MAX-DOAS measurements of HONO slant column densities during the MAD-CAT campaign: inter-comparison, sensitivity studies on spectral analysis settings, and error budget
In order to promote the development of the passive DOAS technique the Multi Axis DOAS – Comparison campaign for Aerosols and Trace gases (MAD-CAT) was held at the Max Planck Institute for Chemistry in Mainz, Germany, from June to October 2013. Here, we systematically compare the differential sla...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2017-10-01
|
| Series: | Atmospheric Measurement Techniques |
| Online Access: | https://www.atmos-meas-tech.net/10/3719/2017/amt-10-3719-2017.pdf |
| Summary: | In order to promote the development of the passive DOAS technique
the Multi Axis DOAS – Comparison campaign for Aerosols and Trace gases
(MAD-CAT) was held at the Max Planck Institute for Chemistry in Mainz,
Germany, from June to October 2013. Here, we systematically compare the
differential slant column densities (dSCDs) of nitrous acid (HONO) derived
from measurements of seven different instruments. We also compare the
tropospheric difference of SCDs (delta SCD) of HONO, namely the difference of
the SCDs for the non-zenith observations and the zenith observation of the
same elevation sequence. Different research groups analysed the spectra from
their own instruments using their individual fit software. All the fit errors
of HONO dSCDs from the instruments with cooled large-size detectors are
mostly in the range of 0.1 to 0.3 × 10<sup>15</sup> molecules cm<sup>−2</sup>
for an integration time of 1 min. The fit error for the mini MAX-DOAS is
around 0.7 × 10<sup>15</sup> molecules cm<sup>−2</sup>. Although the HONO
delta SCDs are normally smaller than
6 × 10<sup>15</sup> molecules cm<sup>−2</sup>, consistent time series of HONO
delta SCDs are retrieved from the measurements of different instruments. Both
fits with a sequential Fraunhofer reference spectrum (FRS) and a daily noon
FRS lead to similar consistency. Apart from the mini-MAX-DOAS, the systematic
absolute differences of HONO delta SCDs between the instruments are smaller
than 0.63 × 10<sup>15</sup> molecules cm<sup>−2</sup>. The correlation
coefficients are higher than 0.7 and the slopes of linear regressions deviate
from unity by less than 16 % for the elevation angle of 1°. The
correlations decrease with an increase in elevation angle. All the
participants also analysed synthetic spectra using the same baseline DOAS
settings to evaluate the systematic errors of HONO results from their
respective fit programs. In general the errors are smaller than
0.3 × 10<sup>15</sup> molecules cm<sup>−2</sup>, which is about half of the
systematic difference between the real measurements.<br><br>The differences of HONO delta SCDs retrieved in the selected three spectral
ranges 335–361, 335–373 and 335–390 nm are considerable (up to
0.57 × 10<sup>15</sup> molecules cm<sup>−2</sup>) for both real measurements
and synthetic spectra. We performed sensitivity studies to quantify the
dominant systematic error sources and to find a recommended DOAS setting in
the three spectral ranges. The results show that water vapour absorption,
temperature and wavelength dependence of O<sub>4</sub> absorption, temperature
dependence of Ring spectrum, and polynomial and intensity offset correction
all together dominate the systematic errors. We recommend a fit range of
335–373 nm for HONO retrievals. In such fit range the overall systematic
uncertainty is about 0.87 × 10<sup>15</sup> molecules cm<sup>−2</sup>, much
smaller than those in the other two ranges. The typical random uncertainty is
estimated to be about 0.16 × 10<sup>15</sup> molecules cm<sup>−2</sup>, which
is only 25 % of the total systematic uncertainty for most of the
instruments in the MAD-CAT campaign. In summary for most of the MAX-DOAS
instruments for elevation angle below 5°, half daytime measurements
(usually in the morning) of HONO delta SCD can be over the detection limit of
0.2 × 10<sup>15</sup> molecules cm<sup>−2</sup> with an uncertainty of
∼ 0.9 × 10<sup>15</sup> molecules cm<sup>−2</sup>. |
|---|---|
| ISSN: | 1867-1381 1867-8548 |