=Paper=
{{Paper
|id=Vol-1152/paper11
|storemode=property
|title=Experience from the Use of the Interactive Model- and GIS-based Information and Decision Support System LandCaRe-DSS for the Development of Economic Effective Application Strategies of Agriculture to Climate Change
|pdfUrl=https://ceur-ws.org/Vol-1152/paper11.pdf
|volume=Vol-1152
|dblpUrl=https://dblp.org/rec/conf/haicta/WenkelMBWK11
}}
==Experience from the Use of the Interactive Model- and GIS-based Information and Decision Support System LandCaRe-DSS for the Development of Economic Effective Application Strategies of Agriculture to Climate Change==
https://ceur-ws.org/Vol-1152/paper11.pdf
Experience from the Use of the Interactive Model- and
GIS-based Information and Decision Support System
LandCaRe-DSS for the Development of Economic
Effective Application Strategies of Agriculture to
Climate Change
Wenkel, K.-O.1; Mirschel, W.1; Berg, M.1; Wieland, R.1; Köstner, B.2
1
Centre for Agricultural Landscape Research (ZALF), Institute of Landscape Systems
Analysis, Eberswalder Straße 84, D-15374 Müncheberg, Germany
2
Technical University of Dresden, Institute of Hydrology and Meteorology, Chair of
Meteorology, Pienner Str. 23, 01737 Tharandt, Germany
Abstract The expected change of climate will influence agriculture in a
multifactorial manner. Up to now there are few suitable tools and methods for
agricultural farms and other decision makers to give helpful answers on
questions like: How will climate change influence regional agriculture and
ecosystems and what will be possible adaptation strategies for agriculture. The
newly developed information and decision support system, called LandCaRe-
DSS, closes this gap. It is designed as a user friendly, interactive, model-based,
and spatial-oriented information and decision support system, can be used on
different spatial scales, and supports long-term and ensemble simulations on a
high spatial resolution. The LandCaRe-DSS system assists strategic planning
in agriculture and sustainable development of rural areas (regions) and
provides answers concerning the effects and costs of possible adaptation
measures. Being still a prototype, LandCaRe-DSS is parameterized and
validated for two German regions at present. The impact models used in
LandCaRe-DSS, the experience of collaboration with stakeholders, and the
possibilities and limitations of the LandCare-DSS are described just like
practical experience gathered during the system development from the
modelling and implementation point of view. First results of scenario
simulations based on different climate regionalization methods are presented
and discussed.
1 Introduction
Agriculture is a fundamental human activity, supporting the livelihood of everyone
on this planet. Of the nearly 14 billion hectares of ice-free land on Earth, about 10%
are used for crop cultivation, while an additional 25% of land is used for pasture.
Over 2 billion tons of grains are produced yearly for food and feed, providing
_________________________________
Copyright ©by the paper’s authors. Copying permitted only for private and academic purposes.
In: M. Salampasis, A. Matopoulos (eds.): Proceedings of the International Conference on Information
and Communication Technologies
for Sustainable Agri-production and Environment (HAICTA 2011), Skiathos, 8-11 September, 2011.
129
roughly two-thirds of total direct and indirect protein intake; a mere 10% of this total,
or 200 million tons, is traded internationally (Tubiello et al, 2007).
Perhaps the most important challenge that agriculture will face in coming decades
is represented by the need to feed increasing numbers of people while conserving soil
and water resources. Existing projections indicate that future population and
economic growth will require a doubling of current food production, including an
increase from 2 up to 4 billion tons of grains annually, without significantly
increasing current arable land.
At the same time there is a significant concern about the impacts of climate
change and its variability on agricultural production worldwide. Current research
confirms that many crops would respond positively to elevated CO2 but the
associated impacts of high temperatures, altered patterns of precipitation and possibly
increased frequency of extreme events such as drought and floods will probably
combine to depress yields and increase production risks in many world regions,
widening the gap between rich and poor countries (IPCC, 2007).
Most of the discussion on climate change up to now has focused on mitigation
measures, for example the Kyoto Protocol. Not much attention has been given to
climate-change adaptation, which will be critical for many developing countries but
also for many regions in Europe. Each country should examine separately, how it can
reduce their vulnerability to climate change and increase desirable outcomes with the
lowest costs under consideration of the different regional and local economic and
soil-climate conditions.
Adaptation to climate change requires resilient knowledge on the potential
regional and local impacts of climate and weather extremes. Effects of climate
change on agriculture may be positive or negative, depending on the variability of
weather conditions, site quality, land use and management. Adaptation must consider
sustainability with respect to high plant production without loosing different
ecosystem services like soil protection, purification and recycling of water or
maintenance of biodiversity. Further, adaptation measures should not enhance
climate change but reduce greenhouse-gas emissions. That implies decision making
to consider both socio-economic and ecological consequences of adapted
management.
Recent developments in geographical information systems, in the development of
robust climate impact models, but also better technologies for data acquisition, have
enabled modelling to identify potentials and environmental constraints to crop
production at regional and national levels regarding the expected climate change. By
integrating models in interactive usable decision support systems, farmers and other
stakeholders have a framework, which can be used to find answers related to the
most appropriate management practices to adjust the agriculture to the climate
change.
Because there is a lack of good model based decision support systems for climate
adaptation of agriculture on the regional and local scale, five years ago in Germany
the now finished joint research project LandCaRe 2020 (Land, Climate and
Resources) had been started. LandCaRe 2020 (www.landcare2020.de) investigated
effects of regional climate change on agricultural production as well as water and
matter fluxes to provide a knowledge-based framework for adaptation. The project
consisted of ten sub-projects co-ordinated by the Department of Meteorology at the
130
Technical University of Dresden and conducted at six research institutes in Germany.
Funding had been provided by the German Ministry for Education and Research
within the funding program “Research for climate protection and protection from
climate impacts”.
Central objective of the project LandCaRe 2020 was a web-based, dynamic
decision support system (dDSS), known as LandCaRe-DSS, exemplarily developed
for two contrasting regions of Eastern Germany (dry lowlands of the State of
Brandenburg and a humid mountain area of Free State of Saxony). The conceptual
framework and the integration of different modules within the LandCaRe–DSS are
represented in Figure 1.
Fig. 1. Conceptual framework and levels of integration of different modules in the LandCaRe-
DSS
2 Methods and Materials
System Structure and Characteristics
The LandCaRe-DSS consists of five basic system components:
• Information and advisory system,
• Analysis of climate data and climate impacts on plant phenology,
131
• Climate change impact assessment for agriculture on national level,
• Climate change impact assessment on regional and farm level,
• Simulation and integrated assessment of different agricultural adaptation
strategies to climate change.
The basic principle of operation, which can be characterized as an iterative
procedure from the scenario definition, the evaluation of different agricultural farm
management adaptation strategies, up to decision, what is the best adaptation strategy
for the concrete farm, is shown in Figure 2.
Fig. 2. Basic principle of operation of LandCaRe-DSS
As distinguished from other Decision Support Systems (DSS) the LandCaRe-DSS
offer the following special features:
• Interactive
(The user decides which simulations and calculations to execute and runs
almost all models by himself.)
• Dynamic
(A large variety of simulations can be run, analyzed and compared with each
other by the user. The chosen preconditions will affect the simulation results.)
• Spatial-oriented
(The user chooses the desired level of detail by zooming between national,
regional or farm level. Based on this choice different models can be activated
for execution.)
• Web-based
(Central support, control and update of the entire DSS software and all
supporting data)
• Extendable
(Open for further add-ons; frequent update of information, knowledge and
data.)
132
Short Description of the Functionality
As a prerequisite for the development of a DSS, potential users like agricultural
administrations, farmers, water agencies and agro-businesses were included in the
research process. As far as possible their specific demands of knowledge and
decision support could be considered. The spatial DSS is interactive and dynamic,
because it allows new model runs with various sets of scenarios and parameters by
the user. It includes regional climate trends and weather statistics of the past,
recorded by climate stations of the German Weather Service (DWD) since 1950, as
well as different future climate scenarios based on the ECHAM5 Global Circulation
Model with dynamic downscaling steps (20 to 1 km grid length) using the Europe-
wide regional climate model CLM (Böhm et al. 2006) and results of the statistical-
dynamic model WETTREG (Enke et al. 2005). Besides multi-ensemble climate
scenarios, socio-economic scenarios are provided by the agricultural information
system RAUMIS to derive the actual and potential future land use and to provide
data for the transfer of the system to other regions in Germany. The module
“ecology” comprises simulations of water, carbon and nitrogen fluxes as well as
yield predictions of agricultural crops and grassland. Simulations are performed by
the central, CO2-sensitive model MONICA (MOdel for NItrogen and Carbon in
Agro-ecosystems, Nendel et al., 2011) developed from components of previous in
ZALF developed models like THESEUS (Wegehenkel et al, 2004), HERMES
(Kersebaum, 2007) and AGROSIM (Mirschel & Wenkel, 2007). Further models and
algorithms are used to derive ecological indicators related to nitrogen and water
fluxes, water-use efficiency, primary production, greenhouse-gas emission, soil
erosion and site potential. Modelling is supported by data from FACE (free-air
carbon dioxide enrichment) experiments with agricultural crop rotations at vTI
Brunswick. Besides completed experiments with the C3-species sugar beet, wheat
and barley, the C4-species maize is currently investigated under high CO2 (550 ppm).
The experiments allowed it to include the CO2 fertilizer effect in modelling which is
crucially important for the prediction of crop yield with respect to quality, quantity
and implications on the soil water and energy budget. The project is open for
participation in the user panel and collaboration with other related projects. At the
end of the project, the LandCaRe-DSS verified for the two exemplary regions was
converted into an operative web-based version (http//:www.landcare-dss.de). The
model framework, adapted software and a defined set of required data provide for
future transfer to other regions.
Examples of LandCaRe-DSS Use
Analysis of Climate and Phenological Data
Information about the impacts of climate change on the ontogenesis of agricultural
crops are very important for agro-management . Using the example of winter wheat,
Figure 3 shows the lengths of different ontogenesis stages between sowing and
harvest in comparison for two climate time periods. The DSS-user can chose
different climate scenarios, different time periods and different sowing dates for
133
running the ONTO model. In the result the DSS-user can see the crop reaction and
can draw the consequences for agricultural measures, for instance in spring.
Fig. 3. Length of ontogenesis stages for winter wheat (in days) between sowing and harvest in
comparison between 1975 (inner circle) and 2030 (outward circle).
Climate Change Impact Assessment on National Scale
At the national scale maps with information about changes in crop yields, cropping
structure, farm economies and irrigation demand as a consequence of climate change
are presented to the stakeholders for different time periods. The maps were created
by the research group of vTI Brunswick. Within the LandCaRe-DSS the user can
carry out a statistical analysis which is exemplary shown in Figure 4 (winter wheat
yield at national scale for Germany with high resolution expected in 2025).
134
Fig. 4. Regional winter wheat yields for Germany in 2025 (results of the RAUMIS model
simulation from vTI Brunswick)
Climate Change Impact Assessment on Regional Scale
On the regional scale the ecological impact assessment of climate and land use
changes are realized on a high spatial resolution, i.e. on a minimum pixel size of 1 ha
(100 x 100 m). Most of the models mentioned above can be activated on this level,
but without a coupling to the economic model on this scale. Using different models
calculations are possible for the expected impacts of climate change on yields for
135
arable and grassland, on the potential erosion risk, on the regional actual
evapotranspiration and the whole regional discharge, on the irrigation water demand
and others. At this regional scale a statistical analysis (average, median, histogram
…) is automatically realized. In Figure 5 for the Federal State of Brandenburg (BB)
the irrigation water demand in 2000 (average for BB: 70.7 mm) is compared with the
situation in 2080 (average for BB: 85.4 mm). From Figure 5 it is seen that in 2080
the irrigation area is significant larger than in 2000.
Fig. 5. Distribution of the irrigation water demand in 2000 (left) compared with the situation in
2080 (right) for the Federal State of Brandenburg (simulation with the model ZUWABE)
Local or Farm Scale
At local or farm scale an interactive simulation and integrated impact assessment of
agricultural adaptation strategies to climate change (crop rotation, soil tillage,
fertilization, irrigation, price and cost changes, …) is offered by the LandCaRe-DSS.
The user of the system will be informed about changes of crop productivity (yield,
yield quality), soil fertility (water, carbon and nitrogen contents), water erosion and
farm economy. At farm level the MONICA and YIELDSTAT models are coupled to
the farm economy model (FEM). The LandCaRe-DSS user receives information
about different economic parameters, fertilizer amounts and costs, irrigation water
demands and costs and finally about crop yields and sales profits. For all output
information the variances of results are given based on up to 90 simulation runs. The
results are visualized using normalized bar graphs for a better comparison between
different scenario runs. Figure 6 shows an example for the visualized simulation
results of the model MONICA for a small part of a farm, based at the Google-map
background. The bar graphs are arranged around the fixed part of the farm which is
subdivided in 1 ha (100 x 100 m) pixels. In the upper part first the data of the actual
scenario run can be chosen. Secondly all input information can be activated and
presented. At the left site of the figure there are shown parts of the dynamic results of
136
MONICA as average for the cropping year and as time course (for example the soil
carbon dynamic) for the chosen 30-year time period.
3 Experiences and Conclusions
Most agricultural systems have a measure of built-in adaptation capacity
(„autonomous adaptation“) but the rapid rate of climate change will impose new
pressures on the existing adaptation capacity. Interactive simulation und integrated
impact assessment of agricultural adaptation strategies to climate change (crop
rotation, soil tillage, fertilization, irrigation, price and cost changes, …) are very
important prerequisites to support farmers and other stakeholders to find out cost
effective adaptation strategies to climate change. Only a well informed user can make
good decisions
Fig. 6. Visualization of simulation results for the combined MONICA and FEM models at
farm level for a small part of a farm within the Uckermark district, Germany.
The development of an interactively useable decision support system for climate
change adaption makes high demands on the implementation concept as well as the
modules included into the system. In order to support fast input/response cycles for
climate scenario simulations (long term simulations; often necessary to run multiple
times for decades), it is necessary to utilize the possibilities offered by modern multi-
core computers and often to recode models, available in different programming-
languages, into C++. Another prerequisite is to use a standardized interface to the
137
shared GIS- and parameter database, which feed the models with all necessary
parameters and input data.
Additionally some models had to be modified with regards to contents in order to
make them fit for scenario simulations even in the limited presence of some
regionalized climate variables (e.g. no availability of high temporal resolution
precipitation events for the calculation of the RUSLE's R-factor in the erosion
model). A further important experience is, that the models which are to be included
into decision support systems have to be robust, well documented and above all have
to be separately verified in space and time before their inclusion into the DSS
framework. Trying to find programming mistakes or design errors in a model after
the inclusion complicates things a lot due to the inherent dynamism build into the
DSS and the sheer amount of parameter-combinations.
Because we can use in principle empirical, mechanistic as well as hybrid models
for long term simulations (e.g. to assess the impacts of climate change on yield and
biomass production for agricultural crops), the LandCaRe-DSS supports the
possibility of multi-model-simulations. This is a prerequisite to assess the
uncertainties resulting from different model types, yielding different simulation
results with the same inputs.
Analogously LandCaRe-DSS can be used to evaluate the impacts of different
climate scenarios (A1B, B1, A2, etc.) and different climate regionalization methods
(e.g. CLM, WETTREG, STAR2) onto the simulation results. Experiences from
recent climate impact studies show, that different regionalization methods lead to
different projections of the main important climate elements (radiation, temperature
and precipitation) in time. Because all projections are of the same probability and
interact with the impact-models, the only way to estimate the possible bandwidth of
regional climate changes onto the development of agricultural yields and other
important landscape functions is to use multi-ensemble-simulations (simulations with
data of different climate regionalization methods). For a decision support system this
means, that it has to offer this functionality to the user in such a way, that he can
pursue multi-model and/or multi-ensemble-simulations without difficulty. An
implication of these possibilities is, that the decision support system has to prepare
the often large amount of simulation results in such a way, that the user gains
knowledge instead of is loosing track.
During the development phase of the LandCaRe-DSS a lot of work has been put
into offering the user different kinds of context sensitive help to empower him to
create and run the envisioned scenario simulations on his own. But the first results
from the test phase show, that a reasonable use of such a complex model-based
decision support system at the farm as well as the landscape planning level is only
possible either involving the model's creators or specially trained people.
Furthermore we have learned that a decision support system is never finished. With a
growing user base also the requirements rise and/or change. Furthermore the
scientific knowledge changes quickly and leads usually to new and more powerful
models. This is especially true, if the system is being developed at the cutting edge.
Thus only if it can be assured that the system's developer-team is able to complete
and improve the system in the long term, it can be expected, that a complex decision
support system can enter successfully into the real world.
138
Some important experiences could be attained on the first real applications of the
models in climate and climate-impact studies regarding the quality of the
geographical data in digital maps and climate projection data, offered by
climatologists. Regarding the Geo-data it had to be observed, that often content
doesn't match or locations are imprecise. This leads to problems during the automatic
generation of inputs for the impact-models. Even more problematic is the fact that
the climate projections of different climate regionalization methods sometimes differ
heavily and extreme events which are of great importance for provision strategies
either are not available at all, or their estimations are highly uncertain. These
objectively existing uncertainties in the climate projection data lead to a large
bandwidth of expected climate-effects and increase the difficulty in the search for
cost effective climate adaption strategies. Making available more reliable regional
climate projections which are better harmonized to each other thus proves to be a key
duty for the climate-impact research and adaption of agriculture to the climate
change in the future.
Acknowledgements. This contribution was supported by the German Ministry for
Education and Research, the German Ministry of Food, Agriculture and Consumer
Protection, and the Ministry of Infrastructure and Agriculture of the Federal State of
Brandenburg (Germany).
References
1. Böhm, U. et al. (2006) CLM - the climate version of LM: Brief description and
long-term applications. COSMO Newsletter, 6.
2. Enke, W. et al. (2005) A novel scheme to derive optimized circulation pattern
classifications for downscaling and forecast purposes. Theor. Appl. Climatol.
82:51-63.
3. Fischer, G., Shah, M., Tubiello, F. N. and Velhuizen, H. (2005) Socio-economic
and climate change impacts on agriculture: an integrated assessment, 1990-2080.
Phil. Trans. R. Soc. (2005) B, 2067-2083.
4. IPCC (2007) Impacts, Adaptation and Vulnerability. Contribution of Working
Group II in the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change, Summary for Policemakers. Cambridge University Press,
Cambridge (2007) UK, 7-22.
5. Kersebaum, K.C. (2007) Modelling nitrogen dynamics in soil-crop systems with
HERMES. Nutr. Cycl. Agroecosys. 77, 39-52.
6. Mirschel, W. and Wenkel, K.-O. (2007) Modelling soil-crop interactions with
AGROSIM model family. In: K.C. Kersebaum, J.-M. Hecker, W. Mirschel and
M. Wegehenkel (Editors), Modelling water and nutrient dynamics in soil crop
systems. Springer, Stuttgart, 59-74.
139
7. Nendel, C., Berg, M., Kersebaum, K.C., Mirschel, W., Specka, X., Wegehenkel,
M., Wenkel, K.-O. and Wieland, R. (2011) The MONICA model: Testing
predictability for crop growth, soil moisture and nitrogen dynamics. In:
Ecological Modelling 222 (2011): 1614-1625.
8. Schmidhuber, J. and Tubiello, F. N. (2007) Global food security under climate
change. Proc. Nath. Acad. Sci., USA, December 11 (2007) 104; 19703-19708.
9. Tubiello, F. N., Soussana, J. F. and Howden, M. S. (2007) Crop and pasture
response to climate change. Proc. Nath. Acad. Sci.,USA, December 11 (2007)
104; 19686-19690.
10. Wegehenkel, M., Mirschel, W. and Wenkel, K.O. (2004) Predictions of soil water
and crop growth dynamics using the agro-ecosystem models THESEUS and
OPUS. J. Plant Nutr. Soil Sci. 167, 736-744.
140