Upcoming and recent developments
Data and baseline
The 2007 base period (2006-2008 average) is in preparation, the curent base period is 2004 and the baseline to 2020. Triggered by a change in the statistical system at FAO and the need to integrate data on bio-fuels, the way the global data base and baseline are sourced are currently re-defined. National time series are available until 2010 as part of the baseline generation which thus integrates information about rather recent developments in agricultural markets.
Recently, a pilot project has started to integrate irrigation water demand and supply into CAPRI.
CAPRI is currently extended to allow for the analysis of the discussed of the so-called "greening measures" of the CAP. Prelimininary scenarios on leaked information had been run and where e.g. been presented during the September 2011 training session in Braunschweig.
Different projects link CAPRI to GTAP. Two different versions are successfully applied. Teams at LEI and ILR have developped a sequential calibration link in the context of the SEAMLESS project, where GTAP's agricultural sectors are highly aggregated. That solution mimics the iterative link between supply and market modules in CAPRI. A second option has been explored by Tom Hertel at Purdue and Wolfgang Britz at ILR where the simulation behaviour of CAPRI's arable crop sector is mapped into land substitution elasticities of GTAP new land use module.The approach is discussed in a paper dealing with indirect land use impacts of European biofuel mandates. That approach is currently being refined and expanded to domestic agricultural policy shocks . As an outcome of these activities, an organized session on CAPRI-GTAP was launched at the GTAP conference in Bangkok in summer 2010.
CAPRI-RD will also develop regional CGEs at the NUTS2 level to be linked with the CAPRI supply models to model impacts of rural development policies on agriculture as well as on upstream and downstream sectors. The technical link is working, and the generation of the regional SAMs for all European NUTS 2 regions is progressing as planned.
In a project for DEFRA, the English Agricultural Ministry, impacts on farm income and land use from removing Pillar I support are analyzed. To do so, a land supply module has been integratd into CAPRI which allows for substitution between non-agricultural and agricultural land as well as between arabale and permanent grassland.
Major pillar II measures (LFA payments, N2K payments, agri-environmental measures) are the first time integrated in CAPRI in a so far operational, but preliminary fashion.
During 2009 and beyond, some structural changes are undertaken in CAPRI's code to increase modularity and transparency. Equally, a good practise for GAMS coding in CAPRI is under development, and the GAMS code adjusted to it. These changes are part of quality management work package in the FP7 project CAPRI-RD, and had been discussed for quite a while in the network.
Triggered by the tremendous increase in model size by the integration of the farm types, the supply models are now solved in parallel in several GAMS processes. Equally, a larger part of the post-model processing is now run in parallel. As a consequence, model runs could be sped up considerably. Equally, systematic decomposistion of key results of the system had been integrated in the post model processing.
In order to allow for light installations, a GDX file comprising all start values for standard simulations has been introduced. Consequently, large GDX files comprises data base and baseline results are not longer be required for a simulation run.
JAVA code allowing the generation of HTML based documentation of the GAMS code has been integrated in the GUI. The HTML based documentation will be linked with the methodological documentation. That step is deemed necessary to increase transparency in the system, and decrease the costs of contribution to the code. The HTML pages comprise also information about the relation between the local copy and the master version on the server. The first steps are described in a technical document.
A new version of the Graphical User Interface has been realesared. It applies a new conceptual approach to handle CAPRI tasks, integrates a batch execution facility, the HTML based GAMS code documentation, improved access to meta data and an online help system. It also now comprises a scenario editor and in-built SVN functionalities, statitistics such as outlier detection, integrate transparently the WEKA machine learning package and allows converting a shapefile into a geometry for use in the CAPRI GUI. Equally, an expression evaluator had been integrated in the exploitation tools so that simple arithmetic operations can been performed directly while viewing, reducing the storage and procesing needs post-model.
The next training session is scheduled for September 9-13th 2013, in Oslo, Norway. It will build on the successful "learning by doing" format of the autumn 2012 session in Piran, Sloveniw, the 2011 session in Braunschweig, Germany and the 2010 session in Cirencester, UK: small groups will analyze given scenarios from different perspectives (such as trade and market impacts, impacts on the farming sector, environmental impacts, welfare and CAP budget) based on the GUI, and report their findings back to all particitipants. The format allows newcomers to interact with trainers and more advanced users, while giving time to the more experienced network members to discuss methodology or to try new approaches to analysis. The training session will especially focus on spatial analysis.
The weeks before of the training session is also the time where all features of the system are intensively tested.