Estimation of Surface NO₂ Volume Mixing Ratio in Four Metropolitan Cities in Korea Using Multiple Regression Models with OMI and AIRS Data

Abstract

Surface NO₂ volume mixing ratio (VMR) at a specific time (13:45 Local time) (NO₂ VMR_ST) and monthly mean surface NO₂ VMR (NO₂ VMR_M) are estimated for the first time using three regression models with Ozone Monitoring Instrument (OMI) data in four metropolitan cities in South Korea: Seoul, Gyeonggi, Daejeon, and Gwangju. Relationships between the surface NO₂ VMR obtained from in situ measurements (NO₂ VMR_In-situ) and tropospheric NO₂ vertical column density obtained from OMI from 2007 to 2013 were developed using regression models that also include boundary layer height (BLH) from Atmospheric Infrared Sounder (AIRS) and surface pressure, temperature, dew point, and wind speed and direction. The performance of the regression models is evaluated via comparison with the NO₂ VMR_In-situ for two validation years (2006 and 2014). Of the three regression models, a multiple regression model shows the best performance in estimating NO₂ VMR_ST and NO₂ VMR_M. In the validation period, the average correlation coefficient (R), slope, mean bias (MB), mean absolute error (MAE), root mean square error (RMSE), and percent difference between NO₂ VMR_In-situ and NO₂ VMR_ST estimated by the multiple regression model are 0.66, 0.41, −1.36 ppbv, 6.89 ppbv, 8.98 ppbv, and 31.50%, respectively, while the average corresponding values for the other two models are 0.75, 0.41, −1.40 ppbv, 3.59 ppbv, 4.72 ppbv, and 16.59%, respectively. All three models have similar performance for NO₂ VMR_M, with average R, slope, MB, MAE, RMSE, and percent difference between NO₂ VMR_In-situ and NO₂ VMR_M of 0.74, 0.49, −1.90 ppbv, 3.93 ppbv, 5.05 ppbv, and 18.76%, respectively.

1. Introduction

The main anthropogenic source of nitrogen dioxide (NO₂) is fossil fuel combustion, while natural sources of NO₂ include lightning, forest fires, and soil emissions [1,2]. In particular, since NO₂ is emitted in large quantities in automobile exhaust gas, NO₂ is often used as an indicator of traffic-related air pollution in urban areas [3]. In terms of its effect on human health, long-term NO₂ exposure can lead to respiratory depression and respiratory illness [4,5,6,7,8]. In addition, it is a precursor of aerosol nitrate, tropospheric ozone, and the hydroxyl radical (OH), the main atmospheric oxidant [9]. It is therefore important to measure NO₂ and various methods are used, with chemiluminescence, a well-known technique for measuring surface NO₂ volume mixing ratio (VMR) [10]. In situ measurements such as the chemiluminescence method are, in general, more accurate than remote sensing techniques, but require a large number of in situ instruments to provide the spatial distribution of the NO₂ VMR at high resolution [11]. In recent years, NO₂ vertical column density (VCD) has been measured from satellites that can monitor NO₂ at global scale over a short time scale. Space-borne sensors that have observed global distributions of NO₂ are the Global Ozone Monitoring Experiment (GOME) aboard European Remote Sensing-2 (ERS-2) (1995–2003), Scanning Imaging Absorption Spectrometer for Atmospheric Chartography/Chemistry (SCIAMACHY) aboard Environmental Satellite (Envisat) (2002–2012), the Ozone Monitoring Instrument (OMI) aboard EOS-AURA (2004–present), and GOME-2 aboard the Meteorological Operational satellite (MetOp)-A (2007–present) and MetOp-B (2012–present) [12,13,14,15,16,17]. In many countries, air quality regulation requires surface NO₂ VMR so the NO₂ VCD obtained from satellites cannot be used directly. In recent years, studies have been conducted to investigate the feasibility of estimating the surface NO₂ VMR using the NO₂ VCD obtained from satellite measurements and, in particular, the correlation between the NO₂ VCD obtained from satellite measurements and the surface NO₂ VMR.

Ordónez et al. [18] reported the correlation between tropospheric NO₂ VCD and the NO₂ VCD measured by GOME and ground based in situ devices in Milan. Kharol et al. [3] estimated the annual average ground-level NO₂ concentrations in North America using chemical transport model (GEOS-Chem) data and OMI NO₂ columns and also reported the annual trend of the estimated ground-level NO₂ concentrations. However, no studies have attempted to estimate the surface NO₂ VMR at higher temporal resolutions such as hourly and monthly using the NO₂ VCD measured by satellites.

In this present study, we estimate for the first time the surface NO₂ VMR at a specific time (13:45 Local time (LT)) (NO₂ VMRST) and the monthly mean surface NO₂ VMR (NO₂ VMRM) using two linear regression models and a multiple regression model with the tropospheric NO₂ VCD obtained from OMI (Trop NO₂ VCDOMI) in five metropolitan cities. In addition, the performance of each regression method is evaluated by comparing the estimated surface NO₂ VMRs with those obtained from in situ measurement (NO₂ VMRIn-situ).

2. Study Area and Period

A large amount of anthropogenic NO_X is emitted in Northeast Asia including China, Korea and Japan [19]. Especially, the annual mean NO₂ tended to increase in Seoul from 1995 to 2009 [20]. The study areas were selected where the surface NO₂ VMR is continuously measured in Korean metropolitan cities (Figure 1). Metropolitan cities such as Busan and Incheon where the OMI pixel covers both sea and land are excluded since there are no surface NO₂ data available over the sea. Therefore, the selected areas are Seoul, Gyeonggi, Daejeon, and Gwangju. Seoul is covered by four OMI pixels and is divided into eastern and western areas (West Seoul and East Seoul). The study period is the nine years from 2006 to 2014. This is split into a seven-year training period (2007–2013) to determine the coefficients of the regression models used in this study, and two years of validation (2006 and 2014) when the surface NO₂ VMRs estimated from the resulting three regression models are evaluated by comparison with the in situ data. The three regression models used in this study are described in detail in Section 3.

2.1. Data

The data used in this study are Trop NO₂ VCD_OMI and Atmospheric Infrared Sounder (AIRS) boundary layer height (BLH_AIRS), atmospheric temperature (Temp_AIRS) and pressure (Press_AIRS), together with in situ measurements of NO₂ VMR_In-situ, surface temperature (Temp_In-situ), surface pressure (Press_In-situ), surface dew point (Dewpoint_In-situ), surface wind speed (WS_In-situ), and surface wind direction (WD_In-situ) (see

Table 1. Satellite and in situ data used in this study.

	Data		Time (LT)
Satellite	Trop NO2 VCD	OMI Level3 NO2 Daily data (OMNO2d)	13:45
	BLH, Temperature, Pressure	AIRS/Aqua L3 Daily Support Product (AIRS + AMSU) V006 (AIRX3SPD)	13:30
In situ	Surface NO2 VMR	Air Korea	13:00 and 14:00
	Surface Temperature, Surface Pressure, Surface Dew point, Surface Wind Speed, Surface Wind Data	AWS (Automatic Weather System)

).

2.1.1. Ozone Monitoring Instrument (OMI) Data

The Trop NO₂ VCD_OMI data were obtained from OMI Level3 NO₂ Daily Data (OMNO2d) provided by the NASA Goddard Earth Sciences Data and Information Services Center (http://disc.sci.gsfc.nasa.gov/Aura/data-holdings/OMI) [17,21,22]. OMI is a nadir-viewing UV–visible (270–500 nm) spectrometer aboard the Aura platform launched in July 2004 [23]. Aura is a polar orbiting satellite with an overpass time of 13:45 LT. The spectral resolution of the OMI is about 0.5 nm and the spatial resolution is 13 × 24 km at nadir. Cloud-screened NO₂ data (Level-3 OMI NO₂ Cloud-Screened Total and Tropospheric Column NO₂ (V003)) are used in the present study (Cloud Fraction <30%).

2.1.2. Atmospheric Infrared Sounder (AIRS) Data

The BLH_AIRS, Temp_AIRS, and Press_AIRS used in this study were obtained from the AIRS/Aqua L3 Daily Support Product (AIRS + AMSU) 1 degree × 1 degree V006 (AIRX3SPD.00) from NASA Goddard Earth Sciences Data and Information Services Center (http://disc.sci.gsfc.nasa.gov/uui/datasets/AIRX3SPD_V006/summary?keywords=%22AIRS%22) [24,25,26]. The AIRS/Advanced Microwave Sounding Unit (AMSU) is a sounding suite launched in May 2002 aboard Aqua [26,27]. Aqua is a polar orbiting satellite with an overpass time of 13:30 LT and a horizontal spatial resolution of 40 km at nadir.

2.1.3. In Situ NO₂ Data

The NO₂ VMR_In-situ data were obtained from Air Korea (http://www.airkorea.or.kr/last_amb_hour_data). Since NO₂ VMR_In-situ is available hourly, the average of the values at 13:00 and 14:00 LT is used to be closer to the OMI overpass time. In a previous study [18], the in situ measurements were grouped into five different NO₂ levels: clean, slightly polluted, averagely polluted, polluted, and heavily polluted. Many stations are located close to roads and are exposed to emissions. In addition, the in situ NO₂ data from stations within GOME pixels (320 × 40 km) were averaged, since in situ measurements are only representative of a small fraction of the satellite ground scene. In the present study, the NO₂ VMR_In-situ obtained from in situ measurements located close to streets were excluded in this study. We used the average of three or more NO₂ VMR_In-situ from stations located at least 2 km from each other.

2.1.4. In Situ Meteorological Data

The Temp_In-situ, Press_In-situ, Dewpoint_In-situ, WS_In-situ, and WD_In-situ used in this study are Automatic Weather System (AWS) data provided by the Korea Meteorological Administration (http://sts.kma.go.kr/jsp/home/contents/statistics/newStatisticsSearch.do?menu=SFC&MNU=MNU). Since meteorological data are available hourly, the average of the data at 13:00 LT and 14:00 LT is used. The surface wind data, especially wind direction can be impacted by local topography and interferences.

3. Methodology

In this study, NO₂ VMR_ST and NO₂ VMR_M were estimated using three regression models with Trop NO₂ VCD_OMI.

Table 2. Regression models used for surface NO₂ VMR estimation in this study.

Model		Equation
M1	13:45 LT and Monthly	${\mathrm{NO}}_2{\mathrm{VMR}}_{insitu}=aTropN{O}_{2}VC{D}_{OMI}^{\left(a\right)}+b$
M2	13:45 LT and Monthly	${\mathrm{NO}}_2{\mathrm{VMR}}_{insitu}=aBLHN{O}_{2}VM{R}_{OMI}^{\left(b\right)}+b$
M3	13:45 LT	Section 3, Multiple regression Equation (1)
M4	Monthly

Notes: (a) NO₂ tropospheric vertical column density obtained from OMI; and (b) ${\mathrm{BLH}\mathrm{NO}}_{2}VM{R}_{OMI}=\frac{TropN{O}_{2}VC{D}_{OMI}\mathrm{Gas}\mathrm{constant}\mathrm{R}Tem{p}_{AIRS}\times {10}^{13}}{\mathrm{Avogadro}\mathrm{constant}\mathrm{NA}BL{H}_{AIRS}Pres{s}_{AIRS}}$ , where the AIRS pressure and temperature are boundary layer mean values, Gas constant R = $8.314472{\mathrm{m}}^3{\mathrm{pa}\mathrm{K}}^{-1}{\mathrm{mol}}^{-1}$ and $\mathrm{Avogadro}\mathrm{constant}$ NA = $6.022\times {10}^{23}{\mathrm{mol}}^{-1}$.

summarizes the three models.

3.1. M1

M1 is the linear regression equation where Trop NO₂ VCD_OMI is used as the independent variable. Figure 2 shows the linear regression between Trop NO₂ VCD_OMI and NO₂ VMR_In-situ at 13:45 LT during the training period, with R² (coefficient of determination), slope, and intercept of 0.47, 0.80 and 11.47, respectively. Figure 3 shows the linear regression between monthly mean Trop NO₂ VCD_OMI and monthly mean NO₂ VMR_In-situ during the training period, with R², slope, and intercept of 0.62, 0.77, and 10.95, respectively. The final form of the M1 equation for estimating NO₂ VMR_ST is shown in

Table 3. Final form of the regression models used for estimating surface NO₂ VMR at a specific time and R² obtained from the regression between NO₂ VMR_In-situ and the corresponding independent variable for the training period.

		Equation	R2
13:45 LT	M1	$N{O}_{2}VM{R}_{ST}=1.71\times TropN{O}_{2}VC{D}_{OMI}-0.68$	0.47
	M2	$N{O}_{2}VM{R}_{ST}=4.19\times BLHN{O}_{2}VM{R}_{OMI}+1.57$	0.38
	M3	$\begin{array}{ll}N{O}_{2}VM{R}_{ST}& =0.000602\times TropN{O}_{2}VC{D}_{OMI}-0.000107\times Tem{p}_{In-situ}\\ & -0.000083\times Dewpoin{t}_{In\text{-}situ}+0.000061\times Pres{s}_{In\text{-}situ}\\ & -0.000002\times BL{H}_{AIRS}-0.002435\times W{S}_{In\text{-}situ}\\ & +0.001190\times WD{}_{In\text{-}situ}-0.039996\end{array}$	0.47

, and that for estimating NO₂ VMR_M in

Table 4. As Table 3 but for monthly mean surface NO₂ VMR.

		Equation	R2
Monthly mean	M1	$N{O}_{2}VM{R}_M=1.23\times TropN{O}_{2}VC{D}_{OMI}+4.74$	0.62
	M2	$N{O}_{2}VM{R}_M=2.92\times BLHN{O}_{2}VM{R}_{OMI}+6.74$	0.59
	M4	$\begin{array}{ll}N{O}_{2}VM{R}_M& =0.657241\times TropN{O}_{2}VC{D}_{OMI}-0.137334\times Dewpoin{t}_{In\text{-}situ}\\ & -0.136096\times Pres{s}_{In\text{-}situ}-0.004331\times BL{H}_{AIRS}-0.770356\times W{S}_{In\text{-}situ}\\ & +2.370956\times WD{\left(west\right)}_{In\text{-}situ}+157.361668\end{array}$	0.63

.

Table 3 and Table 4 show the equations M1, M2, M3, and M4 with the regression coefficients determined from the training period.

3.2. M2

There might exist a minor fraction of the tropospheric NO₂ column in upper troposphere particularly because of lightning. However, the NO₂ amount in upper troposphere could be considered negligible in metropolitan cities, where a significant amount of NO_X is emitted. Therefore, assuming Trop NO₂ VCD_OMI is mostly present within the PBL, the relationship between Trop NO₂ VCD_OMI and the surface NO₂ VMR may change as the PBL varies. However, a minor fraction of the tropospheric NO₂ column can also be in the upper tropospheric, particularly because of lightning. This NO₂ fraction in upper tropospheric might cause either small or negligible reduction in correlations of the OMI NO₂ VCD between and surface NO₂ VMR as the upper part of the troposphere (free troposphere) contribution is assumed to be negligible [28]. To reflect the BLH in the regression equation, Trop NO₂ VCD_OMI is first divided by BLH_AIRS to calculate the NO₂ concentration in the PBL and then converted to the NO₂ mixing ratio in the PBL (BLH NO₂ VMR_OMI) using Temp_AIRS and Press_AIRS [29] as shown Table 2. Only a single OMI pixel contained completely within an AIRS pixel was used. Figure 4 shows the linear regression between BLH NO₂ VMR_OMI and NO₂ VMR_In-situ at 13:45 LT during the training period. Here R², slope and intercept are 0.38, 1.58, and 14.30, respectively. Figure 5 shows the corresponding linear regression for the monthly mean data, with R², slope and intercept of 0.59, 1.71, and 12.75, respectively. The final form of equation M2 to estimate NO₂ VMR_ST is shown in Table 3, and for the monthly values in Table 4.

3.3. M3 and M4

M3 and M4 are multiple regression equations for estimating NO₂ VMR_ST and NO₂ VMR_M. Multiple regression equations consist of a dependent variable, independent variables, and their regression coefficients. In addition to Trop NO₂ VCD_OMI and BLH_AIRS, meteorological factors (surface temperature, dew point, atmospheric pressure, wind direction, and wind speed) are used as candidate independent variables for the multiple regression equation in the present study. In a previous study [30], these meteorological factors were also used as candidate independent variables to estimate surface SO₂ concentration in Shanghai, China. Temperature, pressure, boundary layer height, wind speed, and wind direction were selected as the candidates for independent variables since they are known to either directly or indirectly affect the spatial mixing of NO₂ molecules in boundary layer. Furthermore, temperature and dewpoint were selected as candidates for independent variables as they affect the boundary layer height [31].

The multiple regression equation can be defined by the following equations:

$$\widehat{y}={\beta }_0+{\beta }_1{x}_1+{\beta }_2{x}_2+\dots +{\beta }_n{x}_n+\epsilon$$

(1)

where $\widehat{y}$ and β₀ are the dependent variable (NO₂ VMR_In-situ) and regression coefficient, respectively; $x_1$, $x_2$,..., $x_n$ are the candidate independent variables (Trop NO₂ VCD_OMI, Dewpoint_In-situ, Press_In-situ, Temp_In-situ, BLH_AIRS, WS_In-situ, and WD_In-situ); β₁, β₂, …, β_n are the regression coefficients of the independent variables; and $\epsilon$ is the difference between observations (NO₂ VMR_In-situ) and estimated values (NO₂ VMR_estimate). The regression coefficients can be estimated by least square fitting:

$${\displaystyle \sum _{j=1}^m}{\epsilon }_j^2={\displaystyle \sum _{j=1}^m}{\left(y_j-\widehat{y_j}\right)}^2$$

(2)

where y_j is the observed value with m data points. By minimizing the sum of ${\epsilon }^2$, regression coefficients can be derived. These least square fitting techniques are based on the following assumptions: the linear relationship, a normal distribution and equal variance in the residuals. The least squares regression is sensitive to the presence of some points that are excessively large or small values in the training data [32]. To determine the independent variables ($x_n$) and regression coefficients (${\beta }_n$) included in the final form of equations M3 and M4, we considered the variation inflation factor (VIF) and p-value to ensure their statistical significance. First, we examined the VIF that explains the multicollinearity of a candidate independent variable with regard to other candidate independent variables. The VIF of the j-th independent variable is expressed as:

$$\mathrm{VIF}\left(x_j\right)=\frac{1}{1-R_j^2}$$

(3)

where $R_j^2$ is the coefficient of determination for the regression of $x_j$ against another independent variable (a regression that does not involve the dependent variable $j$). The VIF indicates how much $x_j$ is correlated with the other candidate variables. A candidate independent variable with a very high VIF can be considered redundant and should be removed from the multiple regression equations. Candidate independent variables that do not satisfy the criterion VIF < 10 [33], were excluded from the independent variables. The p-value was also used to select independent variables. The highest still statistically significant p-level was shown by Sellke et al. [34] to be 5%. Among the independent variables that satisfy the VIF criterion, those that also satisfy p-value <0.05 are selected as final independent variables in the multiple regression equations. The independent variables selected for equations M3 and M4 are shown in

Table 5. Final independent variables included in multiple regression equations (M3 and M4).

	Final Selected Independent Variables	p-Value	VIF
M3	Trop NO2 VCDOMI	0	1.26
	TempIn-situ	0.000032	7.02
	DewpointIn-situ	0.000306	7.16
	PressIn-situ	0.009981	3.14
	BLHAIRS	1.73 $\times$ 10−15	1.12
	WSIn-situ	3.86 $\times$ 10−133	1.33
	WDIn-situ	1.7493 $\times$ 10−38	1.07
M4	Trop NO2 VCDOMI	2.4832 $\times$ 10−89	1.64
	DewpointIn-situ	0.000421	6.47
	PressIn-situ	0.034582	6.65
	BLHAIRS	0.000834	2.32
	WSIn-situ	3.86 $\times$ 10−133	1.59
	WDIn-situ	1.699 $\times$ 10−7	1.25

. The final form of equation M3 to estimate NO₂ VMR_ST is shown Table 3, and that for NO₂ VMR_M in Table 4.

4. Results

4.1. Daily Estimates

Figure 6 shows the day-to-day variations of NO₂ VMR_In-situ and NO₂ VMR_ST estimated at 13:45 LT in West Seoul and East Seoul using M1, M2 and M3 in Table 3 for 2006 and 2014. A slightly larger difference in magnitude is found between NO₂ VMR_In-situ and NO₂ VMR_ST obtained with M3 compared to those between NO₂ VMR_In-situ and NO₂ VMR_ST obtained with M1 and M2. However, NO₂ obtained from M3 showed moderate agreement with NO₂ VMR_In-situ in the form of the day-to-day variation. Results for Daejeon, Gwangju, and Gyeonggi are included in the Supplementary Materials.

Figure 7 shows the R, slope, mean bias (MB), mean absolute error (MAE), root mean square error (RMSE) and percent difference between NO₂ VMR_ST and NO₂ VMR_In-situ for the validation period (2006 and 2014). The R obtained with M1 ranges from 0.49 to 0.71, showing better agreement than that with M2 (0.47 < R < 0.65). M3 showed the best correlation with NO₂ VMR_In-situ (0.67 < R < 0.90). The slopes from both M1 and M2 are close to one in East Seoul, whereas they are lower in the other cities. The MB from M1, M2, and M3 ranges from −7.74 to 5.80 ppbv. In all study areas, the MAE (5.79 ppbv < MAE < 8.25 ppbv) of M3 is lower than those (6.58 ppbv < MAE < 11.41 ppbv) of M1 and M2, which means that NO₂ VMR_ST estimated from M3 show moderate agreement with NO₂ VMR_In-situ in terms of magnitude. The RMSE from M3 is found to be lower than those from M1 and M2. The NO₂ VMR_ST from M3 have the lowest RMSE in all study areas (7.21 ppbv < RMSE < 11.37 ppbv). In addition, percent differences estimated from M3 and NO₂ VMR_In-situ are lower in all study areas than from M1 and M2. In estimating NO₂ VMR_ST, M3, which is a multiple regression method with various independent variables as inputs, generally showed good statistical performance except for MB.

4.2. Monthly Estimates

Figure 8 shows the temporal variation of monthly mean NO₂ VMR_In-situ and NO₂ VMR_M estimated using M1, M2 and M4 of Table 4 in West Seoul and East Seoul using monthly mean independent variables during the validation period (see the detailed input data in Section 2.1). Figure 8 shows good agreement in terms of the temporal pattern between the estimated NO₂ VMR_M and monthly mean NO₂ VMR_In-situ. However, we found a large difference between NO₂ VMR_In-situ and NO₂ VMR_M in periods when there was a jump in NO₂ VMR_In-situ between successive months. For example, no models calculated NO₂ VMR_M that were similar to NO₂ VMR_In-situ in December 2006, which is very different from that in November 2006. NO₂ VMR_In-situ (NO₂ VMR_M from M1, M2, and M4) in November and December in 2006 are 19.32 ppbv (15.94, 17.96, and 17.62 ppbv) and 30.30 ppbv (15.94, 17.96, and 17.62 ppbv) in Daejeon, 15.26 ppbv (12.29, 13.87, and 18.09 ppbv) and 32.55 ppbv (12.73, 14.57, and 18.46 ppbv) in Gwangju, 29.31 ppbv (25.86, 25.97, and 22.35 ppbv) and 40.64 ppbv (29.91, 29.15, and 26.85 ppbv) in Gyeonggi, and 31.25 ppbv (22.80, 24.55, and 23.64 ppbv) and 45.93 ppbv (28.65, 28.92, and 26.49 ppbv) in West Seoul. Especially in West Seoul, there are several periods when NO₂ VMR_In-situ changes rapidly compared with the previous month. The NO₂ VMR_M obtained from the three models at these times are in poor agreement with the pattern of monthly NO₂ VMR_In-situ. As described in Section 2, despite the use of NO₂ VMR_In-situ located away from the streets, the in situ measurement sites in West Seoul are located closer to the streets than the in situ measurement sites in Daejeon and Gwangju. This may explain why there are more periods when NO₂ VMR_In-situ changes rapidly from one month to the next. It is difficult to estimate the rapid change of NO₂ VMR near the NO₂ source using regression models that reflect the relationship between the in situ measurements and the OMI sensor covering both source and non-source areas in a single pixel.

Figure 9 shows the R, slope, MB, MAE, RMSE and percent difference between NO₂ VMR_M and monthly mean NO₂ VMR_In-situ in 2006 and 2014. In general, NO₂ VMR_M agreed better with NO₂ VMR_In-situ than did the NO₂ VMR_ST. The value of R from M1, M2 and M4 and monthly mean NO₂ VMR_In-situ ranged from 0.68 to 0.82 in all areas. MB was close to 0 in most study areas. MAE was less than 5 ppbv in Daejeon, Gwangju, Gyeonggi, and East Seoul where there is good agreement between NO₂ VMR_M from M1, M2, and M4 and monthly mean NO₂ VMR_In-situ, whereas MAEs in West Seoul ranged from 5.66 to 6.79. RMSEs between NO₂ VMR_In-situ and NO₂ VMR_M from M1, M2, and M3 are found to be lower than 7 ppbv in the study areas except for West Seoul. In addition, the three models showed percent differences of less than 30% except for the value estimated from M1 in Gwangju.

5. Discussion

In a previous study [18], tropospheric NO₂ VCDs obtained from GOME were compared with tropospheric NO₂ VCDs calculated using NO₂ concentrations obtained from both in situ measurements and the Model of Ozone and Related Tracers 2 (MOZART-2). There are also several previous studies estimating surface NO₂ VMR using satellite data [3,35]. Among them, Kharol et al. [3] estimated the annual variation of ground-level NO₂ concentrations using both GEOS-Chem data and OMI data. However, in the present study, NO₂ VMR_ST and NO₂ VMR_M were estimated for the first time at higher temporal resolution using three regression models with Trop NO₂ VCD_OMI as input.

5.1. Estimation of Surface NO₂ VMRs at a Specific Time (13:45 LT)

• Among the three regression models, the multiple regression model M3 performed best in estimating NO₂ VMR_ST. The linear regression model (M2), in which BLH is used as an independent variable in addition to Trop NO₂ VCD_OMI, has comparable performance to that of the model (M1) which uses Trop NO₂ VCD_OMI as the only independent variable.The BLH varies with latitude [36], but the latitudinal variation of BLH is not well represented since the spatial resolution of the AIRS used in this study is coarser than the spatial resolution of OMI. It might also be associate the BLH_AIRS data quality. We expect better results using BLH data obtained from LIDAR.
• The average difference was found to be 46.04% between NO₂ VMR_In-situ and NO₂ VMR_ST obtained from M1, 44.29% between NO₂ VMR_In-situ and NO₂ VMR_ST obtained from M2, and 31.50% between NO₂ VMR_In-situ and NO₂ VMR_ST obtained from M3 in all cities, while there was moderate agreement in the temporal pattern of NO₂ variation between NO₂ VMR_In-situ and NO₂ VMR_ST obtained from M1, M2, and M3 (Figure 6).
• In terms of statistical evaluation with respect to the in situ data, M3 showed the best performance in general.
• The results produced by M2 are not improved compared to those by M1 which may imply that surface NO₂ VMR is dominantly affected by tropospheric NO₂ column while the BLH effect could be negligible in areas of the present study. It might also be associate the AIRS BLH data quality.

5.2. Estimation of Monthly Mean Surface NO₂ VMRs of a Specific Time (13:45 LT)

• We found good agreement in the temporal pattern between the estimated NO₂ VMR_M and monthly mean NO₂ VMR_In-situ (Figure 8). However, there was a large difference between NO₂ VMR_In-situ and NO₂ VMR_M in the period when there was a clear change in NO₂ VMR_M between one month and the next. Despite the use of NO₂ VMR_In-situ located away from streets, the in situ measurement sites in West Seoul are located closer to streets than the in situ measurement sites in Daejeon and Gwangju. This may explain why there are more periods when NO₂ VMR_In-situ changes rapidly in successive months. It is difficult to estimate the rapid change of NO₂ VMR near NO₂ sources with regression models that reflect the relationship between the in situ measurements and the OMI sensor covering both source and non-source areas in a single pixel.
• In terms of statistical evaluation, the three regression models (M1, M2, and M4) were found to be similar (Figure 9).
• NO₂ VMR_M shows better agreement with the NO₂ VMR_In-situ than does NO₂ VMR_ST. The reason for the better performance in the monthly mean estimation could be attributed to reduced errors in the monthly mean OMI data [37] as well as fewer occasions with sudden monthly changes in NO₂ VMR_In-situ than rapid day-to-day changes in NO₂ VMR_In-situ.

This present study provides the results in the condition of 2 km distance between the in situ NO₂ measurement location and NO_X point source. For a future study, performances of the models need to be investigated depending on the distance between the in situ NO₂ data and point sources. We expect that the regression methods used to estimate the surface NO₂ VMR using Trop NO₂ VCD_OMI will be useful in providing information on surface NO₂ VMR in metropolitan cites on a monthly timescale. In future research, the estimation of surface NO₂ VMR may be attempted at higher time resolution with geostationary satellite sensors (e.g., geostationary environmental monitoring spectrometer (GEMS), tropospheric emissions: monitoring of pollution (TEMPO), and Sentinel-4). In further work, improvements are needed in the input data or the model formulation before the surface NO₂ can be estimated on a daily basis.

6. Conclusions

In this study, monthly and specific time estimates of NO₂ VMR were obtained for the first time using three regression models in four metropolitan cities for two years, 2006 and 2014. The multiple regression model (M3) was found to perform best in estimating NO₂ VMR_ST in all cities. For surface NO₂ estimates at the specific time (13:45 LT), M3 generally gives better R, MAE, RMSE, and percent difference than the other two models (M1 and M2). A comparison between monthly surface NO₂ VMR estimates and those at the specific time showed that agreement with NO₂ VMR_In-situ was better for monthly estimates. In estimating NO₂ VMR_M, three regression models (M1, M2, and M4) showed similar performance. In estimating daily and monthly surface NO₂ VMR variations, when the surface NO₂ VMR changes rapidly, the difference between surface NO₂ VMR estimated from all models and NO₂ VMR_In-situ is found to be large. In future studies, using higher spatial resolution satellites is expected to improve the relationship with in situ measurements. In addition, the use of other independent variables that may co-vary with rapid changes of surface NO₂ VMR should be investigated.