Environmental Engineering Reference
In-Depth Information
Looking into the Future
The different operational stage of the two modelling systems, and the fact that
SEAMLESS-IF comprises SEAMCAP as a specific version of CAPRI, renders a
comprehensive evaluation difficult. Over the last decade, CAPRI was able to provide
analytical support to different EU policy reform steps based on quantitative
impact assessment, while at the same time it was expanded, revised and improved.
However, its limited modularity has led to a rather large and complex system, so
that only a few experienced modellers are able to exploit the full potential of the
system or to successfully change components.
A key success factor for CAPRI, as for many other model networks, has been
human capital in the sense that several key persons continued working with the
system over years. Important for building up the network has been a sequence of
3-year EU framework projects (CAPRI, CAP-STRAT, CAPRI-Dynaspat) centred
on CAPRI development and application, co-ordinated by the core development team.
The network evolved almost naturally during the gradual build up and expansion of
the system over a longer period as the projects around CAPRI allowed training
graduate students in using, modifying and expanding the system. The increasing
demand for agricultural policy impact assessment allowed several of the graduate
students trained in the EU framework projects to continue working as post-docs
with CAPRI at different institutions, being the basis for the successful application
of CAPRI in policy impact assessment studies. Such studies with their typically
tight deadlines require more senior staff, well experienced in applying the system,
and are thus less suited to provide training for graduate students. New teams and
individuals have been attracted to CAPRI by yearly training sessions. The sessions
discussed the general concept of CAPRI, communicated methodological and
technical changes and conducted hands-on training on application. Building up the
network was certainly supported by the growing market for this type of analysis,
which kept competition in the network low. At the same time, the increased demand
for integrative impact assessment and application of model families made it
attractive for research institutions to take CAPRI into their portfolio of tools.
From the technical side, the CAPRI model chain uses GAMS (General Algebraic
Modeling System) as the only modelling software for all major components,
including those dealing with database updates and model parameterization, which
certainly reduces maintenance and development costs. However, this has led to a
less than perfect implementation in technical terms, for example regarding model
structure or documentation of the code. Moreover, GAMS does not support well the
development of re-usable components, since basically any object has a global scope
and the passing of arguments, as usual for other programming languages, is not
supported, hindering encapsulation of components.
From the human resource perspective, the discussion on how to come to a fair
distribution of maintenance costs of the system is far from finalized. These costs
are discouraging when applying for typical research projects which put a focus on
further methodological improvements. Moreover, short-time policy impact studies
Search WWH ::




Custom Search