5.4.3. Integrated Assessment Using Coupled Models and Expert Panels
Box 5-2. Integrated Regional Climate Impact Assessment in Brandenburg
The Brandenburg project (Stock and Toth, 1996) was an integrated regional
climate impact study. A variety of statistically derived climate change
scenarios were used as inputs to various types of ecosystem and forest
growth models, plant development simulations, and hydrological analyses.
Socioeconomic impacts were limited to human health and energy use. Integration
across sectors and climate impacts was limited, however.
The observed climate of 1937-1992 served as a reference scenario. Six
additional scenarios were developed to capture possible changes in climate
under various assumptions about the frequency, length, and type of anomalous
weather conditions. A synthetic year was composed from a warm and wet
spring, a hot and dry summer, a warm and wet fall, and a warm and wet
winter. The frequency of these extreme years was increased by 20% in Scenario
2; Scenario 3 contained the same extreme year only. Scenarios 4 and 6
were created by modifying the frequency of extreme conditions by 20% in
monthly patterns and summer weather conditions, respectively. Scenarios
5 and 7 contained only synthetic years derived from modified monthly and
summer patterns, respectively.
Brandenburg is the driest region in Germany. This condition is aggravated
by the predominance of sandy soils with low water-holding capacities and
by the lack of substantial amount of water inflows to the region. Two
regional water models were used to analyze the hydrological implications
of climate change. Under all scenarios considered, critical low-flow values
fall significantly. This result would create problems for water availability,
water quality, and aquatic ecosystems.
Three different modeling approaches have been used to study possible
impacts on the composition, productivity, and stability of forests and
natural ecosystems. A global vegetation potential model projects a transition
toward steppe as the dominant vegetation form as a consequence of increasing
dryness; it is uncertain, however, that dryness will increase. On the
other hand, two local forest succession models show that the simulated
potential natural vegetation of beech will be replaced almost everywhere
by mixed stands of oak, lime, and fir. Currently, forestry in Brandenburg
relies mainly on fir stands; simulation results show that there is little
chance for improved conditions at current sites even under optimistic
projections of climates. Reduced water availability would result in substantially
reduced yields. Unfavorable secondary impacts-such as more frequent forest
fires and increased pest and insect damages-together with higher management
costs and timber loss are likely to generate significant economic losses.
Additional components of the study looked at possible impacts on yields
in agriculture, human health, and energy production and use. In addition
to preliminary results, the major conclusion in all these areas was that
additional and more detailed studies are required to improve our understanding
of possible threats, as well as the benefits and disadvantages associated
with possible adaptation strategies.
Box 5-3. Integrated Study: Effects of Increasing
Temperatures on the Ecosystems of the Bornhöved Lake District
The Bornhöved lake district is located in the north of Germany, near
the Baltic Sea; it includes agroecosystems, beech forest, alder bog, grassland,
and a lake. The surface of the catchment is about 5 km2, including 1 km2
of lake surface. The dominating soil types are dystric arenosols on the
All results are based on a climate data set of 60 years with a maximum
increase of 2.7ºC; rainfall was assumed to be unchanged. Potential evapotranspiration
(pET) would increase by about 100 mm, but actual evapotranspiration (aET)
would differ only slightly from observed values because soil water storage
would become a limiting factor. During summer, there may be considerably
more drought stress. Soil moisture repletion in autumn and winter would
take considerably longer: The profile would not be filled until January.
As a result, farmers would have to change their land use from spring cereals
to winter crops or install irrigation systems on the light soils.
The extent of the relatively flat catchment area depends on the amount
and spatial distribution of groundwater recharge. Because of higher evapotranspiration,
groundwater recharge would be reduced by 25% (100 mm/year), and the catchment
area of the lake would decrease. Overall, the inflow into Lake Belau could
decline by about 50%.
The discharge of the rivers could drop by about 30% compared with present
conditions. The frequency of low water levels could increase, causing
additional stress for flora and fauna in the rivers. Water quality in
the eutrophic Lake Belau is determined by the load of the river that flows
through the lake. As a result of decreasing discharge, there would be
higher nutrient concentrations in the groundwater and the river water;
the overall nutrient load will increase substantially.
Under unchanged management, nitrogen concentrations in the soil solution
would rise, but the overall loss would decrease slightly. However, a change
in the crop rotation and a reduction of fertilization from the present
340 kg N/ha to 163 kg N/ha would affect the nitrogen leaching much more.
Vegetation studies were conducted for primary production of alder, vegetation
structure in grassland, soil respiration, changes in the length of the
growth period in beech forest, and production of beech and Carex acutiformis.
Zoological case studies were carried out for molluscs, birds, mice, several
invertebrates, and the food chain of the robin. Particularly interesting
is the decrease in the number of long-distance (trans-Sahara) migratory
birds-a phenomenon that also has been observed in other regions of Germany.
One possible explanation could be that with increasing temperatures, the
nests and territories of these birds already have been occupied by nonmigratory
birds and birds with short passage that can respond more quickly to the
changing conditions of rising temperatures.
Changes in agricultural yield were simulated for wheat, corn, rape, grassland,
and beans. Because of the light soils in the research area, corn yield
will decrease dramatically under scenario conditions, and the risk of
bad harvests will increase. An earlier seed time and the use of other
varieties of corn has almost no effect on yields. Because the agriculture
of the region is dominated by dairy production, food for cattle has the
highest priority. To minimize the risk of bad harvests, farmers should
irrigate or replace corn with grassland, which uses winter rainfall more
Under scenario conditions, rivers and lakes will be more severely affected
than terrestrial ecosystems. The structure of the landscape will be more
pronounced between dry uphill patches and downhill regions with access
to groundwater. The soil water storage capacity will become much more
limiting and will increase the local risk of bad harvests; adaptation
of land-use practices would be necessary. Despite numerous efforts to
predict species composition in terrestrial and aquatic ecosystems, predictions
on a local scale are not possible. Changes should be evaluated by a combination
of models and monitoring programs (predictive monitoring).
Source: Hörmann et al., 1995.
A third form of integrated impact assessment is the use of coupled models and
expert panels. The knowledge of a range of experts on various impacts-supported,
where possible, by a suite of models-is combined; consistency is attempted through
common scenarios, soft-linking of models (i.e., using output from one model
as input to another), and discussion. This approach better reflects the full
richness of the literature and the complexity of the issues (including nonmodeled
parts), but it lacks the solid outcome of a model. This approach has been used
for the United Kingdom (CCIRG, 1991, 1996), Finland (SILMU, 1996), the state
of Brandenburg in Germany (see Box 5-2), the Bornhöved lake district in Germany
(see Box 5-3), and land use in England and Wales (Parry et al., 1996). Ensuring
consistency only in scenarios, the UK and Finland studies are least ambitious
in integration yet most comprehensive in areas and topics covered; relevant
findings are discussed elsewhere in this chapter. The Brandenburg and Bornhöved
studies consider only small areas, focusing on the impact of climate change
on ecosystems. The study on England and Wales uses world food prices (from Rosenzweig
and Parry, 1994) and climate change scenarios to investigate land-use changes;
a major result is that, although the aggregate agricultural output of this region
is unlikely to change drastically, the spatial distribution of activities may
well change substantially.