Geoscience Reference
In-Depth Information
14
13
12
11
10
9
8
7
6
5
0
0.2
0.4
0.6
0.8
1
AMSU-A OI
Fig. 4.6
Observation Influence (
OI
) for AMSU-A observations and for October 2011 grouped by
channels
14
13
12
11
10
9
8
7
6
5
0
3
6
9
12
15
18
21
24
AMSU-A DFS %
Fig. 4.7
Degree of Freedom for Signal in percentage (
DFS
) for AMSU-A observations and for
October 2011 grouped by channels. The percentage is relative to the total
AMSU-A DFS
4.4.2
Geographical Map of OI
The geographical map of observation influence for SYNOP and DRIBU surface
pressure observations is shown in Fig.
4.8
. Each box indicates the observation influ-
ence per observation location averaged among all the October 2011 measurements.
Data points with influence greater than one are due to the approximation of the
computed diagonal elements of influence matrix (see
Cardinali et al. 2004
). Low-
influence data points have large background influence (see
4.14
and
4.15
), which is
the case in data-rich areas such as North America and Europe (observation influence
0:2
) (see also Sect.
4.3.3
). In data-sparse areas individual observations have larger
influence: in the Polar regions, where there are only few isolated observations, the
OI
is very high (theoretically
1
) and the background has very small influence on
the analysis.
In dynamically active areas (Fig.
4.8
: e.g. North Atlantic and North Pacific),
several fairly isolated observations have large influence on the analysis. This is
Search WWH ::
Custom Search