NO2 VCD was measured using the ground-based Pandora from March
2012 to May 2014 and tropospheric NO2 VCD was measured using the
ground-based MAX-DOAS from August 2013 to May 2014 in Seoul,
Korea. The NO2 measurements using Pandora and MAX-DOAS were
compared with those obtained by satellite remote sensing from Ozone
Monitoring Instrument (OMI) where the characteristics were intercompared
and analyzed as a function of measurement geometry, cloud amount and
aerosol loading.
The difference between NO2 VCD from Pandora and OMI measurements
were larger when cloud amount and Aerosol Optical Depth (AOD) were
higher. The correlation coefficient between NO2 VCDs from Pandora and
OMI was 0.58 for the entire measurement period, whereas the correlation
coefficient between the two was increased to 0.75 when the cloud amount- 94 -
and AOD were low (cloud amount<3, AOD<0.4). The correlation coefficient
between tropospheric NO2 VCDs from MAX-DOAS(fitting error < 50 %)
and OMI was 0.47 for the entire measurement period, whereas the
correlation coefficient between the two was 0.55 when the cloud amount is
zero and the correlation coefficient between the two was 0.26 when the
cloud amount is more than zero. The difference between tropospheric NO2
VCD from MAX-DOAS and OMI was nearly zero when AOD interval is
between 0 and 0.25. As AOD becomes larger, the biases became higher
which also might be associated with the light path enhancement and
shielding effect of the aerosols.
It can be inferred that the morphology and condition for clouds and
aerosols increase the uncertainty of the retrieved OMI NO2 VCD and OMI
tropospheric NO2 VCD, since the light path enhancement and shielding
effect of the cloud and the aerosols may result in high uncertainty for the
contaminated radiance fraction thus the air mass factor.