Environmental Engineering Reference
In-Depth Information
In conclusion, the results indicate that data assimilation can provide significant
improvement at sites with strong contribution by a local source. Therefore, further
improvement could be expected through utilisation of observations covering also
major remote sources influencing the area of interest.
x10
-7
x10
-7
4.0
3.5
4.0
3.5
3.0
3.0
2.5
2.0
1.5
1.0
2.5
2.0
1.5
1.0
0.5
0.5
Fig. 2.
RMS forecast errors for the reference (crosses) and assimilation runs (boxes). Left: mean
RMS in the control set. Right: median (solid line), 13th and 86th percentiles (dashed lines)
References
H. Elbern, A. Strunk, H. Schmidt, O. Talagrand, Atmospheric Chemistry and Physics
7
(14), 3749
(2007)
M. Sofiev, P. Siljamo, I. Valkama,M. Ilvonen, J. Kukkonen, Atmospheric Environment
40
, 674 (2006)
M. Galperin, in Large Scale Computations in Air Pollution Modelling, ed. By Z. Zlatev, J. Dongarra,
I. Dimov, J. Brandt, P.J. Builtjes (Kluwer Academic Publishers, 1999), pp. 161-172
M. Sofiev, J. Geophys. Res.
107
(2002)
M. Sofiev, Atmospheric Environment
34
, 2481 (2000)
A. Sandu, D.N. Daescu, G.R. Carmichael, T. Chai, J. Comput. Phys.
204
(1), 222 (2004). DOI
http://dx.doi.org/10.1016/j.jcp.2004.10.011
4. Questions and Answers
M. Schaap:
Do you know what is the impact of assimilation of SO
2
on the
performance for SO
4
?
Answer:
Due to the limited availability of SO
4
measurements we have so far been
unable to analyse the effect of assimilation on SO
4
.
P. Builtjes:
Why do you keep the emission update constant over a day, and not on
an hourly basis?
Answer:
The 24 h assimilation window provides a natural time scale for the emission
adjustments. Furthermore, trying to estimate the emission corrections in one
hour steps would most probably lead into a severely underdetermined problem.
