Geoscience Reference
In-Depth Information
forecast model is taken into account to ensure the observations are used in a
meteorologically consistent way. The idea behind 4D-Var data assimilation is
shown in Figure 1.4. For a single parameter
x
the observations are compared with
the short-range forecast from a previous analysis over a 12-hour period. The model
state at the initial time is then modified to achieve a statistically good compromise,
x
a
, between the fit,
J
b
, to the previous forecast,
x
b
, and the fit
J
0
to all observations
within the assimilation window.
J
b
and
J
0
are referred to as cost functions [AND
08]. This 4D-Var approach replaced earlier approaches that were based on the
optimum interpolation method.
Figure 1.4.
The idea of 4D-Var data assimilation technique, see the text
for a detailed explanation (from [AND 08])
Climate re-analyses are also subject to a detailed validation against independent
observations. A detailed description of climate reanalysis goes far beyond the scope
of this chapter; for more details the interested reader should refer to the literature
(e.g. [UPP 04]).
1.1.7.
Climate data providers outside NMHs
Climate data are not only provided by meteorological and hydrological services
but also by other data providers, in particular, universities. The majority of these
data centers provide surface climate data that are either station data or gridded data.
Usually, these data providers use data from NMHs and improve their data quality or
spatial coverage. From the GIS perspective, standardization of these datasets is still
weak and both data formats and metadata are quite heterogenous. Consequently, the
import of the data to GIS needs some data preparation. However, OGC
standardization of the NetCDF format is expected to overcome this shortcoming in
the near future.
A major climate data provider is the Climate Research Unit (CRU) from the
University of East Anglia (UK). In particular, CRU provides long-term climate data
with global coverage, which is an important base for global climate monitoring.
