uplands showed similar challenges for farmers: Progressing soil erosion and degradation as a consequence of unsustainable farming methods, potentials of soil
conservation measures, but lack of adoption owed to prohibitive costs or farmers’
preference of short-term benefit, dependence of cash crop yields on massive
fertilizer inputs are the main trends, which are modified by the level of agricultural
intensification and market access a watershed has reached. Model projections
differed only with respect to maize yield levels in the farther future; the question,
in how far productivity of an unsustainable system can be maintained by fertilizer
inputs or when exactly it will collapse, was answered differently by LUCIA and
FALLOW on one side (where maize yields increased under farmers’ practice) and
the coupled LUCIA-MP-MAS model system (where only the soil conservation
treatment gave increasing yields). In this case, a comparison between MP-MAS
standalone and the coupled version on the same Chieng Khoi dataset might lead to
new insights.
The contributions presented here also show a cross-section of the potential
model adaptation and coupling options needed in order to fit a model to a specific
research question. Interest in and the technical possibilities of combining spatial
scales and temporal resolutions, empiric and mechanistic approaches, have steadily
increased over the last few years, which is reflected in the mushrooming number of
approaches put forward, such as Open Modelling Interface (Open MI) (Moore and
Tindall 2005), CORMAS (Becu et al. 2003), the PCRaster Python Framework
(Karssenberg et al. 2010; Schmitz et al. 2009) and many others.
Within the research of the Uplands Program, interdisciplinary modeling – as an
approach (rather than as a technical tool) that can link scientific domains, has
created added value in terms of:
• Discussing approaches, their interfaces and potential linkages, from the conceptual level to the technical level
• Understanding the perspectives of other disciplines and how these influence
mind models of a researcher’s scientific domain. This widens horizons and
sharpens perception of the limitations of individual approaches
• Developing new and joint strategies of data collection (joint planning from the
start; making use of datasets that were originally not intended for model use, e.g.,
validation) and overcoming the difficulties of data-miners, and
• Jointly interpreting coupled model runs. Here, crossover can lead to innovation
in developing the individual models and in conceptualizing the research.
As a continuation of the participatory approaches used by the Uplands Program,
ongoing projects are continuing the effort to bring the models to the people. The
main target groups for the WaNuLCAS, FALLOW and LUCIA models are students
and researchers at Thai and Vietnamese universities, and staff of the relevant
national land development agencies. As outputs of this work, teaching kits (software plus case study datasets), web-based modeling facilities and new model
modules (programmed on demand by the users) will be available in the near future.
10 Integrated Modeling of Agricultural Systems in Mountainous Areas
427
conservation measures, but lack of adoption owed to prohibitive costs or farmers’
preference of short-term benefit, dependence of cash crop yields on massive
fertilizer inputs are the main trends, which are modified by the level of agricultural
intensification and market access a watershed has reached. Model projections
differed only with respect to maize yield levels in the farther future; the question,
in how far productivity of an unsustainable system can be maintained by fertilizer
inputs or when exactly it will collapse, was answered differently by LUCIA and
FALLOW on one side (where maize yields increased under farmers’ practice) and
the coupled LUCIA-MP-MAS model system (where only the soil conservation
treatment gave increasing yields). In this case, a comparison between MP-MAS
standalone and the coupled version on the same Chieng Khoi dataset might lead to
new insights.
The contributions presented here also show a cross-section of the potential
model adaptation and coupling options needed in order to fit a model to a specific
research question. Interest in and the technical possibilities of combining spatial
scales and temporal resolutions, empiric and mechanistic approaches, have steadily
increased over the last few years, which is reflected in the mushrooming number of
approaches put forward, such as Open Modelling Interface (Open MI) (Moore and
Tindall 2005), CORMAS (Becu et al. 2003), the PCRaster Python Framework
(Karssenberg et al. 2010; Schmitz et al. 2009) and many others.
Within the research of the Uplands Program, interdisciplinary modeling – as an
approach (rather than as a technical tool) that can link scientific domains, has
created added value in terms of:
• Discussing approaches, their interfaces and potential linkages, from the conceptual level to the technical level
• Understanding the perspectives of other disciplines and how these influence
mind models of a researcher’s scientific domain. This widens horizons and
sharpens perception of the limitations of individual approaches
• Developing new and joint strategies of data collection (joint planning from the
start; making use of datasets that were originally not intended for model use, e.g.,
validation) and overcoming the difficulties of data-miners, and
• Jointly interpreting coupled model runs. Here, crossover can lead to innovation
in developing the individual models and in conceptualizing the research.
As a continuation of the participatory approaches used by the Uplands Program,
ongoing projects are continuing the effort to bring the models to the people. The
main target groups for the WaNuLCAS, FALLOW and LUCIA models are students
and researchers at Thai and Vietnamese universities, and staff of the relevant
national land development agencies. As outputs of this work, teaching kits (software plus case study datasets), web-based modeling facilities and new model
modules (programmed on demand by the users) will be available in the near future.
10 Integrated Modeling of Agricultural Systems in Mountainous Areas
427