Ground-based multi-axis differential optical absorption spectroscopy
(MAX-DOAS) measurements of aerosols and tropospheric nitrogen dioxide
(
Nitrogen oxides (
For more than 2 decades, satellite nadir measurements of atmospheric
backscattered sunlight in the UV–vis range have provided daily global
tropospheric column measurements of atmospheric
For about 2 decades, the multi-axis differential optical absorption
spectroscopy (MAX-DOAS) technique (Hönninger et al.,
2004) has been widely used for retrieving the vertical and horizontal
distribution of trace gases and aerosols in the troposphere (e.g., Sinreich
et al., 2005; Ortega et al., 2015). MAX-DOAS instruments perform
observations of scattered sunlight in the visible and ultraviolet (UV)
spectral ranges at multiple elevation angles towards the horizon, leading to
increased sensitivity to absorbers situated close to the surface
(Hönninger et al., 2004).
The motivation of the present work is to investigate the use of
multi-azimuthal MAX-DOAS measurements in Uccle (Belgium) to validate the
TROPOMI tropospheric
The paper is organized as follows: in Sect. 2, the measurement site and
the MAX-DOAS experimental setup are described, followed by the DOAS
analysis and the retrieval methodologies applied to the MAX-DOAS
observations and their validation. Section 3 focuses on the
tropospheric
The MAX-DOAS instrument operated at BIRA-IASB (Koninklijk Belgisch Instituut
voor Ruimte-Aeronomie – Institut royal d'Aeronomie Spatiale de Belgique) is
an improved version of the system described in Clémer et al. (2010). Developed to contribute to the CINDI-2 intercomparison campaign
in Cabauw in September 2016 (Kreher et al., 2020), it was
subsequently installed on the rooftop of the Royal Meteorological Institute
(RMI) in Uccle (50.8
The MAX-DOAS dual-scan instrument is composed of three main parts: (1) an optical head mounted on a sun tracker, (2) a thermoregulated box with two spectrometers (UV and visible) and (3) the acquisition unit. Optical fibers connect the optical head with the two spectrometers.
The optical head is equipped with a filter wheel that allows for switching
between skylight and direct sun measurements. The UV optical fiber consists
of a
Installed indoors, the thermoregulated box is equipped with visible and UV
grating spectrometers covering the wavelength ranges of 405 to 540 nm and
300 to 390 nm, respectively. The UV spectrometer is from Newport (model
74086) with a spectral resolution of 0.4 nm. To block the visible light and
to reduce the stray light in the UV wavelength region, a band-pass filter
(U-340 Hoya) is used. The output of the UV spectrometer is connected to a
back-illuminated UV-enhanced charge-coupled device (CCD) detector system
(Princeton Instrument Pixis 2K). The visible spectrometer from Horiba (model
Micro HR) has a spectral resolution of 0.7 nm and is also mounted on a
back-illuminated CCD system (Princeton Instrument Pixis 100). Both CCD
detectors are cooled at 223 K (using multistage Peltier system). The
overall spectrometric unit is thermally stabilized to better than 1
To control the data acquisition, two computers are used. One records spectra coming from the visible spectrometer and controls the sun tracker, while the second, synchronized with the first one, records the spectra from the UV spectrometer.
In order to measure in dual-scan (elevation
MAX-DOAS experimental setup.
The dual-scan experimental setup of the BIRA-IASB MAX-DOAS instrument (overlaid onto the OpenStreetMap (OSM) standard layer) (left panel). The lines are color-coded according to the three different experimental setups (Table 1), with a line length of 20 km each. The colored dots show the different types of in situ stations around the MAX-DOAS instrument (see Sect. 4.2). The right panel shows the location of the MAX-DOAS, AERONET and ceilometer instruments at the measurement site. ©OpenStreetMap contributors 2019. Distributed under a Creative Commons BY-SA License.
The spectra measured in both sub-modes are analyzed using the QDOAS spectral fitting software developed at BIRA-IASB (Fayt et al., 2011) for the retrieval of atmospheric trace gas abundances in the UV, visible and near-infrared spectral ranges. The DOAS technique consists of a separation between narrow absorption features characteristic of molecular species and a spectral background resulting mainly from Mie and Rayleigh scattering as well as instrumental effects (Platt and Stutz, 2008). Its primary product is the differential slant column density (dSCD), which represents the light-path-integrated trace gas concentration in a measured spectrum relative to the amount of the same absorber in a reference spectrum. In the present case, daily noon zenith spectra are used as a reference.
DOAS settings for
Same as Table 2 for the UV spectral range.
The aerosol extinction coefficient and
The MMF algorithm uses the optimal estimation method (OEM; Rodgers, 2000) formalism, the VLIDORT (Spurr, 2006) version 2.7 radiative transfer model (RTM) as a forward model and a Levenberg–Marquardt (LM) iteration scheme.
MMF works in linear measurement space and logarithmic retrieval space.
Further details about this algorithm can be found in
Friedrich et al. (2019). MMF was one of the retrieval codes
used during the CINDI-2 campaign (Tirpitz et al., 2020), and it also
participated in the round-robin comparison of profiling algorithms as part
of the fiducial reference measurements for the Ground-Based DOAS Air-Quality
Observations (FRM
An important parameter in the OEM approach is the a priori profile. In the
present study, exponentially decreasing a priori profiles are used for both
aerosols and
The pressure and temperature profiles are prescribed using 20-year
monthly averaged data extracted from the European Centre for Medium-Range
Weather Forecasts (ECMWF) ERA-Interim reanalysis (see Beirle et al., 2019)
for the location of Uccle. The retrieval altitude grid consists of
20 layers of 200 m thickness between the surface and 4 km of altitude. The
surface albedo is set to 0.06 and the aerosol optical properties, such as the
single-scattering albedo and the asymmetry parameter, are taken from
colocated AERONET measurements. Regarding the retrieval wavelengths,
aerosol extinction vertical profiles are retrieved at 360 and 477 nm and the
Each retrieval is quality-checked based on three different criteria. First,
the degrees of freedom (DOFs) should be larger than 2. This ensures that the
profile information comes mostly from the measurements and not from the
a priori profile. Second, the relative root mean square error (RMSE) of the
difference between measured and calculated differential slant column
densities with respect to the zenith spectrum of each scan should be
smaller than 15 %. This excludes local minima. Third, the AODs should be
smaller than 5 because of the high profile uncertainties on the trace gas
retrieval in such conditions (Hendrick et al., 2014). The
abovementioned criteria are applied to the
The presence of aerosols and clouds in the atmosphere can strongly affect
the MAX-DOAS trace gas retrieval (Frieß
et al., 2006; Gielen et al., 2014; Wagner et al., 2004, 2014). In order to
exclude MAX-DOAS measurements strongly influenced by the presence of clouds,
a cloud-filtering approach is applied using a colocated thermal infrared
pyrometer. The pyrometer determines the total cloud-cover fraction based on
the temperature data over a field of view of 6
The uncertainties of the vertical profiles retrieved by MMF include three
types of errors (Rodgers, 2000): (1) the smoothing error, which
represents the difference between the retrieved and the true profile due to
the vertical smoothing, (2) the noise error, which represents the
uncertainty arising from the dSCD measurement, and (3) the error coming
from the forward model. In Table 4, an overview of the main error sources for
the
Error budget overview of the MMF-retrieved
We also consider the systematic uncertainty on the
The dual-scan MAX-DOAS retrieval strategy refers to the near-surface
The parameterization approach used in this study is an adaptation of the one
introduced in Sinreich et al. (2013). It consists of a
conversion of
If sufficient aerosols are present in the lower troposphere (boundary
layer), the measured dSCD at two low elevation angles (in the present study,
1 and 2
The unknown variable in Eq. (
For each measurement, the differential effective path lengths can be
calculated as a ratio of the measured
The introduction of a unitless correction factor (
In this section, the abovementioned parameterization method is applied to our dual-scan MAX-DOAS measurements in Uccle.
To estimate MLH, we use the
The dAMF depends on the geometry (SZA, RSAA and elevation angle) as well as
the aerosol and trace gas concentration profiles. For its calculation, we
used VLIDORT (Spurr, 2006) version 2.7. The dAMF of
Panels
The correction factor
To estimate
For the analysis, only measurements at SZA smaller than 80
Furthermore, when the
To estimate uncertainties on the retrieved
The second important error source is related to the correction factors,
which depend on the air mass factor and MLH
Error budget overview of the parameterized
Qualitative information about the horizontal distribution of
During one measurement, four useful pieces of information can be used in
order to estimate the distance of the VMR VMR VMR
This information is further exploited in Sect. 4.1, where the seasonal
variation of the dual-scan MAX-DOAS measurements is presented.
Comparison of monthly averaged MLH diurnal variations as
estimated by the BIRA-IASB MAX-DOAS measurements and the colocated
ceilometer. The error bars for both datasets represent the standard
deviation (
Scatter plot of the monthly average MLH diurnal variation values of the MAX-DOAS and the colocated ceilometer. The color bar separates the data by season.
Visible range: comparison between
Same as Fig. 5 for the UV channel.
TROPOMI
To validate the dual-scan parameterization method used in this study, two
different approaches are adopted. First, the MLH, which is used in the
calculation of the correction factors, is compared with MLH measurements
using a colocated ceilometer. Second, the
To validate the MLH estimations, we use a colocated Vaisala CL51 ALC ceilometer operated by RMI. With this instrument, the MLH is retrieved according to an algorithm based on the direct analysis of the backscatter gradient and variance (Haij at al., 2007; Haeffelin et al., 2016; Menut et al., 1999).
Figure 3 displays the diurnal variation of monthly averaged MLH values
derived from the ceilometer and MAX-DOAS data during one full year, from March 2018 until March 2019. As can be seen, the MAX-DOAS data capture the
diurnal variation of the MLH measured by the ceilometer well. The corresponding
scatter plot is presented in Fig. 4. Both datasets are highly correlated
(
Tropospheric
The second approach to validate the parameterization technique consists of
comparing the retrieved
Panels
Flying onboard the S5P satellite platform, TROPOMI is a
passive grating imaging spectrometer covering the UV–visible (270–500 nm),
near-infrared (710–770 nm) and shortwave infrared (2314–2382 nm)
spectral ranges (Veefkind et al., 2011). TROPOMI measures the
solar backscattered earthshine radiance in a push-broom configuration. With
a full swath width as wide as 2600 km, TROPOMI provides daily global
coverage with a true-nadir pixel size of
Developed at KNMI (van Geffen et al., 2018), the tropospheric
The present study is based on RPRO and OFFL datasets of the TROPOMI L2
tropospheric
Box-and-whisker plots of MAX-DOAS horizontal effective light paths
(dL
In Fig. 9, the seasonal variation of the MAX-DOAS near-surface
Seasonally averaged near-surface
The near-surface
As already noted, the retrieved
Box-and-whisker plots representing, for each season, the
tropospheric
The in situ telemetric air quality network (Bruxelles
Environnement/Leefmilieu Brussel) of the Brussels region is used to
verify the retrieved near-surface
The present work uses hourly
In some directions, several in situ stations are located in proximity to the
MAX-DOAS line of sight so that MAX-DOAS
Based on the different categories of in situ stations, three groups were
created, similarly to the study of Kramer et al. (2008): (1) urban
background, (2) urban and (3) traffic. The in situ data were interpolated on
the MAX-DOAS time grid and compared with the retrieved MAX-DOAS near-surface
VMR. For the comparison, the in situ dataset was averaged in bins of 2.5 ppb length each. In Fig. 11, the results of this comparison show that the
MAX-DOAS near-surface
Scatter plots of binned MAX-DOAS and in situ
The worst agreement is found at traffic and urban stations (
To validate the TROPOMI tropospheric A first comparison is performed by selecting only MAX-DOAS data in the main
azimuthal direction (35.5 To improve the spatial coincidence between MAX-DOAS and TROPOMI
observations, a second comparison is performed by using the dual-scan
MAX-DOAS observations: measurements in every MAX-DOAS azimuthal direction
are compared with a weighted average of TROPOMI columns as measured in
coincident pixels with the weighting being determined by the MAX-DOAS
horizontal segment crossing every pixel. The impact of possible systematic uncertainties on the satellite retrieval
(in particular the a priori profile shape) is investigated.
In order to increase the number of colocation pairs, we compare both UV and
visible MAX-DOAS measurements with TROPOMI. It should be emphasized that UV and
visible MAX-DOAS VCDs correspond to different dLeff (
In this first approach, the MAX-DOAS tropospheric
We compare TROPOMI daily measurements with MAX-DOAS
Time series of the tropospheric
Figure 13 presents scatter plots of TROPOMI tropospheric
Seasonal scatter plots between the tropospheric
These results indicate that the discrepancy between TROPOMI and MAX-DOAS
measurements is significant during all seasons and particularly marked
during winter and autumn. A possible explanation could be differences
in the air masses probed by the two instruments. The use of only one
satellite pixel, even if its direction with respect to the MAX-DOAS line of
sight is taken into account, is not necessarily the most appropriate
comparison method. One azimuthal MAX-DOAS measurement samples air masses
along a light path of several kilometers in a fixed direction, which
corresponds to more than one TROPOMI pixel, as outlined in Fig. 7. One
expects this horizontal sampling effect to be more marked in winter and
fall given the larger
In a second step, we compare TROPOMI tropospheric
Results displayed in Fig. 14 show that the agreement between TROPOMI and
MAX-DOAS datasets is significantly improved, especially in terms of
correlation (
Seasonal scatter plots between the tropospheric
To identify the origin of the persisting underestimation of TROPOMI
Boersma et al. (2004) presented a thorough analysis of satellite
tropospheric
Clouds can have a major impact on tropospheric
Like clouds, aerosols can affect the accuracy of tropospheric
Surface albedo is another parameter having a significant influence on
satellite tropospheric
However, we should keep in mind that the surface albedo values used in the
TROPOMI
The TROPOMI
A way to test how uncertainties on the a priori profile influence the
TROPOMI
Mean TROPOMI averaging kernels (blue line), median
MAX-DOAS
Using the formula described in Appendix A and daily median MAX-DOAS
concentration profiles derived in the main azimuthal direction using the MMF
algorithm, a modified version of the TROPOMI tropospheric
Example of median daily MAX-DOAS visible
Figures 15 and 16 present a priori
Figure 17 presents validation results corresponding to the recalculated
TROPOMI
Seasonal scatter plots between the tropospheric
In conclusion, the change of the a priori profile in the TROPOMI retrieval has a significant impact on the agreement between the satellite and MAX-DOAS measurements, leading to a satisfying closure of the validation study. Although based on a different approach, these results are in agreement with the recent studies of Ialongo et al. (2020) and Judd et al. (2019).
A total of 1 year of S5P/TROPOMI tropospheric
The dual-scan parameterized
Summary table of the regression analysis parameters derived by the three validation exercises.
The seasonal variability of the
In a second step, MAX-DOAS data were used to validate TROPOMI tropospheric
Further, a detailed investigation of the main ancillary parameters used for
the AMF calculation in the TROPOMI tropospheric
In conclusion, our study shows that dual-scan MAX-DOAS measurements
conducted in an urban area offer the possibility (1) to better characterize
the spatial variability of short-lived pollutants like
We start from the general formula used to derive the
For optically thin conditions valid in the blue spectral range in which
In addition, the vertical sensitivity of the
When comparing satellite and ground-based measurements (here from a MAX-DOAS
instrument), it is a good practice to smooth the ground-based reference
profile using the satellite AK (see, e.g., Eskes and Boersma, 2003):
An alternative approach is to recalculate the satellite VCD using the
MAX-DOAS profile as a priori in the satellite retrieval. Only the AMF is
concerned and, similarly to Eq. (A2), we can write
The datasets generated and analyzed in the present work are available from the corresponding author on request.
ED undertook the development and validation of the dual-scan MAX-DOAS
retrieval strategy in Uccle, exploited the MAX-DOAS retrievals during
1 year, performed the validation of TROPOMI tropospheric
The authors declare that they have no conflict of interest.
This article is part of the special issue “TROPOMI on Sentinel-5 Precursor: first year in operation (AMT/ACP inter-journal SI)”. It is not associated with a conference.
We gratefully acknowledge the Belgian Federal Science Policy Office (BELSPO). The authors would like to thank the AERONET team for providing valuable data.
This research has been supported by the Belgian Science Policy Office BELSPO (Supplementary Researcher, grant no. 60.11.41.30.51).
This paper was edited by Lok Lamsal and reviewed by two anonymous referees.