CC BY
17
RUSSIAN JOURNAL OF EARTH SCIENCES, VOL. 11, ES4004, doi:10.2205/2010ES000437, 2010
Tropospheric NO2 trend over St. Petersburg (Russia) as measured from space
D. V. Ionov1
Received 23 March 2010; accepted 7 April 2010; published 29 April 2010.
For the period 1996-2009, combined data set of tropospheric NO2 measurements by satellite instruments GOME, SCIAMACHY and OMI has been analyzed to derive possible trend over the city of St. Petersburg in Russia. The time series of the monthly NO2 columns for these 14 years have been fitted with a linear function superposed on an annual seasonal cycle. Resulting trend estimate provides a linear growth of ~ 4% per year, which generally correlates with an independent assessments done by NOx inventory studies. KEYWORDS:
Atmospheric nitrogen dioxide; satellite remote sensing; tropospheric pollution.
Citation: Ionov, D. V. (2010), Tropospheric NO2 trend over St Sci., 11, ES4004, doi:10.2205/2010ES000437.
Introduction
Nitrogen dioxides (NOx = NO+NO2) have a large impact on human health, climate and atmospheric chemistry. In the dense populated regions NOx are intensively emitted by all combustion processes and contribute to photochemical smog production and nitric acid deposition. Starting from 1995 the global mapping of tropospheric NO2 has become possible with help of satellite instruments GOME (1995-2003) [Burrows et al., 1999] aboard ERS 2, SCIAMACHY (from 2002) aboard Envisat [Bovensmann et al., 1999], OMI (from 2004) aboard EOS-Aura [Levelt et al., 2006] and GOME 2 aboard MetOp [Munro et al., 2006]. A number of studies have been devoted to detection of trends in tropospheric NO2 over different regions of globe. To the first-order approximation the local changes in NO2 columns can be assumed to be proportional to changes in local NOx emission. Thus, persistent efforts to reduce emissions from traffic and industry, adopted in a number of developed countries (e.g. European Union and USA), are expected to show the decrease in tropospheric NO2 column. On the contrary, the regions of rapid economic development must be characterized with tropospheric NO2 growth (e.g. East Asia). Indeed, several studies have revealed statistically significant negative NO2 trend over Western Europe and Eastern USA, and positive trend over China. Richter et al. [2005] analyzed tropospheric NO2 columns retrieved from GOME and SCIAMACHY in 1996-2004 and found substantial reduction in NO2 over Europe (~ 20%) and Central East Coast of the
1Faculty of Physics, St.Petersburg State University, St. Petersburg, Russia
Copyright 2010 by the Russian Journal of Earth Sciences. http://elpub.wdcb.ru/journals/rjes/doi/2010ES000437.html
Petersburg (Russia) as measured from space, Russ. J. Earth.
USA (~ 10%), but significant increase of ~ 50% over the industrial areas of China. Van der A et al. [2006] studied trend in NO2 using the data of similar satellite measurements in 1996-2005, and showed large growth over eastern China - up to 20% per year (city of Shanghai). Further, Van der A et al. [2008] extended that study to the 10-years data set of satellite observations in 1996-2006 and revealed reductions in NO2 over Europe and eastern USA (up to 7% per year), with a strong increase over Asia (up to 29% per year in China), but also some industrial regions of Siberia in Russia - up to 11% per year (city of Novosibirsk).
Here the 14-years combined data set of NO2 measurements by GOME, SCIAMACHY and OMI in 1996-2009 is analyzed to quantify the trend in tropospheric NO2 column over the city of St. Petersburg in Russia. First, a brief description of satellite instruments and tropospheric NO2 retrieval is given. Further, the data analysis is described and an overview of long-term NO2 time series, acquired over St.Petersburg in 1996-2009, is given. Then, accomplished trend analysis of tropospheric NO2 is presented. Finally, the results of study are discussed and further work is proposed.
Tropospheric NO2 Measurements From Space
Along with ozone, the global monitoring of NO2 is a key objective of NASA’s Ozone Monitoring Experiment (OMI). In this regard, OMI succeeds two other space experiments
- ESA’s GOME and SCIAMACHY. GOME (Global Ozone Monitoring Experiment) was launched in April 1995 onboard the ERS 2, into a sun-synchronous polar orbit [Burrows et al., 1999]. It is a nadir-viewing UV-visible spectrometer that measures the solar irradiance and the solar radiation
0 1 15 2 3 4 5 6
[10 molecules cnr2]
Figure 1. Annual map of tropospheric NO2 vertical column over the Baltic Sea region, produced from the global data of Aura OMI monthly mean tropospheric NO2 in 2004-2009 (available at http://www.temis.nl).
backscattered from both the atmosphere and the Earth’s surface. Besides providing global maps of atmospheric ozone, GOME was also the first spaceborne instrument capable of measuring the total column amount of NO2. The trace gas column densities given in the GOME data products (ozone and NO2), are derived by the Differential Optical Absorption Spectroscopy (DOAS) technique [Platt, 1994]. GOME was followed by the SCanning ImAging spectroMeter for Atmospheric CartograpHY (SCIAMACHY) onboard ENVISAT in 2002, which could also measure NO2 vertical columns [Bovensmann et al., 1999].
The work presented here is based on tropospheric NO2 columns retrieved from the satellite data by the Royal Netherlands Meteorological Institute [Boersma et al., 2004, 2007]. The algorithm is based on the
combined retrieval-assimilation-modelling approach. A chemistry-transport model, driven by high-quality meteo-
rological fields and assimilated NO2 information from previously observed orbits, provides best guess profiles of NO2. These model forecast fields are collocated with satellite observations, and the radiative transfer modelling in the retrieval is performed based on the model trace gas profile and temperature profiles. The modelled stratospheric NO2 distribution is employed to derive a tropospheric column by subtracting the modelled (assimilated) stratosphere from the measured column. The retrieval is coupled to cloud top height and cloud fraction retrievals derived from the satellite data, and also coupled to high quality albedo maps. The global dataset of such monthly averaged tropospheric NO2 columns retrieved from the GOME, SCIAMACHY and OMI instruments is publicly available through (http://www.temis.nl).
For better accuracy, only measurements with an estimated cloud fraction less than ~ 0.2 are used in the calculation of monthly mean NO2 fields. Detailed error budget of tropospheric NO2 retrieval is presented in the work of Boersma et al. [2004]. According to that study, a meaningful estimate of the tropospheric vertical column over the heavily polluted continental areas can be given with a precision of 35-60%.
Long-Term NO2 Data Over St. Petersburg
As an illustration of the opportunity to monitor urban emissions from space, an annual distribution of tropospheric NO2 over the Baltic Sea region was averaged for the years 2004-2009 of OMI measurements (see Figure 1). The map shows almost no NO2 in the troposphere of northern Europe (< 1 x 1015 molecules cm-2), but an area of high values in the south-west of the region - Belgium, Netherlands and Germany (up to ~ 18 x 1015 molecules cm-2), and a spots of polluted air over the cities of Moscow (up to ~ 18 x 1015 molecules cm-2) and St. Petersburg (up to ~ 9 x 1015 molecules cm-2).
To analyze NO2 trend over St. Petersburg, the data records from 3 satellite instruments were assembled: the GOME data in 04/1996-06/2003, the SCIAMACHY data in 07/2002-09/2009 and the OMI data in 10/2004-09/2009. The monthly mean tropospheric NO2 columns, available at TEMIS website are gridded on a different scale, depending on the instrument - 0.25° x 0.25° for GOME and SCIAMACHY and 0.125° x 0.125° for OMI. To filter out the effects of different spatial resolution, rectangle of 1 ° latitude by 2° longitude centered over the city of St. Petersburg (59.95°N, 30.32°E) was selected to average the individual grid pixels. Resulting time series of monthly mean data by GOME, SCIAMACHY and OMI over St. Petersburg are shown in Figure 2. The measurements demonstrate high month-to-month variation - tropospheric NO2 column is changing from ~ 1 to ~ 10 x 1015 molecules cm-2.
To achieve more significant results, both GOME and SCIAMACHY data are used in 2002-2003, and both SCIAMACHY and OMI are used since October 2004. However, to homogenise the full data record 1-year periods of overlap measurements (by SCIAMACHY and OMI in
Figure 2. Time series of monthly mean tropospheric NO2 column, as measured by ERS-2 GOME, ENVISAT SCIAMACHY and Aura OMI over St. Petersburg. The merged data for the full period of 1996-2009 are shown by grey curve.
10/2004-09/2005, GOME and SCIAMACHY in 07/200206/2003) were used to fit and apply the necessary adjustments. For this, the corresponding monthly data were spatially convolved from their original resolution to the uniform grid of 1°(latitude) x2° (longitude), and the yearly averaged field of tropospheric NO2 column over Eurasia (40-80° N, 0-180°E) was calculated for each instrument. Resulting maps
- GOME and SCIAMACHY in 2002-2003, SCIAMACHY and OMI in 2004-2005 - are presented on Figure 3, demonstrating reasonable agreement between satellite instruments. Thus, the correlation coefficient for the pixel-to-pixel comparison results in 0.95 for GOME versus SCIAMACHY and 0.96 for SCIAMACHY versus OMI. Regression equations, computed from these comparisons, were used to adjust GOME data to SCIAMACHY, and SCIAMACHY to OMI. The average of two measurements, when available, was taken
- GOME and SCIAMACHY in 2002-2003, SCIAMACHY and OMI since 2004. Resulting time series of tropospheric NO2 column over St. Petersburg is plotted as a separate line on Figure 2 (“merged data”).
Trend Analysis
Following similar studies [Ionov et al., 2009; Poberovskii et al., 2007], a simple statistical model was used to describe long-term variation of tropospheric NO2 column:
Yt = A + BXt + Cßt + Nt
(1)
where Yt represents the monthly NO2 column of month t, Xt is the number of months after January 1996, Nt is the residual noise, A represents the NO2 column in January 1996, B is the monthly trend in NO2, and C is the amplitude of seasonal component ^t. Thus, coefficients A, B and C are the fit parameters of the multiple linear regression.
The precision of the trend B is estimated from the variance œn of remainder, the data set length n (in years) and autocorrelation of Nt, [Weatherhead et al., 1998]:
OB
O N
(2)
4 of 7
0 1 15 2 3 4 5 6
[10 molecules crrr2]
Figure 3. Comparison of annual tropospheric NO2 by satellites, spatially convolved from their original resolution to the uniform grid of l°(latitude)x2°(longitude): GOME (left top) versus SCIAMACHY (right top) in 2002-2003, SCIAMACHY (left bottom) versus OMI (right bottom) in 2004-2005.
ES4004 ionov: tropospheric no2 trend ES4004
i:
O
z
0_n------1---1----1----1-----1----1----1----1----1----1-----1----1
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN
Figure 4. Seasonal component of tropospheric NO2 variation, obtained from the satellite measurements over St. Petersburg in 1996-2009. The standard deviation of monthly averaged is shown by grey line.
Then, to decide if the trend estimate is statistically significant a commonly used rule is applied, considering that a trend B is real with 95% confidence level if \B/aB| > 2
[Weatherhead et al., 1998].
To compute seasonal component ^t, each of the 12 months was averaged over the full data record for the years 19962009. Resulting time series of ^t obtained from the satellite measurements over St. Petersburg, along with the standard deviation for each month, is presented on Figure 4. According to that data, high NO2 columns are observed in winter and early spring - up to ~ 6 x 1015 molecules cm-2 in March; the minimum values are found in summer - about ~ 2 x 1015 molecules cm-2. Such typical seasonal cycle is caused by the longer lifetime of NOx during the cold period, increased local NOx emissions from the municipal heating system and frequent episodes of calm and temperature inversion, facilitating the accumulation of pollutants in the wintertime surface air.
Figure 5 shows a measured time series of monthly average tropospheric NO2 column over St. Petersburg and the fitted function (model approximation). The solid green line is the model fit and the remainder between model and measurements is denoted by the squares. Resulting trend estimate provides a linear growth of 4.3 ± 1.0% per year (growth is the percentage calculated with respect to the mean NO2 column, 3.3 x 1015 molecules cm-2). The fitted trend value is statistically significant (\B/aB \ > 2).
Figure 5. Time series of monthly average tropospheric NO2 column over St. Petersburg and the fitted function (model approximation). The solid green line is the model fit and the remainder between model and measurements is denoted by the squares. The linear component is shown by red line.
Figure 6. Inventory estimates of NOx emission from transport of St. Petersburg as given by EMEP (http://emep.int), EDGAR (http://www.mnp.nl/edgar) and JSC “SRI Atmosphere” (http://www.nii-atmosphere.ru).
Discussion, Conclusions and Further Work
Fossil fuel combustion is known to be responsible for more than 50% of the total production of NOx. According to the studies done at the Scientific Research Institute for Atmospheric Air Protection, JSC “SRI Atmosphere” (http://www.nii-atmosphere.ru) the share of traffic in the total emission of anthropogenic pollutants at St. Petersburg has increased from 56% in 1990 to 78% in 2005 and 92% in 2008. The number of vehicles officially registered in St. Petersburg has changed from ~ 1.0 million in 2000 to ~ 1.5 million in 2008 - with an average linear growth rate of ~ 6% per year. Inventory estimate, reported by the Committee for Nature Use, Environmental Protection and Ecological Safety of St. Petersburg (http://infoeco.ru) for NOx emission from transport was 115 GgN yr-1 in 2008. The calculations of this estimate, provided by JSC “SRI Atmosphere” show, that this value has dropped down manifold beginning from 1988, then started to grow in 1995 but dropped again in 1998, and now it is growing steadily since 2002 with a linear rate of ^ 4% per year (see Figure 6). Konovalov [2007] combined satellite measurements of tropospheric NO2 with a continental scale air quality model (CTM CHIMERE) over the European part of Russia in 1996-2005, to elaborate so-called “top-down” emission estimates, and also found a positive and statistically significant trend of NOx emission during summer at St. Petersburg (3.6 ± 1.6% per year). Thereby,
the trend estimate derived in the present study (^ 4% per year) is in agreement with an expected growth of anthropogenic emissions at St. Petersburg during last decade.
Looking at the time series of tropospheric NO2 over St. Petersburg (see Figure 5), one may notice some decline, starting from 2006. A similar feature was found recently by Sitnov [2009], who analyzed tropospheric NO2 measured by OMI over the region of Moscow in 2004-2009 and suggested that the observed growth has stopped in 2006, and now the tropospheric NO2 is decreasing. Indeed, municipal inventory surveys of St. Petersburg report the growth of annual surface NO2 concentration from 29 mkg m-3 in 2001 to 57 mkg m-3 in 2006, but then it dropped back to 36 mkg m-3 in 2008 [Golybev and Sorokin, 2009]. The duration of that decline is too short to find any reliable explanation, but it indicates the need of additional parameter (e.g. local meteorology) to be introduced into the trend analysis.
Existing emission inventories based on the so-called “bottom-up” estimates (e.g. compilation of individual sources statistics), are known to be inaccurate. For instance, two spatially gridded inventories - the European Monitoring and Evaluation Programme (EMEP, http://emep.int) and Emission Database for Global Atmospheric Research (EDGAR, http://www.mnp.nl/edgar) provide considerably different estimates of NOX emission from the road transport source over the region of St. Petersburg (see Figure 6). A number of studies have demonstrated already the perspectives to verify and improve available emission estimates with
the help of satellite data [e.g. Konovalov et al., 2008; Leue et al., 2001; Martin et al., 2003, 2006]. A further development of current work will be to combine the data of satellite measurements with the regional chemical-transport model and improve the estimate of anthropogenic NOx emission from the city of St. Petersburg.
Acknowledgment. This study was supported in part by the Ministry of Education and Science of the Russian Federation (project no. RNP.2.1.1.4166, RNP.2.2.1.1.3836).
References
Boersma, K. F., H. J. Eskes, E. J. Brinksma (2004), Error analysis for tropospheric NO2 retrieval from space, J. Geophys. Res., 109, D04311. doi:10.1029/2003JD003962 Boersma, K. F., H. J. Eskes, E. J. Brinksma (2007), Near-real time retrieval of tropospheric NO2 from OMI, Atmos. Chem. Phys., 7, 2103-2118.
Bovensmann, H., J. P. Burrows, M. Buchwitz, J. Frerick, S. Noel, V. V. Rozanov, K. V. Chance, A. H. P. Goede (1999), SCIAMACHY — mission objectives and measurement modes, J. Atmos. Sci, 56(2), 127—150. doi: 10.1175/1520-0469(1999)056<0127:SM0AMM>2.0.C0;2 Burrows, J. P., et al. (1999), The Global Ozone Monitoring Experiment (GOME): Mission concept and first scientific results, J. Atmos. Sci., 56, 151—175. doi: 10.1175/1520-
0469(1999)056<0151:TGOMEG>2.0.CO;2 Golybev, D. A., N. D. Sorokin (2009), Environmental pro-
tection, nature use and ecological safety of St. Petersburg in 2008, annual review, Sesame-Print, St. Petersburg.
Ionov, D. V., Y. M. Timofeev (2009), Regional space monitoring of nitrogen dioxide in the troposphere, Izv. Atm. Ocean. Phys., 45(4), 434—443. doi:10.1134/S0001433809040045 Konovalov, I. B. (2007), Regional differences in the decadal variations of atmospheric emissions of nitrogen oxides in the European part of Russia: results of inverse modeling based on satellite data, Dokl. Earth Sci., 417(9), 1424—1427.
doi:10.1134/S1028334X07090280 Konovalov, I. B., M. Beekmann, J. P. Burrows, A. Richter (2008), Satellite measurement based estimates of decadal changes in European nitrogen oxides emissions, Atmos. Chem. Phys., 8, 2623—2641.
Leue, C., M. Wenig, T. Wagner, O. Klimm, U. Platt, B. Jahne (2001), Quantitative analysis of NOX emissions from GOME satellite image sequences, J. Geophys. Res., 106, 5493—5505. doi:10.1029/2000JD900572
Levelt, P. F., E. Hilsenrath, G. W. Leppelmeier,
G. H. J. van den Oord, P. K. Bhartia, J. Tamminen, J. F. Haan, J. P. Veefkind (2006), Science objectives of the Ozone Monitoring Instrument, IEEE Trans. on Geosc. and Remote Sensing, 44(5), 1199—1208.
Martin, R. V., D. J. Jacob, K. Chance, T. Kurosu, P. I. Palmer, M. J. Evans (2003), Global inventory of nitrogen oxide emissions constrained by space-based observations of NO2 columns, J. Geophys. Res., 108, 4537. doi:10.1029/2003JD003453 Martin, R. V., et al. (2006), Evaluation of space-based constraints on global nitrogen oxide emissions with regional aircraft measurements over and downwind of eastern North America, J. Geophys. Res., 111, D15308. doi:10.1029/2005JD006680 Munro, R., M. Eisinger, C. Anderson, J. Callies, E. Corpaccioli, R. Lang, A. Lefebvre, Y. Livschitz, A. P. Albinana (2006), GOME 2 on Metop: From in-orbit verification to routine operations, 2006 EUMETSAT Meteorological Satellite Conference, Helsinki, Finland, 12-16 June, 2006, EUMETSAT, Helsinki, Finland.
Platt, U. (1994), Differential Optical Absorption Spectroscopy (DOAS), Air Monitoring by Spectroscopic Techniques, Chem. Anal., 127, edited by M. W. Sigrist, 27—76, Wiley-Interscience, Hoboken, N. J.
Poberovskii, A. V., A. V. Shashkin, D. V. Ionov,
Y. M. Timofeev (2007), Variations of NO2 content
in the region of St. Petersburg from the ground-based and satellite measurements of scattered sunlight, Izv. Atm. Ocean. Phys., 43(4), 112—120. doi:10.1134/S0001433807040123 Richter, A., J. P. Burrows, N. Hendrik, C. Granier, U. Niemeier (2005), Increase in tropospheric nitrogen dioxide over China observed from space, Nature, 437(7055), 129—132.
Sitnov, S. A. (2009), Analysis of spatial-temporal variability of tropospheric NO2 over Moscow megapolis using OMI spectrometer (Aura satellite) data, Dokl. Earth Sci., 429(9), 1511—1517. doi:10.1134/S1028334X09090219 Van der A, R. J., D. H. M. U. Peters, H. Eskes, K. F. Boersma, M. Van Roozendael, I. De Smedt, H. M. Kelder (2006), Detection of the trend and seasonal variation in tropospheric NO2 over China, J. Geophys. Res., 111, D12317. doi:10.1029/2005JD006594 Van der A, R. J., H. Eskes, K. F. Boersma, T. P. C. Van Noije, M. Van Roozendael, I. De Smedt, D. H. M. U. Peters,
H. M. Kelder, E. W. Meijer (2008), Trends, seasonal
variability and dominant NOx source derived from a ten year record of NO2 measured from space, J. Geophys. Res., 113, D04302. doi:10.1029/2007JD009021
Weatherhead, E. C., et al. (1998), Factors affecting the detection of trends: Statistical considerations and applications to environmental data, J. Geophys. Res., 103(D14), 17,149— 17,161. doi:10.1029/98JD00995
D. V. Ionov, Faculty of Physics, St.Petersburg State University, St. Petersburg, Russia ([email protected])
