kazakhstan has increased sixfold.In 2014, the highest per capita household coal consumption occurred in Poland (165 kgoe/cap), followed by Kazakhstan (157 kgoe/cap), and Mongolia (104 kgoe/cap). 8,9Kazakhstan experiences unique metrological conditions where the temperature in winter time may go down to −40°C in the north.Characterizing NO x emissions in this large, energy-intense country bears relevance to air quality and climate impacts for the region, country, and world.The World Bank found that the air quality monitoring network in Kazakhstan may experience unrepresentative sampling because the number of monitoring sites are low, and the selection of monitored pollutants does not comply with recognized priority pollutants. 10To address this data need, we apply satellite retrievals of NO 2 to assess NO x trends and spatial variability in Kazakhstan.
Long-term observations of satellite NO 2 columns can constrain NO x emission inventories and significantly reduce model uncertainties in simulations studies. 113][14][15] This decrease resulted from a combination of the implementation of emission control devices on thermal power plants, the shuttering of inefficient plants, and stricter vehicle emission standards. 16iu et al. 17 found that a high NO x concentration does not necessarily correspond to a long lifetime, and the correlation between NO x lifetime and NO 2 tropospheric vertical column density is rather low (R 2 ¼ 0.22), probably due to the complex NO x chemistry, which is also affected by meteorological and chemical variability, such as variations in UV flux, water vapor, and VOC levels. 18urrently, only a few studies regarding the atmosphere of Central Asia are available.In particular, no studies were found in the literature that investigates NO 2 concentrations variability in Kazakhstan.The objective of this study was to examine the temporal evolution of tropospheric NO 2 in four major cities of Kazakhstan including Astana, Shymkent, Almaty, and Ekibastuz from 2005 to 2016 using the data from ozone monitoring instrument (OMI) launched on the NASA Aura satellite in July 2004.

OMI Tropospheric NO Columns
Our study uses data from OMI to assess air quality patterns and trends in Kazakhstan.OMI is a nadir-viewing wide-field imaging spectrometer data on the board of NASA Aura satellite.Its ground resolution is 13 km × 25 km, spectral and spatial detection is 780 × 576 pixels, respectively, and can detect UV with wavelength 350 to 500 nm.The instrument observes Earth's backscattered radiation in the ultraviolet and a visible range of 350 to 500 nm with a wide-telescope feeding two imaging grating spectrometers.Aura was launched on July 14, 2004, into a Sun-synchronous polar orbit at ∼705-km altitude with a local equator crossing time of 13:45 in the ascending node. 3,19he hyperspectral capabilities of the instrument can improve the accuracy and precision of the total ozone concentration.Also, it allows for accurate radiometric and wavelength selfcalibration over the long term.The Dutch OMI NO 2 (DOMINO) retrieval algorithm consists of three steps.First, the slant columns of NO 2 are obtained from the OMI reflectance spectra in the 405-to 465-nm window using the DOAS technique. 20Second, the stratospheric and tropospheric contributions to the OMI NO 2 slant columns are separated by assimilating the slant columns of a global chemistry and transport model. 21,22We used monthly average level-3 product.Finally, tropospheric slant column is converted to a vertical column with the tropospheric air mass factor.
The tropospheric NO 2 column retrievals used are from the version-2 DOMINO data product for OMI 22 obtained through the TEMIS25 website.The ground pixel size of the OMI retrievals is 13 to 24 km with a daily global coverage.This study employs the monthly average level 3 OMI NO 2 product available at 0.25 deg × 0.25 deg because these data are gridded files on a fixed grid.One grid cell (0.25 deg × 0.25 deg) falls in each city.Since cities are not perfectly square shaped, at the edges there might be some uncertainties.In the case of Almaty, which is surrounded by mountains, the grid cell did not cover mountains.

Study Area
We focus on the four major cities of Kazakhstan including Astana, Shymkent, Almaty, and Ekibastuz from 2005 to 2016.These four cities are among most populated and largest urban areas where potential NO 2 emissions exist.
Almaty (43.22°N, 76.85°E) is located in the valley of big and small Almatinka rivers at the foot of Trans-Ili Alatau mountains (Fig. 1).The city is surrounded by mountains and hills, limiting horizontal movement of air.Wind generally blows southward during the day, bringing cold and polluted air from the North and northward at night. 23The population of Almaty has increased from 2005 to 2016 by 42% (Fig. 2). 24Sources of NO 2 in Almaty include thermal power plants, motor vehicles, boilers from military garrisons of regional operational units of the Ministry of Defense, boiler companies, and building material manufacturing.Emissions from transportation account for ∼80% of NO x emissions in Almaty. 23As of 2010, over 500,000 vehicles were registered in Almaty, of which 54.6% were produced from 1996 to 2003, 20.9% from 2003 to 2008, and only 2.4% recently manufactured, suggesting that engines do not comply with modern emission standards. 23hymkent (42.34°N, 69.69°E) is an industrial city with the population of 902,000 by 2016.The city population has increased from 2005 to 2016 by 68% (Fig. 3). 25The main sources of NO 2 come from stationary and mobile sources (Fig. 4).The stationary sources include lead, petrochemical, chemical, thermal power plants, and uranium mining enterprises.Mobile sources have been reported to be the main source of formaldehyde, nitrogen dioxide, and carbon monoxide. 26The estimated total emissions of harmful substances in the atmosphere in the city of Shymkent show that the ratio of hazardous substances emitted by stationary sources to harmful substances released by mobile sources average 1 to 4.5 annually. 27kibastuz (51.73°N, 75.32°E) is located in the Pavlodar region of the Republic of Kazakhstan (Fig. 5).An increase in population was observed for Ekibastuz from 2006 to 2016 by 8.6% (Fig. 6).One of the largest coal deposits in Kazakhstan is in Ekibastuz, where high-ash coal is extracted by an open-pit method, and fuels two of the largest power plants in the country.    2 The population of Almaty has increased from 2005 to 2016 by 42%. 24hile 72% of electricity in Kazakhstan is produced by thermal power plants, Ekibastuz State District Power Plant-1 (SDPP-1) contributes to 12% of the demand. 28,29High-ash coals with inadequate ash collectors at local plants lead to significant emissions of harmful substances, such as NO x and SO 2 into the atmosphere.Since Ekibastuz uses brown coal with a high content of minerals (more than 30%), pollutant emissions from coal thermal power plants extend to  Fig. 7 The population of Astana increased by 65% from 2005 to 2016. 24he entire northeast of Kazakhstan, Siberia, and Mongolia.30 Minor industrial activities account for ∼20% of the emissions.31 SO 2 and NO x were detected at a distance of 119 km.32 In December 1997, Kazakhstan's capital city changed from Almaty to Astana (51.16°N, 71.47°E). Asa result industrial outputs and the gross regional products have increased by 11 and 90 times, respectively.While the trend of migration to Almaty did not decrease from 2005 to 2016, the population migrating to the new capital, Astana, increased.Figure 7 shows the population increase by 65% from 2005 to 2016 in Astana. 24This shift in capital, construction, and transport by roads and railways increased dramatically, introducing increased sources of dust and air pollutions.To offset air pollution, the government organized projects such as the construction of a large (190 ha) forest park.In addition, the area of landscaping in Astana increased from 67.9 ha in 1997 to 1061.5 ha in 2009.33 Figure 8 shows the current map of Astana.

Results and Discussion
Annual averages of OMI columns over Kazakhstan from 2005 to 2016 show the high NO 2 vertical tropospheric columns over the megacities (Almaty, Shymkent), as well as the towns with big thermal power plant (Ekibastuz) (Fig. 9).The annually averaged NO 2 vertical columns over Kazakhstan are low, ∼0.8 × 10 15 molecules∕cm 2 .However, there exists "hotspots" in the country, such as the southern cities of Almaty, Shymkent, and the northern city Ekibastuz (Fig. 9).
The city of Almaty showed the most striking changes in NO 2 density from 2005 to 2016 (k ¼ 0.0236).As shown in Fig. 10, the vertical columns peak annual average increased from 4.8 × 10 15 molecules∕cm 2 to 6.4 × 10 15 molecules∕cm 2 .Another striking increase (k ¼ 0.0151) came from another major city in the south, Shymkent (Fig. 10), where the vertical columns increased from 1.4 to 3.1 × 10 15 molecules∕cm 2 .
Overall increase in NO 2 concentrations between 2005 and 2016 over Almaty and Shymkent cities could be attributed to the rise in the population (Figs. 2 and 3, respectively) and vehicle count.Another reason for increased NO 2 levels can be credited to the growing number of chemical factories and power plants.
In 2013, the government implemented two big projects: emission trading system (ETS) and Euro-4 standard for cars. 34,35When observing total NO 2 columns, the ETS and ICAP potentially contributed to the decrease in NO 2 over Almaty (20%) and Ekibastuz (17%) from 2012 to 2014 (Fig. 10).When the government transferred the Combined Heat and Power-2 (CHP-2) to the Almaty city territory in 2015, the NO 2 is shown to slightly increase in this area (14%). 36fter 2015, the optimization of technical processes power stations and employing natural gasdriven buses stabilized NO 2 level in hot spot cities, such as Almaty (19% change of NO 2 over the Almaty). 36n comparing the urban area of Almaty with the rural area of Esik (43.35°N, 77.46°E), satellite data show that NO 2 density in the rural area of Esik is four times lower than Almaty city for the period of 2005 to 2016.
In the capital city of Astana (formerly Akmola), located in the northern part of the country, shows an 11% decrease from 2005 to 2016 (Fig. 10).Despite Astana's relatively high population, the observed average NO 2 column during 2005 to 2016 is 1.2 × 10 15 molecules∕cm 2 , which is as low as rural areas elsewhere in the country (Table 1).
Figure 11 shows the monthly NO 2 tropospheric column over Kazakhstan in year 2014.During winter time there are some gray patterns, specifically in the north part, which shows missing data due to the extended snow cover.Therefore, in this study the evaluation of NO 2 concentration over northern part of country during winter is not reliable.Another important point from Fig. 11 is that NO 2 level changes over the whole country and hotspots during the year.The overall pattern over country changes from blue (low NO 2 ) to light blue (high  NO 2 ) as season changes from winter to summer, while pattern over hot spots changes from yellow (very high NO 2 ) to light blue (high NO 2 ) during this period.Figure 11    Trend line equation Figure 12 shows the seasonal behavior of NO 2 retrievals in cities, as well as the country average.The maximum NO 2 peak for Almaty, Shymkent, and Ekibastuz occurs during wintertime (at 9.85 × 10 15 molecules∕cm 2 , 5.30 × 10 15 molecules∕cm 2 , and 4.25 × 10 15 molecules∕cm 2 , respectively), which is consistent with Ref. 22. Two factors account for this wintertime peak in Kazakhstan: the heavier use of combustion power plants for heating and the longer lifetime of NO 2 in the atmosphere.In the summertime, urban to rural NO 2 density may vary by a factor of 4. This ratio is amplified in the winter, with urban areas showing 20× more NO 2 density than rural areas (Fig. 12).Rural regions are typically at the national average of NO 2 concentration, as the majority of Kazakhstan's area is rural.As seen in Fig. 12, the rural areas stay relatively stagnant in their NO 2 concentrations, with the maximum NO 2 occurring in the summertime, though the difference between maximum and minimum value is low (0.5 × 10 15 molecules∕cm 2 ).

Conclusion and Future Work
In this study, Kazakhstan is suggested as a case study for applying satellite data to assess air quality and energy sectors.As a country with rapid industrial and social growth, and without an established air quality monitoring network, the value of satellite data to assess emissions is particularly high.
This study analyzed retrievals of the vertical column density of NO 2 from the OMI over atmosphere of Kazakhstan between 2005 and 2016.Observations are coincided with other papers: (1) the dominant NO 2 source regions correspond primarily to areas with high population and large industry; (2) hot spots are strongly correlated with electricity usage and fuel sources; and (3) in northern hemisphere hot spots peak NO 2 is much higher during winter.OMI observations have shown that hotspot cities in Kazakhstan are industrial megacities with high population (Almaty, Shymkent) and a small town with thermal power plants providing 72% of electricity in Kazakhstan (Ekibastuz).The rest of the country has approximately the same amount of NO 2 concentration.From the rural sample region of Esik, the NO 2 levels stayed much lower year round, having concentrations that are 1∕4 (in summer) to 1∕20 (in winter) of its urban neighbor, Almaty.Significant increase in NO 2 over 2005 to 2016 was recorded in the growing cities of Almaty and Shymkent, while NO 2 columns in the rest of the country fluctuated insignificantly.From analyzing the seasonal behavior of NO 2 over Kazakhstan, we show a wintertime peak for hotspot cities due to the increased need for heat and increased lifetime of NO 2 .In comparing the annual columns in 2012 and 2014, it was possible to estimate that the implementation of the Euro-4 standard and the ETS reduced NO 2 up to 40% in Almaty, the city with the highest number of automobiles in country.Discrepancies for other two cities can be used to identify specific areas needing further studies.This work could be expanded by a comparison with ground-based monitoring data, model estimates of column NO 2 , and/or other satellite data products.As next-generation NO 2 data become available-include the Netherlands Space Office and European Space Agency Tropomi instrument launched in 2017, and the upcoming geostationary Sentinel 4 geostationary satellite will offer new sources of NO 2 over Kazakhstan and the region to improve our characterization of energy and emissions in this rapidly changing country.
The transboundary influence from other countries can be investigated in the future using modeling tools, such as GEOS-Chem.Particularly, biomass burning in China (East of Kazakhstan) during the winter time can introduce NO 2 to Kazakhstan as shown in Fig. 11, although the global wind in Kazakhstan is westerly wind.We will address this point as our future work using GEOS-Chem study.

Fig. 1
Fig. 1 Almaty Google map.Almaty is located in the valley of big and small Almatinka rivers at the foot of Trans-Ili Alatau mountains.

Fig. 4
Fig. 4 Shymkent Google map.The main sources of NO 2 over Shymkent come from stationary and mobile sources.

Fig.
Fig.2The population of Almaty has increased from 2005 to 2016 by 42%.24

Fig. 5
Fig. 5 Ekibastuz Google map.Ekibastuz is located in the Pavlodar region of the Republic of Kazakhstan.

Fig. 6
Fig.6An increase in population was observed for Ekibastuz from 2006 to 2016 by 8.6%.25

Fig. 8
Fig. 8 Astana Google Map.This shift in capital, construction, and transport by roads and railways increased dramatically, introducing increased sources of dust and air pollutions.

Fig. 10
Fig. 10 NO 2 column trend over Kazakhstan and its cities between 2005 and 2016.
also shares valuable information regarding the impact of majorly polluted neighbor countries including China and Russia on NO 2 density in Kazakhstan.It is understood from Fig.11that from October to February, Kazakhstan is influenced by the border regions of China and Russia.

Fig. 11
Fig. 11 Illustration of the monthly NO 2 tropospheric column over Kazakhstan in year 2014.December shows missing data due to extended snow coverage.

Table 1
The trend line equations for NO 2 variations from 2005 to 2016 for different cities of Kazakhstan.