|
Remaining problems with the correction for the interference
The correction for the ozone interference described in the
main background correction section
is working quite well, much better than the previous correction
and the number of problem cases is reduced dramatically.
But the problems are not complete over yet, as described below. For the
moment we do not know how we could set up a better correction of the
interference problem and thus avoid problems like those mentioned here.
Problems at high SZA
For SZA well above 80 degrees there can still be a problem, as the SO2 vertical column observed on the North of Scandinavia (top graphique):
the nadir state in the middle near Scandinavia shows a few ground pixels with
vertical column values of more than 10 DU. There are no reports of
volcanic activity on Iceland around that time, nor is there any
anthropogenic source nearby. Clearly, the seemingly high SO2 values in this
graph are artifacts.
For this reason we have decided to use an "adaptative" SZA according to the day of year and to the hemisphere (North or South).
Our study of an optimal SZA gives the falling empiric values of the limitation of SZA for the North hemisphere (SZAnorth) and the South hemisphere SZAsouth:
Day of year |
SZAnorth |
SZAsouth |
21/03 |
75� |
80� |
21/06 |
70� |
80� |
21/09 |
75� |
80� |
21/12 |
80� |
70� |
Day of year |
SZAnorth |
SZAsouth |
21/03 |
75� |
80� |
21/06 |
70� |
75� |
21/09 |
75� |
80� |
21/12 |
75� |
70� |
The SO2 peak values shown in the top image (SZA=80�) satisfy the
criteria for exceptional SO2 concentration
and would thus trigger a notification, but clearly this is a "false notification". For this reason we use an empirical "adaptative" seasonal SZA to avoid such false notification.
Apparent band of higher SO2 at mid-latitudes
The monthly average SO2 slant column plots show in most cases a kind of band
of higher and noisier SO2 signal between about 30 and 60 degrees North.
See for example the plot for March 2007 shown here, which shows a
particularly clear band. A few months show no or a very weak band at the
same latitude. A similar band exists on the Southern hemisphere, but there
it is much less pronounced. All in all the SO2 slant column lies for low and
mid latitudes mostly between +0.3 and -0.3 DU, which can be
considered as the noise level on the zero-SO2 background.
Note from the plot shown here that the South Atlantic
Anomaly is clearly visible as enhanced noise in the SO2 slant column.
It is unclear what the origin (reason) for this band is. Most likely it has
to do with not entirely adaquate correction of the ozone interference. But
it is strange that the band lies at slightly lower latitudes than where the
largest variation of ozone along latitudes is observed, as can be seen by
comparing the monthly average total ozone column with the SO2 plot above.
|
|