A macroeconomic assessment of impacts and adaptation to climate change in Europe

Full text

CICERO Report 2009:06
A macroeconomic assessment of impacts
and adaptation to climate change in
Europe
Asbjørn Aaheim, Helene Amundsen, Therese Dokken, Torgeir Ericson, Taoyuan Wei
September 2009
CICERO
Center for International Climate
and Environmental Research
P.O. Box 1129 Blindern
N-0318 Oslo, Norway
Phone: +47 22 85 87 50
Fax: +47 22 85 87 51
E-mail: admin@cicero.uio.no
Web: www.cicero.uio.no
CICERO Senter for klimaforskning
P.B. 1129 Blindern, 0318 Oslo
Telefon: 22 85 87 50
Faks: 22 85 87 51
E-post: admin@cicero.uio.no
Nett: www.cicero.uio.no

Tittel:
Title: A macroeconomic assessment of impacts and
adaptation to climate change in Europe
Forfatter(e):
Author(s): Asbjørn Aaheim, Helene Amundsen,
Therese Dokken, Torgeir Ericson, Taoyuan Wei
CICERO Report 2009:06
sider
CICERO Report 2009:06
50 pages
Finansieringskilde:
Financed by: NFR
Prosjekt:
Project: EUs 6. rammeprogram (ADAM)
Prosjektleder:
Project manager: H Asbjørn Aaheim
Kvalitetsansvarlig:
Quality manager: Knut H Alfsen
Nøkkelord:
Keywords: Virkninger av klimaendringer.
Tilpasning. Makroøkonomiske modeller
Sammendrag:.
Abstract:
Språk:
Language of report: English
Rapporten kan bestilles fra:
CICERO Senter for klimaforskning
P.B. 1129 Blindern
0318 Oslo
Eller lastes ned fra:
http://www.cicero.uio.no
The report may be ordered from:
CICERO (Center for International Climate and
Environmental Research – Oslo)
PO Box 1129 Blindern
0318 Oslo, NORWAY
Or be downloaded from:
http://www.cicero.uio.no

Contents
1 Introduction ..................................................................................................................................... 3
2 The Integration of climate change impacts ..................................................................................... 4
3 Regions and sectors in GRACE ...................................................................................................... 5
4 Economic impacts of climate change in Europe ........................................................................... 12
4.1 THE CLIMATE CHANGE SCENARIO ......................................................................................................... 13
4.2 AGRICULTURE ...................................................................................................................................... 13
4.3 FORESTRY ............................................................................................................................................ 18
4.4 FISHERIES ............................................................................................................................................. 21
4.5 ENERGY................................................................................................................................................ 22
4.5.1 Demand ........................................................................................................................................... 22
4.5.2 Electricity generation ...................................................................................................................... 25
4.6 MANUFACTURING INDUSTRIES ............................................................................................................. 28
4.7 TOURISM .............................................................................................................................................. 28
4.8 TRANSPORT .......................................................................................................................................... 30
4.9 SERVICES ............................................................................................................................................. 32
4.10 NON SECTOR-SPECIFIC IMPACTS ........................................................................................................... 33
5 Implementation of impacts functions ............................................................................................ 33
5.1 AGRICULTURE ...................................................................................................................................... 34
5.2 FORESTRY ............................................................................................................................................ 34
5.3 FISHERIES ............................................................................................................................................. 36
5.4 ELECTRICITY SUPPLY ........................................................................................................................... 37
5.5 ENERGY DEMAND ................................................................................................................................. 38
5.6 TOURISM .............................................................................................................................................. 38
5.7 EXTREME EVENTS ................................................................................................................................ 39
5.8 SEA-LEVEL RISE ................................................................................................................................... 40
6 Results ........................................................................................................................................... 41
7 Conclusions and remaining questions ........................................................................................... 44
Acknowledgements
This report summarises work carried out within the the ADAM project, which was funded by the 6
th
Framework Programme of the European Commission and the Norwegian Research Council. The
project has been co-funded by the ALIANSE project from Norwegian Research Council.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
1
Summary
This report provides a documentation of the integrated macroeconomic general equilibrium
model GRACE_adapt, and presents results from the macroeconomic analysis of impacts and
adaptation in Europe in case of a +2 °C and +4 °C increase in global mean temperature.
GRACE_adapt has been developed for the ADAM project to address economic impacts of
climate change and resulting adaptation, interpreted as the economic responses to climate
impacts. The development of the integrated computable general equilibrium model
GRACE_adapt represents an attempt to bring consistency between adaptation to climate
change and economic behaviour. The model is based on the idea that adaptation to climate
change can be interpreted within the context of economic behaviour: Climate change may
lead to changes in the availability of economic resources, or shifts in the demand for
commodities or services, either from production sectors (shift in technologies) or from final
demand (shift in preferences). The resulting shift from one general equilibrium prior to
climate change to a new general equilibrium posterior to climate change can be interpreted as
adaptation.
GRACE_adapt is a general equilibrium model for an optional number of regions and
economic sectors. The model divides Europe into eight main regions. The selection of regions
was made subject to three criteria. First, each region has to be geographically connected, and
have similar climatic conditions. Second, the population of all regions should be of similar
size. Third, income per capita should be of the same order of magnitude within a region. The
model divides the economy into 11 sectors of production expected to be particularly affected
by climate change, i.e. sectors that utilize climate sensitive natural resources or sectors for
which the demand is sensitive to the climate (the sectors are listed in Table 3.3).
Contrary to most other integrated models where the impacts of climate change are aggregated
into a “total cost of damage”, the impacts of climate change in GRACE_adapt are attached to
specific economic activities or variables. Moreover, barriers to adaptation are introduced by
splitting each of the eight European regions represented in the model into 9 to 11 provinces,
between which we impose constraints to the mobility of primary input factors. Adaptation to
climate change can thereby be analysed endogenously by means of the assumptions
underlying the behaviour of economic agents.
In addition to providing a documentation of the modelling, this report presents the first results
of the model runs, which addresses one of the two questions that were raised as the
background for the ADAM project; namely to compare the costs of mitigation to aim at +2 °C
increase in global mean temperature with the need for adaptation if the global mean
temperature increase by +4 °C.
We found that adaptation contributes to reduce the macroeconomic impacts of climate
change, but that the share of costs that is reduced by adaptation decreases as global mean
temperature increases. The analysis gives some justification of the +2 °C target advocated by
the EU: The +2 °C case shows small economic impacts, with positive changes in some
regions and negative in others. For higher temperature increases, the impacts are negative in
all regions. We also note that the impacts are non-linear: A +1 °C increase in the global mean
temperature results in a doubling of the costs of climate change. Adaptation takes place in
terms of changes in quantities, but changes in prices also constitutes an important part of the
adaptation process when comparing GDP in the different cases. The total effect of adaptation
is substantial, and reduces the impacts by approximately 80 - 85 percent in many regions.
However, this large effect of adaptation may partly be explained by the assumptions in the
model that the world outside Europe is unaffected by climate change. Both the level of
economic impacts and the degree to which adaptation mitigates the initial economic costs of

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
2
climate change vary greatly both between the eight European regions and within the regions.
The southern regions are the most affected by climate change, whereas the low-income
regions seem to take the most advantage from adaptation. This can be explained by the
change in relative prices, which turn out in favour of the low-income regions. The low-
income regions also exhibit relatively large variations within the region. This can also be
considered a result of the changes in relative prices: Some provinces are better capable of
taking advantage of this than others. Also the most affected regions exhibit large variations
across provinces. In some regions there is a strong correlation between low income per capita
and losses in GDP or reductions in wages on the province level, but such a pattern cannot be
found in all regions.
It has to be emphasised that the results of the present study are preliminary and subject to
notable weaknesses. First, we present only a static comparison between equilibrium under
different assumptions about climatic conditions. To get a more realistic picture of the
economic consequences, one should run scenarios over a given time period. In that case, a
closer examination of the barriers to adaptation would have to be considered, because barriers
will diminish over time. Second, we have assumed that the rest of the world is unaffected by
climate change. As mentioned, the effect of adaptation depends heavily on the change in
relative prices, also on the world market. Implementation of impacts in the rest of the world is
therefore likely to affect the results considerably. Hence, the levels of impacts reported here
are probably biased, which is why we do not want to highlight them. However, we believe
that we have gained some insights when it comes to the comparison of European regions and
variability within regions.
We would also like to draw the attention to the uncertainties about the estimates of the impact
of climate change. The model is based on external assessments of impacts by activity. This
requires, in principle, a comprehensive set of impact studies in which the relationships
between climate indicators and impacts on activities can be based. There are, however, few
available cost estimates of the impacts of climate change, not to mention adaptation. Those
that exist will therefore have to be stretched to their limits, in the sense that a specific estimate
will be used as if it applies to more cases than actually addressed in the referred study. In
some cases results will have to be transferred to apply in other sectors or sub-sectors, and in
some cases to other regions.
Moreover, the impacts that were taken into account probably represent moderate estimates,
because some effects of climate change have been disregarded. Health effects, for example,
which have economic consequences only to the extent that economic activities are affected,
are not included at all in the numerical analysis, and all impacts are imposed by means of
expected values. The most severe economic impacts of climate change are related to extreme
events, which are stochastic by nature. Stochastic events of certain magnitudes may
themselves generate negative impacts, which have been disregarded here. A closer study of
these impacts was, however, addressed in Work package A2 of the ADAM project.
Finally, there have been no attempts to include non-market impacts of climate change, which
may turn out to be substantial. Hence the results of this study must be considered a “pure”
economic assessment of impacts. A full evaluation of the impacts of climate change requires
inclusion of many more aspects.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
3
1 Introduction
The starting point of the ADAM project was to compare the costs to Europe of limiting the
increase in global mean temperature to +2 °C by mitigation efforts with the costs of having to
adapt to an increase of +5 °C. Later, it was concluded that an increase in global mean
temperature at +5 °C is probably too pessimistic even under a high emissions scenario, and
the adaptation case was moderated to somewhere between +4 °C and +4.5 °C. While tools
for assessing the costs of mitigation were readily available long before the project started, the
attention to the costs of impacts and adaptation has been limited, although so-called integrated
models have become more common over the past decade. These models are, however, still in
their infancy, and one of the general critiques has been that adaptation is poorly addressed, if
at all.
To do an assessment of the costs of adaptation on the macroeconomic level, new modelling
tools were therefore required. This is the background for the development of GRACE_adapt.
The model is based on the idea that adaptation to climate change can be interpreted within the
context of economic behaviour: Climate change will change the national wealth, such as the
value of natural resources, the capital stock and the labour force. Changes in climate will,
moreover, lead to a shift in the demand for certain commodities or services, both in
production sectors (shift in technologies) or from final demand (shift in preferences). The
resulting shift from one general equilibrium prior to climate change to a new general
equilibrium posterior to climate change can be interpreted as adaptation.
GRACE_adapt differs from standard computable general equilibrium models in two respects.
First, the impacts of climate change are integrated activity by activity in the model with
separate impact functions for each sector in each region. Thus, an impacts function for each
activity is needed. For example, the impacts of climate change to agriculture have to be
attached to the activity within the agricultural sector where the impacts arise. Thus, a
reduction in agricultural productivity because of a drier climate can be interpreted as a
reduction in productivity of land. The productivity of land is represented by the input of
natural resources, and hence, the impact manifests itself in a reduction in the value of natural
resources for agriculture. Damage to buildings because of extreme events is, on the other
hand, introduced as damage to the capital stock in the sector.
Most other integrated models include impacts of climate change by means of aggregated
damage functions. Then, the aggregated “damage cost of climate change” is subtracted from
the output of the economies. The activity based implementation of impacts in GRACE_adapt
has the advantage that economic behaviour is made consistent with adaptation to climate
change instead of being considered something different that has to be determined “outside”
the model.
A challenge in analysing adaptation in the context of general equilibrium models is that
economic markets respond immediately without time to adapt. Thus, capital, labour and
natural resources switch from one sector to another within the same economy without
frictions. Within larger regions, adaptation to climate change would hardly involve major
problems if this was the case. The second difference between GRACE_adapt and standard
integrated models is that adaptation is subject to frictions across sub-regions. Each region is
split into sub-regions, called provinces, between which there are constraints to the mobility of
primary input factors or goods. The result is that “local markets” develop for factors subject
to mobility constraints, so that wage levels can vary between provinces within a region.
The model is based on external assessments of impacts by activity. This requires, in principle,
a comprehensive set of impact studies on which the relationships between climate indicators
and impacts on activities can be based. There are, however, few available cost estimates of
the impacts of climate change, not to mention adaptation. Those that exist, henceforth called

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
4
basic estimates, will therefore have to be stretched to its limits, in the sense that a specific
estimate will be used as if they apply to more cases than actually addressed in the referred
study. In some cases results will have to be transferred to apply in other sectors or sub-
sectors, and in some cases to other regions.
When transferring the estimates between sectors or regions, it is nevertheless possible to take
into account known conditions that affect impacts, thereby making estimates site or sector
specific even when based on studies elsewhere. These are:
 The basic estimates refer to a certain scenario for climate change. A transfer will
take into account that the climate changes differently in different regions.
 A basic estimate applies for a certain sector with a given activity. In most cases
the sector is a sub-sector of the sectors in the GRACE_adapt model. Transferring
the estimate requires that the sub-activity is adjusted to reflect the relative
importance of the activity within the aggregated GRACE_adapt sector.
 Basic estimates may apply to parallel sectors and commodities or in more general
contexts. To draw parallels or to generalise, information which is insufficient to
provide a basic estimate may nevertheless apply as support.
This report describes how basic estimates are used to adjust the parameters of the model, and
provides some results from the analysis of impacts and adaptation. First, we describe in
general what kind of information GRACE_adapt needs to integrate impacts of climate
change. Next, the sectors and regions are described and defined. Moreover, we point out the
sectors where we expect to identify impacts of climate change and suggest functional
relationships between adjustments and climatic parameters. Then, we present results from
studies of climate change impacts to distinguish basic estimates from supportive information.
Finally, we come up with the adjusted estimates by sector and region.
2 The Integration of climate change impacts
GRACE_adapt is a general equilibrium model for an optional number of regions and
economic sectors. For practical purposes it is necessary to limit the number of sectors and
regions to a certain extent. Impacts of climate change are attached to specific economic
activities of the model, in accordance with the ideas presented in more detail in Aaheim and
Schjolden (2004) and in Aaheim, A. (2007). In general, the way economic activities are
adjusted for a change in climate can be divided into three categories:
i) A change of input to economic sectors. That is, a given change of climate will
force sectors to use a different composite of input to produce a unit of output;
ii) A change in the availability of primary resources (natural capital, real capital and
labour);
iii) A change of preferences, or final deliveries.
The numbers that enter the model are the changes in aggregated volumes by sector and region
attributable to climate change. The volume is measured by the value in a base year. In this
context prices are used as a means to aggregate different physical units within a sector and
across sectors. Thus, the volumes are meant to represent physical amounts, and the
adjustments i) – iii) are interpreted as a change of the technology in a sector: climate change
affects the number and composition of inputs needed to produce a given amount of output.
The initial adjustments are made under the assumption that the economic agents adjust their

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
5
input and consumption bundles in fixed proportions. Hence, adjustments are imposed to the
model as if independent of economic behaviour.
It should be noted, however, that aggregation of firms and groups of final demanders into
sectors and final demand, respectively, implies that there is a certain economic behaviour
underlying the term “technology”. This is due to the assumption that each aggregate is
composed by elements such that the cost of producing the aggregate is minimized. When
applied to studies of impacts of and adaptation to climate change, this has two implications.
The first is that the “technology” of a sector is a result of adaptation within the aggregated
sector. If, for example, climate change improves the agricultural conditions for crops but
worsen the conditions for cattle, some farmers will probably switch from cattle to crops
because the value of their land increases if used for crops but decreases if used for cattle. To
tell how far this substitution will go one may use the basic economic first-order condition that
the marginal value of cattle land equals the marginal value of crop land. However, the
agricultural sector comprises both crops and cattle in GRACE_adapt. The substitution
between sub-sectors therefore has to be represented by an exogenous shift of the technology
in agriculture, even though this change is, in fact, brought about by economic market
behaviour.
The second implication is that available measures to adapt to climate change in order to
reduce the costs of impacts are assumed to be implemented proactively if the expected
benefits exceed the costs. In other words, adjustments of the technology made to represent
impacts of climate change should be net of proactive adaptation. Analysing adaptation by
macroeconomic models does not, therefore, mean that assessments of adaptation are avoided,
but one has to make a clear distinction between proactive implementation of adaptation
measures and autonomous adaptation.
3 Regions and sectors in GRACE
GRACE_adapt divides Europe into eight main regions. The choice of regions were made
subject to three criteria. First, each region has to be geographically connected, and have
similar climatic conditions. Second, the population of all regions should be of similar size.
Third, income per capita should be of the same order of magnitude within a region. The main
regions are described in Table 3.1.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
6
Region
Countries
Pop
Area
GDP pr capita
million 1000
km
2
€
Range
*
Baltic states Estonia, Latvia, Lithuania,
Poland
45.7 188.2 6 397 0.41
British islands
Ireland, United Kingdom
64.6
121.6
30 553
0.29
Central Europe East Czech rep. Hungary, Slovakia,
Romania, Bulgaria
55.3 219.7 5 962 1.17
Central Europe North
Austria, Germany, Switzerland
98.1
186.1
28 391
0.45
Southern Europe Cyprus, Greece, Italy, Malta,
Slovenia
72.0 178.7 22 689 0.56
Central Europe West Belgium, France, Luxembourg,
The Netherlands
88.2 240.0 28 371 0.31
Iberian peninsula
Spain, Portugal
51.0
230.6
19 486
0.35
The Nordic countries Denmark, Finland, Iceland
Sweden, Norway
24.6 485.9 36 944 0.61
Other developed
countries
Developing countries
*) For each region the range is the difference between the highest and lowest GDP/capita divided by the region
average
Table 3.1. Regions in the GRACE_adapt model
To match all three criteria, it is necessary to make compromises. Population ranges from 24.5
mill. in the Nordic countries to 98 million in Central Europe – North. For this reason it might
be advantageous to couple the Nordic countries with the Baltic states, which gives a total
population of approximately 70 mill. However, the income level differs substantially between
two regions. Moreover, the climatic conditions vary to a certain extent, partly due to the fact
that the Nordic region covers the largest area among all the European regions. The largest
relative range of income inequality by country within a region is Central Europe – east. This
is because 1) this region has the lowest average income level, and 2) the low income in
Bulgaria. The population of Southern Europe is at the average of all regions and the four
countries in this region have a similar climate, although the income differences are substantial
between the country with the lowest income (Malta) and the country with the highest income
(Italy).
For each of the eight main regions, the basic economic model is calibrated by use of separate
national accounts data as reported by the GTAP database (Dimaranan, 2006). Impacts of
climate change are integrated by means of estimated impacts functions, which are reported in
Section 5. The idea of attributing impacts to economic activity is that the model determines
adaptation to climate change when solving for the new equilibrium after the impacts have
been implemented. An underlying assumption of this interpretation is that all the available
resources within a region can switch from one activity to another without other barriers than
the opportunity cost. This is a strong assumption, indeed, and particularly inappropriate when
addressing adaptation to climate change. There are reasons to believe that a major share of the
costs of adaptation arise as a result of “market frictions”. Labour, real capital and natural
resources cannot be moved easily around within a region in order to make them less sensitive
to climate change.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
7
To take frictions such as these at least to some extent into account, each region is further
divided into between 9 and 11 sub-regions, called provinces, between which we can impose
restrictions on the mobility. As there are no national accounts data on the province level, the
economic activities of a province in the current version of GRACE_adapt is a mirror of the
total economic activity of the region, but divided according to each province’ share of total
GDP. With more information about sectoral composition in the different provinces, a more
sophisticated split would be possible.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
8
Bal ti c States litu lt0 Lithuania
lves lv0 Latvia ee0 Estonia
pnow pl42 Zachodn. pl63 Pomorskie
pnoe pl62 Warm.-Mazur. pl34 Podlaskie
pnmw pl43 Lubuskie pl41 Wielkopolskie
pcem pl61 Kuj aw.-Pomor. pl11 Lódzkie
pma z pl12 Mazowieckie
psoe pl31 Lubelskie pl32 Podkarpackie
pms e pl33 Swietokrz. pl21 Malopolskie
psla pl22 Slaskie
psow pl51 Dolnosl. pl52 Opolskie
British Islands ukno ukm Scotland ukc North Ea st
uknw ukd N-West UK
uksw ukl Wales ukk South West
york uke Yorksh
ukwm ukg West Midl.
ukem ukf East Midl .
ukse ukj Sout-Eas t UK
lond uki London
ukea ukh Ea st UK
irel ukn Nor. Ireland ie0 Ireland
Central Europe Ea st buea bg33 Severoizt. bg34 Yugoiztochen bg42 Yuzhen t
buwe bg31 Severozap. bg32 Severen t bg41 Yugozapaden
rves ro11 N-W Rom. ro42 W Rom
rces ro12 Centru ro41 S-V Oltenia
rest ro21 N-E Rom ro22 S-E Rom
rsub ro31 S-Muntenia ro32 Bucur.-Ilfov
slov sk0 Slovakia
hwes hu10 Közép hu21 Közép hu22 Nyugat hu23 Dél
hea s hu31 Észak hu32 Észak-Alföld
cwes cz01 Praha cz02 Strední Cechy cz04 Severozápad cz03 Ji hozápad
ceas cz05 Severových. cz06 Jihovýchod cz07 Str. Morava cz08 Moravskosl.
Central Europe North dbad de1 Baden-Würt.
dbay de2 Bayern
des t de3 Berl i n de4 Brandenb. dee Sachsen-Anhalt
dnor def Schl.-Holstein de8 Meckl enb. de6 Hamburg
dnib de5 Bremen de9 Ni edersachs en
dnwe dea Nordrh.-Westf.
dmws dec Saarland de7 Hessen deb Rheinland
ds ea ded Sachsen deg Thüringen
swi t ch Switzerland
aust at Austria
Southern Europe gsou cy0 Cyprus gr3 Attiki gr4 Kriti
gnor gr1 Voreia Ellada gr2 Kentriki Ellada
inea si0 Slovenia itd1 Bolzano-Bozen itd2 Trento itd3 Veneto itd4 Venezia Giulia
inwe itc1 Pi emonte itc2 Valle d'Aosta itc3 Liguria
imwe itd5 Emilia-Rom i te1 Toscana
ilau ite4 Lazi o ite2 Umbria
isea i te3 Marche itf1 Abruzzo itf2 Molise itf4 Puglia
icam itf3 Campania
ilom itc4 Lombardia
isou itf5 Basilicata itf6 Calabria itg1 Sicilia itg2 Sardegna mt0 Mal ta
Central Europe Wes t file fr1 Îl e de Fra nce
fpar fr2 Bassin Parisien
fcal fr3 Pas de Calais
fes t fr4 Es t France
fea s fr5 Ouest France
fsoe fr6 S-O France
fces fr7 Centre-Est
fmod fr8 Méditerra née fr9 French oversea
nsoe nl4 Zuid-Nederl nl2 Oost-Nederland
nwen nl1 Noord-Nederl nl 3 West-Nederland
bel u be1 Bruxel les be2 Vlaams be3 Wallonne be34 Luxemb (B)
Iberian Peninsula pnor pt11 Norte
pcen pt16 Centro (PT) pt17 Lisboa
espc pt15 Algarve pt18 Al entej o pt20 Açores (PT) pt30 Madeira (PT) es53 I l l es Ba l ears es 63 de Ceuta (ES) es64 Melilla (ES) es70 Canarias (ES)
snwe es11 Galicia es12 Asturias es13 Cantabria
snor es21 Pais Vasco es22 Navarra es23 La Ri oj a es24 Aragón
sma d es30 Madrid
scas es41 León es42 la Mancha es43 Extrema dura
scat es51 Cataluña
svam es 52 Val enci a na es62 Murcia
sand es61 Andalucia
The Nordic Countries fnor fi13 Itä-Suomi fi1a Pohjois-Suomi fi19 Länsi-Suomi
fsou fi18 Etelä -Suomi fi20 Åland
dcop dk01 Copenh
dsou dk02 Sjælland dk03 Syddanmark
dnor dk04 Midtjylland dk05 Nordjylland
ssto se11 Stockholm
smi d se12 Ö Mel l a nsv
ssou se21 Smål + i s l s e22 Sydsverige
swmi se23 Västsv s e31 N Mel l a nsv
nord se32 Midt Norrl se33 Övre Norrl. is00 Iceland no07 N-Norge no06 Trøndel.
nea s no01 Oslo area no02 Hedm.-Oppl no03 Telem-Busk
nwes no04 Agder-Rogal no05 Vestlandet
Table 3.2. Regions and provinces GRACE_adapt and associated NUTS codes

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
9
The provinces are based mainly on aggregation of NUTS 2 level regions in Europe. Each
province is geographically connected, with an exception of the Iberian islands.
Sector
Direct impacts
Affected sub-sectors
Agriculture
Productivity of land
Crops, Vegetables, Livestock
Forestry
Biomass growth
Fisheries
Recruitment and migration of stock
Ocean, Coastal, Farming
Manufacturing
Transport
Routes, mode choice tourism
Land, Sea, Air
Service sector
Demand for holiday tourism, energy demand
Electricity On renewables, cooling in thermal plants
Refined oil
Crude oil
Gas
Demand for heating purposes
Coal
Demand for heating purposes
Table 3.3. Sector-list of GRACE_adapt, main impacts and sub-sectors with potential
differences of impacts
The model divides the economy into 11 sectors of production, listed in Table 3.3. The choice
of sectors is meant, firstly, to separate those particularly affected by climate change. These are
sectors that utilize climate sensitive natural resources or sectors for which the demand is
sensitive to the climate. Table 3.3 indicates the main direct impacts and point out certain sub-
sectors for which climate impacts are likely to be substantial. Secondly, energy sectors are
represented to a certain degree to allow for comprehensive policy analyses of emission
control.
The impacts listed in Table 3.3 are specific impacts to each sector. There are also more
general climate change impacts, such as extreme events and sea-level rise. These are in most
cases integrated by a loss in terms of the total availability on real capital to reflect, for
example, damage to buildings. Table 3.4 lists the most important general effects, suggests
impacts and how they are incorporated into the model
Category
Impact
Representation in model
Water supply To agriculture, health, irrigation Value of land, value of labour
Sea-level rise Flooded areas, degradation of construction
fundaments, fresh water supply
Capital stock
Extreme events Damage to buildings and construction,
damage to arable land, injuries and killed
people, forests
Capital stock, labour. Productivity
in agriculture
Biodiversity Medicine potential, ecosystem resilience,
food supply
Value of land
Health
Vector borne diseases, heat/cold stress
Labour stock
Table 3.4. Cross-over impacts of climate change and possible representation in model

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
10
Final demand in the model consists of private and public consumers, investments and exports.
Final demand may be affected directly by climate change, in the case of electricity demand
for cooling and energy demand for heating. The demand for tourist services is likely to be
affected. Health effects may affect the demand for health services, but also the supply of
labour.
The indirect economic impacts of climate change arise because of the integration of sectors in
the economy. Restoring capital loss in one sector triggers the demand from the building and
construction sectors, a change in the productivity of land in agriculture changes the demand
for labour and capital, etc. These indirect impacts are expressions of the socioeconomic
adaptation processes, which are addressed by the market behaviour of the model. This
behaviour is basically a result of how the demand is modelled. GRACE_adapt applies so-
called nested CES-trees (Constant Elasticity of Substitution) to determine demand. A nested
CES tree consists of aggregates, X, which are being produced by two or more input factors V
i
,
which themselves are aggregates or input factors, according to CES production functions. In
the two-factor case, the CES function is written as
The parameter ρ is chosen on the basis of assumptions about the elasticity of substitution,
which is 1/(1 – ρ), while A and α are calibrated by reference to the GTAP database. Figure 3.1
displays the nested CES tree for the production sectors.
Straight lines between the aggregate and the input factors indicate that the input factors are
substitutes. An angled line indicates no substitution, or a so-called Leontief technology
(aggregates C and H). The implication is that a change of relative prices between input factors
affects the composition of the input when substitution is assumed, but not when a Leontief
technology is assumed. The bottom line shows the input from the specified sectors, which the
model describe as an aggregate of input from sectors domestically and abroad. These so-
called Armington aggregates were imported and domestic deliveries are substituted at a
relatively high rate.
Impacts of climate change on economic activities may now be divided into three categories:
i) it may affect the stock of the primary input factors natural resources, capital and
labour directly.
ii) the productivity (output) may be affected directly.
iii) the productivity (demand) of one of the input factors or aggregates may be affected.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
11
X
A
B
C
D
E
F
G
H
Capital LabourNature
Agric.
Services
Forestry Fisheries
Crude oil
Transport
Coal
Ref. oil
Electr.
Gas
Manuf.
Figure 3.1. Production structure in GRACE-adapt
The corresponding demand tree for the final deliveries are displayed in Figure 3.2:
D
E
F
G
Agric.
Services
Forestry Fisheries Manuf. Transport Crude oil
Coal
Ref. oil
Electr.
Gas
Final
Figure 3.2. Structure of final demand in GRACE-adapt
For primary factors, impacts are represented by adjustments of the availability of resources,
that is, the stock of capital and natural resources or the availability of labour. For impacts on
output, the scale parameter of the CES aggregate may be adjusted according to the
productivity change. Being CES- or Leontief aggregates, which are independent on scale,
impacts on sub-aggregates (aggregates other than “top”) are represented by adjusting the

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
12
distribution parameter of the aggregate that is being affected. For example, a change of
energy demand due to climate change means that the commodity aggregate consists of less
input of energy per unit of “commodity”, which leads to an adjusted distribution parameter.
Adjustments imply that the model is recalibrated according to exogenous information about
the climate. These adjustments are made independently as if this information is produced
without taking market responses into account. In other words, it is assumed that impact
studies on which the adjustments are made, estimate impacts only partially. Running the
model thus results in a new equilibrium, which in the context of the modelling is interpreted
as adaptation to climate change.
The present version of the GRACE model is limited to economic impacts only. Several
studies suggest that non-market impacts may be more important when evaluating the threat of
climate change. These impacts may be related to impacts to health, which have economic
consequences only to the extent that economic activites are affected, or to “services” provided
by the nature such as wild-life, or simply the pleasure of a certain, known environment. Even
more important are the possibilities of major disasters, such as a collapse of ecosystems.
Athough possible in principle, an economic evaluation of the costs of such a collapse with the
perspective taken in the GRACE model would be of limited interest, however, as the
possibility of disasters raises the question of how major risks can be dealt with.
4 Economic impacts of climate change in Europe
This section presents a survey of the literature about expected impacts of climate change on
economic sectors in Europe. The purpose of this survey is to provide a reference to create
estimates of econonomic impacts by sector in all European countries. These estimates are
results of more or less subjective interpretations of the results in the literature, which are not
always easy to compare and transform into numerical terms. The estimates will be used
further in Section 5 as “observations” in the estimation of sectoral impacts functions.
The knowledge of economic impacts of climate change is fragmented. Results from different
studies are based on different climate scenarios and methods, which partly explains why they
come up with different results. In order to make results from different studies comparable,
they have to be adjusted to fit into the same climate scenario and in some cases transferred
from regions where the study was carried out to regions with similar conditions. These
adjustments were made primarily on the basis of our own interpretations, and without any
attempt to apply stringent methods. The estimates used for each separate sector should
therefore be considered as rough estimates only. The estimated impacts are implemented
sector by sector and region by region in the CGE model, in line with the approach presented
in Aaheim and Schjolden (2004).
This report comments on the impacts in nine main sub-sections. The first sub-section covers a
brief overview of the underlying expected changes in climatic variables. The following sub-
sections discuss the main economic impacts by each sector separately. They give a short
presentation of the sector followed by a description of sector specific general impacts of
climate change. This discussion aims at providing the broad overview. Attempts to establish
functional relationships between indicators for climate change and economic activities are
presented in Section 5.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
13
4.1 The climate change scenario
The reference climate change scenario used is a scenario from the Prudence project (REF),
also used in the PESETA Project (PESETA, 2007). PESETA applied the A2 and B2 global
emission scenarios from IPCC (Nakicienovic and Swart, 2000). Our study is based on the A2
scenario, which expects a global warming of 3.1 degrees for the period 2071-2100 compared
to the control period 1961-1990. Based on maps showing temperature and precipitation
changes, the expected changes in the different regions were approximated to those shown in
Table 4.1.
Region °C change in mean
annual temperature
Percent change in
annual precipitation
Baltic States
+ 3.0
+ 13
British Islands
+ 2.5
0
Central Europe East
+ 3.5
- 5
Central Europe North
+ 3.5
+ 5
Southern Europe
+ 4.0
- 15
Central Europe West
+ 3.5
0
Iberian Peninsula
+ 4.0
- 25
Nordic Countries
+ 3.5
+ 10
Table 4.1. Expected changes in climate parameters. From baseline 1961-1990
compared to 2071-2100 in temperature (in degree Celsius) and precipitation (in
percentages) in Europe.
Source: PESETA(2007).
4.2 Agriculture
The contribution from agriculture to the total output varies from 6 percent in the Nordic
countries and northern parts of central Europe to 15 and 18 percent the Baltic States and
eastern parts of central Europe, respectively (see Table 4.2). Also the relative importance of
sub-sectors within the agricultural sector differs. However, fruit and vegetables is the least
important in terms of their contribution to total output in all regions, and the sub-sector called
“other crops” is the most important, while livestock is second. Most of the output from
agriculture is delivered as intermediates to other sectors, such as manufacturing industries and
service sectors. The model determines the implications of changes in agriculture on these
deliveries.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
14
Region Share of
total output
Sub-sectors share of sector total output
Fruit&vegetables
Other crops
Livestock
Baltic States
15
12
54
34
British Islands
6
3
57
40
Central Europe East
18
9
58
34
Central Europe North
6
11
52
37
Southern Europe
7
13
52
35
Central Europe West
7
7
53
40
Iberian Peninsula
9
15
48
37
Nordic Countries
6
4
50
46
Table 4.2: Overview of the agricultural sector in percentages.
Source: GTAP (Dimaranan, 2006)
Climate change affects agriculture regions and sub-sectors of the agricultural sector in
different ways. On average, it is expected that the productivity of crops in Europe will
increase slightly (Alcamo et al., 2007). Crops respond both to increased temperature and to
increased atmospheric concentration of CO
2
. In general, the increase in CO
2
-concentrations
is expected to offset the negative effect of increased temperature and decreased soil moisture
(Long et al., Ainsworth, Leakey, Nosberger & Ort, 2006). However, according to projections
using FACE technology studies (Free Air Carbon Enrichment), the effect of higher CO
2
concentration is lower than previously assumed. Experiments by Long et al (2006) projected a
13 percent increase in crop yields if the concentration of CO
2
increases to the expected 2050
level at 550 ppm, compared to 380 ppm in 2006. This is approximately 50 percent less than
expected by earlier enclosure studies. In terms of temperature, middle and higher latitudes of
Europe will benefit from an extended potential growing season when temperature increase
and cropping areas may expand northwards (Maracchi, Sirotenko, & Bindi, 2005). Southern
parts on the other hand, may experience increased respiration. Accelerated development may
advance maturation of certain crops and cause reduction of yields (Rötter & van de Geijn,
1999). A further complexity is related to the level of tropospheric ozone, which is affected by
climate change, and may also offset the positive effect of CO
2
(Reilly et al. (2007)
Livestock can be influenced through direct effects on animals health, reproduction and
growth, and indirect effects such as through impacts on productivity of forage crops and
pastures (Maracchi et al., 2005). Higher temperatures are likely to generate more frequent
heat stress and higher water consumption (Turnpenny et al., 2001), which may affect
currently warm regions negatively. In cooler regions, milder cold periods will probably be be
beneficial due to reduced feed requirements, lower energy costs and increased survival
(Maracchi et al., 2005). Further benefits may arise from longer growth seasons in these
regions. While droughts may reduce productivity of grasslands in some areas, the overall
productivity of European grasslands is expected to increase (Kammann et al., 2005), making
possible stocking rates higher (Parsons et al., 2001), given that nutrient and water supply is
sufficient (Lüscher et al., 2004)
1
1
Found in Olesen (2006)
.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
15
In general, there is a relatively large literature on the impacts on crop yields, but significantly
less on the impacts on livestock and fruit and vegetables. The estimated changes in these sub-
sectors are, therefore, very rough. For all regions, impacts of climate change on the sub-sector
called “other crops” are based on the preliminary results of the PESETA Project (2007). The
PESETA study used the crop model DSSAT to quantify the physical impacts of climate
change on agriculture. DSSAT includes models that simulate phenological development and
growth in response to environmental factors and management. The PESETA agriculture case
study modelled European crop yield changes due to climate change for the years 2020 and
2080. The results of the modelling were presented in maps, which form the basis for the
approximations on yield changes for the sub-sector “other crops” at country level in our
study. Moreover, we calculate the weighted average impact in each region. When a country is
not distinguishable at the map, such as Malta, we use the average yield change for the rest of
the countries in the region. The PESETA Project uses two global scenarios, A2 and B2, from
the IPCC’s SRES. The present study uses the results based on the A2 scenario, expecting a
global warming of 3.1 degrees for the period 2071-2100 compared to the control period 1961-
1990. For the other sub-sectors, livestock, fruit and vegetables, different studies have been
used in the different regions. In the Baltic States, agriculture contributes 15 percent to total
output. Compared to the rest of Europe, those sub-sectors vulnerable to climate change (fruit,
vegetables and other crops) constitute a relatively high share of the agricultural sector (66
percent). According to the PESETA
2
Based on this result we assume that the productivity of fruit and vegetables in Estonia, Latvia
and Lithuania increase by eight percent by the year 2080, under a regional temperature
increase of 3°C and precipitation increase of 13 percent. Expected changes in fruit and
vegetables in Poland are based on a study by Stuczyiñski et al (2000). They used two models.
We chose the results of the model that gives the climate change results more close to the
climate scenario used for the rest of the agricultural sector. The GISS model predicts an
increase in mean annual temperature by 4°C (winter +5°C, summer +2-3°C) and 15 percent
increase in precipitation, but the study does not report at what time in the future. It is expected
that potato yields decrease by 20-30 percent (Stuczyiñski et al., 2000). On this basis, we
assume that fruit and vegetable yields will decrease by 25 percent by 2080 in Poland.
study, the yield might increase by up to ten percent by
the 2080s relative to the period 1961-1990 in Estonia, Latvia and Lithuania while expected
yields in Poland varies within the country. The results show a decrease between zero and five
percent in central parts of Poland, while other parts exhibit a positive change of yield. The
changes in fruit and vegetables in Estonia, Latvia and Lithuania are taken from a study by
Karing et al (1999). This study focuses on Estonia, but it is assumed here the same holds for
Latvia and Lithuania. A potato dynamic model was used to calculate possible yields under
climate change, and we assume the impacts on potatoes are also applicable to other vegetables
and fruits. It was found that the two climate change scenarios with temperature increases
between 1°C and 3°C will increase potato yields between 6 and 8 percent by 2050 (Karing et
al., 1999).
Studies of expected changes in livestock production due to climate change in this region are
scarce. However, Stuczyiñski et al. (2000) expect poultry production to increase substantially,
as the cost of keeping animals will decrease as a result of in increase in productivity of
grasslands and an extension of grazing periods. Based on this, we assume that the output of
livestock will increase by 5 percent by 2080 in the Baltic region, as a rough estimate. The
2
The results of the PESETA study are based on a relatively optimistic assumptions about adaptation,
where farmers can use as much additional irrigation water and/or fertilizers as wished, without any
constraint.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
16
overall effect of climate change on total output of agriculture in the Baltic region turns out
slightly negative. This is because of the dominance of Poland, which contributes
approximately 85 percent of total output of agriculture in the region. The expected negative
impacts in the country outweigh the positive effects in the rest of the region.
Agriculture contributes only 6 percent of total output on the British Islands, but a substantial
decrease in yields is expected. In southern parts, the decrease can be as high as 30 percent
(PESETA, 2007). Wolf (2002) did a study on potato simulation models under climate change
in Europe, including a site in Oxford, England. The use of two climate scenarios from Hadley
Centre for the period around 2050 resulted in zero to slight increase in Oxford, both with and
without irrigation (Wolf, 2002). Combining this with the expected negative effects on crop,
we assumed no change in the total output of fruit and vegetables. For estimates on livestock,
we use studies by Parsons et al (2001) and Turnpenny et al (2001). Using IPCC 92 IS92
scenarios for 2050 the result of their studies are the following; the profitability of grazing
systems is likely to be small but positive, while the effect on intensive systems is likely to be
small, but negative. As a result, we assume no change in livestock. Given the relative
importance of crops in the agricultural sector in the region, we expect total output to decrease
by approximately 8 percent.
In the region Central Europe East, PESETA has projected up to 30 percent decrease in yields
in some areas of Bulgaria, and a positive change in the other countries. In most areas of
Hungary, Slovakia and Romania the increase could be as much as 15-30 percent. The share of
agricultural output in total output in the region is with 18 percent relatively high, but the
importance of this sector within the region differs. In Bulgaria the economy highly depends
on agriculture, as this sector contributes to 30 percent of total output. This means that a
projected decrease in yields up to 30 percent has a strong effect on GDP in Bulgaria. Our
estimates on fruit and vegetables are based on the same study as used for the British Islands
(Wolf, 2002), in addition to PESETA projections for crops. The British study of potato
simulation models under climate change in Europe included a site in Debrecen, Hungary. The
result was no change to a slight decrease (Wolf, 2002). This runs counter to the expected
effects on crops in other studies, and we therefore expect no change in fruit and vegetables in
Hungary, and assume the same will hold for Romania and Slovakia. Due to the expected
strong negative impact of climate change on crops in Bulgaria, we expect that there will be a
negative impact on fruit and vegetables as well. However, there is little research to build this
estimate on, and we therefore choose somewhat more moderate estimates for the impacts in
the other countries in this region than those for Bulgaria, namely..
For livestock, there are no studies from this region. However, it is assume the results from the
British Islands are applicable, which implies that climate change will not change the
productivity (Parsons et al., 2001; Turnpenny et al., 2001). Even though the expected impacts
in most of the countries in this region are positive, the total effect is negative because of the
relative size of the agricultural sector in Bulgaria.
Apart from an area in western Germany, Central Europe North can expect a positive change
in crop yields, especially in Austria, where the projected increase range from 15 to 30 percent
(PESETA, 2007). In a study on wine quality, using the SRES A2 scenario, it is predicted that
climate changes in cool climate regions like the Mosel Valley and Rhine Valley in Germany
could lead to more consistent vintage quality of wine and even ripening of warmer climate
varieties as temperature increase, at least to a certain point (Jones, White, Cooper, &
Storchmann, 2005). On this background, we assume better conditions for growing fruits and
vegetables in the region, and estimate an increase of 5 percent in total output by 2080. No
studies on expected changes in livestock production due to climate change has been found for
this region, so as for Central Europe East, we expect no changes.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
17
The changes in crop yields projected by PESETA vary throughout Southern Europe, from up
to 30 percent decrease in southern Greece to a 30 percent increase in Slovenia and southern
Italy (PESETA, 2007). Estimates of the changes in production of fruit and vegetables
resulting from climate change in the region point in different directions (Jones et al., 2005;
Viner, Sayer, Uyarra, & Hodgson, 2006). Viner et al (2006) expect negative impacts on
fruits, and decrease in the production of wine and olive oil, while Wolf (2002) calculated both
negative and positive effects on potatoes in a study site in Italy. In this study, we assume a 1
percent decrease in total output of fruit and vegetables for the region, except in the case of
Slovenia. Based on the results of the PESETA Project (2007), which expect an average of
22,5 percent increase in crop yields, we assume a 5 percent increase in fruit and vegetables.
Given that a 15 percent decrease in precipitation is expected (PESETA, 2007), it seems
reasonable to expect that the productivity of grasslands will decrease, with a negative effect
on livestock production (Parsons et al., 2001). As a result, we expect the cost of keeping
animals will increase, taking into account the shortage of irrigation water available, and
estimate a 5 percent decrease for countries in the region, apart from Slovenia. The overall
effect of climate change on the agricultural sector is expected to be negative, given that the
negative impacts in the rest of the region outweighs the positive impacts expected in Slovenia.
According to PESETA (2007), the change in crop yields in Central Europe West will most
likely turn out negative. A study by Jones et al (2005) on wine quality found that for some of
the grape varieties in France, the climate was already too warm for the predicted optimum.
Considering the importance of wine production in this region, we assume that climate change
will affect the productivity of fruit and vegetables negatively in the region, and assume an
estimated 1 percent decrease in total output of this sub-sector. Similar to the British Islands,
no change in the livestock sub-sector is assumed. The effect on total agricultural output is
then expected to be negative.
The changes estimated for the productivity in agriculture on the Iberian Peninsula are
negative and large. As much as 15-30 percent decrease in crop yields is expected in most of
the region (PESETA, 2007). In a study from the Mediterranean area, a 2 percent decrease of
potato yields was estimated in Portugal, while the estimate for Spain was a 9 percent increase,
given a global temperature increase of 2 degrees by 2060 (Giannakopoulos, Bindi, Moriondo,
LeSager, & Tin, 2005). However, these estimates presuppose that the additional water
requirements can be met. Given that the region is already scarce of water, and the negative
projected crop yields estimated by PESETA, we assume a 5 percent decrease in total output
of fruit and vegetables in Portugal, and no change in Spain. According to our estimates, the
Iberian Peninsula will be the far most affected region in Europe. We assume that the direct
impact on agriculture is an 11 percent decrease. In both Spain and Portugal, agriculture
accounts for 9 percent of total output, and local communities and areas highly dependent on
agriculture are expected to be hit hard. Viner et al (2006) found that livestock production is
likely to be negatively affected because of increasing costs of keeping animals. Based on this,
we assume a 1 percent decrease in total output of livestock.
The agricultural sectors in the Nordic Countries are likely to benefit from improved
conditions. A Danish study projects an increase in the mean yields of winter wheat between
21 and 37 percent, depending on methods for scenario application (Olesen et al., 2007). The
projected crop changes in PESETA are 15-30 percent higher in the 2080s compared to
baseline (PESETA, 2007). Potato yields have been estimated to increase by more than 50
percent in Finland, given a temperature increase of 2 degrees, increased precipitation by 8
percent and CO2 concentration of 515 ppm (Carter, Saarikki, & Joukainen, 2000). This result
is in line with the strong increases projected by Wolf (2002). In this study it is assumed that
the output of fruit and vegetables in the Nordic countries increases by 50 percent by 2080.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
18
Animal production is expected to be less affected by climate change, but nevertheless, the
impact is expected to be positive due to indirect effects (Hildén et al., 2005). We assume a
positive impact on total output of livestock of 5 percent in the region. All in all, the
agricultural sector in this region appears to benefit from climate change.
Table 4.3 summarizes the assumptions about the impacts to the agricultural sectors in the
present study, indicating ranges for changes in the productivity.
Region
Expected changes
Fruit&vegetables
Other crops
Livestock
Total
Baltic States
---
+
+
-
British Islands
0
--
0
-
Central Europe East
-
-
0
-
Central Europe North
+
+
0
+
Southern Europe
-
-
-
-
Central Europe West
-
-
0
-
Iberian Peninsula
-
---
-
--
Nordic Countries
+++
+++
++
++
Table 4.3: Evaluation of change on agricultural sub-sectors by region: 0: no change,
single: 1-10 percent, double: 11-20 percent, triple: more than 20 percent
4.3 Forestry
The direct economic impacts of forestry in Europe today are relatively small (see Table 4.4)
although the area covered by forest is generally high in all regions except for the British
Islands. To what extent the forest is used for production varies, from a 5 percent share in
Central Europe North, to 91 percent in the region Central Europe West.
In addition to timber production, forests play an important role in protecting biodiversity,
prevent soil erosion and protect infrastructure and settlements from natural hazards, such as
rock fall and avalanches (Lexer et al., 2002). Forests also provide important recreational
areas, they are important to the cultural identity, for regulation of ecological systems, etc. The
impact of climate change on forest is expected to follow the same pattern as productivity in
agriculture. Due to a decrease of current tundra area (White, Cannell, & Friend, 2000) and
higher tree lines in mountainous areas (Moen et al., 2004), forest area is expected to increase
in the Northern areas of Europe. In terms of how much forestry contributes to total output,
this means forest area is expected to increase in economies where forestry is already relatively
important. In southern areas on the other hand, number of droughts and fire risk is expected to
increase, both in terms of length of season and severity (Arnell et al., 2005). As a result,
forested areas in southern areas are likely to contract (Metzger et al., 2004). The expected
causes are changes in temperature, precipitation and increased CO
2
and ozone, similar to the
effects in the agricultural sector (see Chapter 4.2). Changes in species distribution will be
commented below for each specific region. The estimated climate related changes in the
forestry sector in the GRACE_adapt model are based mainly on the expected changes in net
primary productivity (NPP) of biomass in a study by Fronzek and Carter (2007). NPP of
biomass is a less than perfect economic estimate for the total output of forestry, as increased

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
19
growth rates may reduce the quality of the timber (Broadmeadow, Ray, Sing, & Poulson,
2003). However, we assume it can be a suitable indication, in terms of direction and size of
the expected change. Fronzek and Carter (2007) used a model called the Miami Model, which
is based on an empirical relationship between long-term temperature and precipitation and
measurements of NPP of natural vegetation, to project mean estimates of percentage change
in NPP of biomass between 1961-1990 and 2071-2100 on the basis of the SRES emission
scenarios A2 and B2. We use the estimates from the A2 simulation. The largest changes were
found in The Nordic region, where the expected increase in the most northern areas is above
40 percent while the largest negative impact is in Spain, where the decrease might be more
than 20 percent (Fronzek & Carter, 2007). These findings are similar to findings from other
studies (Alcamo et al., 2007).
Region Share
of
output
Volume shares Value of product removals
Of total land
Used for
production
3
Industr.
roundwood
Wood fuel
Non-wood
products
Baltic States 0.64 35 54 93 5 2
British Islands
0.03
11
44
80
1
19
C Europe East 0.45 30 41 85 7 8
C Europe North
0.09
34
5
71
15
13
Southern Eur
0.07
34
28
65
7
29
C Europe West
0.10
27
91
72
7
21
Iberian Penin
0.20
37
21
61
4
35
Nordic countries
0.58
52
78
85
5
10
Table 4.4: Overview of volume and value of the forestry sector in percentages.
Sources: (Dimaranan, 2006; FAO, 2005).
The share of total output from the forestry sector is relatively high in the three Baltic States
Estonia, Latvia and Lithuhania . The risk of fire, pest outbreaks and wind damages are
expected to increase (Arnell et al., 2005; Kellomäki et al., 2000). In Poland, expectations of a
decrease in the supply of coniferous and increase of deciduous species are reported
(Kellomäki et al., 2000). There are differences throughout the region: the northern parts of the
region can expect biomass to increase, while southern parts should expect a decrease. In total,
the impact of climate change on NPP is expected to be positive (Fronzek & Carter, 2007).
In the British Islands, a change in species composition is expected, with an increase of
coniferous species in northern parts and a decrease of deciduous species in southern parts.
Southern parts are also likely to face higher risk of drought and fire, while northern parts
might expect an increase in the risk of wind and snow damages (Kellomäki et al., 2000). NPP
of biomass is expected to increase in northern parts and decrease in southern parts of the
region (Fronzek & Carter, 2007). The estimated total effect is slightly positive. This is
supported by other research as well. Without including the CO2 effect, Broadmeadow, Ray
and Samuel (2005) found that broadleaf timber productivity is likely to benefit from climate
change across the majority of the British Islands. In a case study from a rural estate in
3
Share of forested land

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
20
Scotland, Viner et al (2006) came to the similar conclusions; with an increase in expected
annual growth.
In Central Europe East, forestry is a relatively large contributor to total output in Slovakia
and Bulgaria. According to an expert assessment on impacts of climate change on forests in
Europe, increased temperature and precipitation is likely to decrease timber supply from
Norway spruce, European beech and Oak in Slovakia (Kellomäki et al., 2000).
Central Europe North can expect higher risks because an increase in pest, bark beetle, forest
fires and wind damages is expected (Kellomäki et al., 2000). A negative impact on
productivity is expected in Germany (Fronzek & Carter, 2007; Lasch, Lindner, Erhard,
Suckow, & Wenzel, 2002; Viner et al., 2006), but this is outweighed by the expected
increases in productivity in Switzerland and Austria (Fronzek & Carter, 2007; Viner et al.,
2006). The total effect on the forestry sector in the region is expected to be positive.
In Southern Europe, one expects a negative impact of climate changes on the forestry sector
in all countries except Slovenia. Increased frequency of droughts and fire risk are expected
(Giannakopoulos et al., 2005). As a consequence, the NPP of biomass is will decrease
(Fronzek & Carter, 2007).
In Central Europe West, the risk of fire is likely to increase. NPP of biomass is expected to
decrease in France and Luxembourg, but stay approximately the same in Belgium and the
Netherlands (Fronzek & Carter, 2007). The total effect is, then, negative.
The Iberian Peninsula is the region most negatively affected by climate change in Europe,
and the forestry sector is not an exception. Droughts and the risk of fire is likely to increase,
and NPP of biomass is expected to decrease throughout the region, and by more than 20
percent in southern Spain (Fronzek & Carter, 2007; Viner et al., 2006).
The projections for the Nordic Countries, on the other hand, show benefits for the forestry
sector from climate change. This is also a region where forestry is relatively important
compared to the rest of Europe, particularly in Finland. The tree line is expected to expand
towards higher altitudes and higher latitudes as the temperature increases. The region is likely
to see a decrease in coniferous species, but an increased supply of deciduous species
(Kellomäki et al., 2000; Moen et al., 2004; Sykes & Prentice, 1996). The productivity of
forests is likely to increase, especially in northern parts of the region (Fronzek & Carter,
2007; Kellomäki et al., 2000; Sonesson, 2004).
Region
Expected changes
Baltic States
+
British Islands
+
Central Europe East
-
Central Europe North
+
Southern Europe
-
Central Europe West
-
Iberian Peninsula
--
Nordic Countries
+++
Table 4.5: Main impacts per region in the forestry sector.
0: no change, single: 1-10 percent, double: 11-20 percent, triple: more than 20 percent.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
21
4.4 Fisheries
Fisheries contribute less than a half percent to total output in all regions except Southern
Europe (see Table 4.6). In Greece, the fishery sector contributes more than 7 percent to total
output. The impacts of climate change differ in many respects between aquaculture, or fish
farming, and commercial fishing. While impacts on fish farming depend to a large extent on
the ability to control pests, commercial fishing is vulnerable to changes in the location of fish
stocks. Hence, the impacts to the sector also depend on the contribution to total output from
fish farming. This contribution varies from 2 percent in Central Europe North to 51 percent in
the Nordic Countries.
Although there is a general opinion that fisheries will be affected significantly by climate
change, the knowledge of biodiversity and ecosystem impacts is still limited, and it is difficult
to predict in what directions the changes will go. Socio-economic costs related to climate
change are also uncertain, but they may become substantial for enterprises and communities
that highly depend on coastal and marine ecosystems (Alcamo et al., 2007). Because of these
limitations we do not base the GRACE_adapt model on specific relationships between
climate indicators and economic activity, but we include assumptions about stock changes for
illustrative purposes. These are presented in Section 5.3.
Region
Percent
of GDP
Fresh-
water and
diad.
Demer-sal
marine
fish
Pelagic
marine
fish
Crustacea
ns and
molluscs
Commerci
al fishing
Aqua-
culture
Baltic States
0.1
12
12
70
6
90
10
British Islands
0.1
16
27
37
19
77
23
C Europe East
0.1
89
1
6
4
85
15
C Europe North
0.1
23
22
45
10
98
2
Southern Europe
0.8
8
25
24
33
96
4
C Europe West
0.1
4
26
44
24
77
23
Iberian peninsula
0.4
3
26
43
24
95
5
Nordic countries
0.2
12
49
36
4
49
51
Table 4.6: Overview of fisheries in each region in percentages.
Sources: GTAP(Dimaranan, 2006), Eurostat (2005a) and FAO (2007).
Despite limited knowledge, the sensitivity of fish stocks to changes in sea conditions seems to
be undisputable, such as the relation between temperature and stock production (Dutil &
Brander, 2003), the distibution of species (Hiscock, Southward, Tittley, & Hawkins, 2004;
Roessig, Woodley, Cech, & Hansen, 2004) and recruitment (Clark, Fox, Viner, & Livermore,
2003). Distributions of North Sea herring have been found to respond markedly to increases
in sea temperatures (Toresen, 2000), and climate change is predicted to drive species ranges
towards the poles (Perry, Low, Ellis, & Reynolds, 2005, Drinkwater, 2005).
Aquaculture is a part of the fisheries sector and accounts for approximately 30 percent of the
total output (EC, 2006). Expanded geographic distribution and range due to warmer sea
temperatures is likely to give opportunities for new species (Beaugrand & Reid, 2003), but
ecosystem changes may have negative impacts, such as outbreaks of harmful algae and
decreased dissolved oxygen events, which will increase operational costs. Further, increased
storm-induced damage on facilities and equipment will increase capital cost (Alcamo et al.,
2007).

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
22
4.5 Energy
Both energy demand and energy supply will be affected by climate change. A part of the
energy demand is due to regulation of indoor temperature, both for space heating and space
cooling. On the supply side, renewable energy sources for electricity are based on climatic
conditions, such as wind speed for wind power and run-off for hydro power. Thermal power
plants, moreover, require cooling water, which may be restricted at high temperatures or
droughts. We assume that there are no direct effects of climate change on the extraction of
fossil fuels.
The five energy sectors in GRACE_adapt (crude oil, gas, coal, refined oil, electricity)
contribute between 2 and 6 percent to total GDP in the eight EU regions. The contribution is
highest in the Baltics and in Central Europe East, although the oil and gas sector are important
in the Nordic countries and on the British islands as well due to their affluence on fossil
resources. The pattern indicates the importance of energy to the lower-income European
economies, as the relative contribution from electricity and refined oils increase with lower
income levels.
4.5.1 Demand
Temperature and energy demand are linked together in a rather intuitive way; higher
temperatures lower energy demand for heating purposes in winter, and increases energy
needed for cooling devices, such as air conditioners, in the summer. These effects are referred
to as the heating effect and the cooling effect (De Cian, Lanzi, & Roson, 2007; Tol, 2002).
Seasonal fluctuations in energy demand are closely related to temperature, but the sensitivity
differs across energy carriers. Variations related to heating purposes affect the consumption of
refined oils, gas and electricity, but differently and to different extent across regions. Cooling
affects only the demand for electricity (Kirkinen et al., 2005).
In a study on the effect of climate change on energy demand, Bigano, Bosello and Marano
(2006) found that an increase in temperature affects the demand for different energy carriers
differently in households and in production sectors. The so-called temperature elasticity of
energy demand (percent change in demand due to one percent change in temperature) is, in
general, small and not statistically significant in service and industry sectors. Households, on
the other hand, do respond to temperature changes. The response in energy demand to
changes in temperature depends on the temperature level in the region and we assume that
this is proportional to the residents’ share of total final energy consumption (TFC) in the
region.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
23
Region Hot/cold Residents share of TFC for different energy
vectors in percentages
Electricity
Gas
Oil products
Baltic States
Cold
25.5
29.3
5.1
British Islands
Cold
33.6
58.5
4.7
Central Europe East
Cold
27.8
35.0
2.8
Central Europe North
Cold
27.4
45.6
15.7
Southern Europe
Hot
24.0
41.0
11.3
Central Europe West
Cold
32.7
38.1
10.1
Iberian Peninsula
Hot
26.3
17.4
6.7
Nordic Countries
Cold
29.6
21.1
4.9
Table 4.7: Residents’ use of total final energy consumption (TFC).
Source: 2005 Energy Balances in thousand tonnes of oil equivalent (ktoe) on a net
calorific value basis (IEA, 2005).
A study of the impact of temperature change on residential energy demand for a series of
energy goods reveals that the cooling effect outweighs the heating effect in hot countries and
that the heating effect outweighs the colling effect in cold countries in terms of electricity
demand (De Cian et al., 2007). For other energy sources, such as coal and gas, the effect of an
increase in temperature on consumption is negative. Moreover, the effect was found to be
similar for hot and cold countries. In order to estimate expected changes in energy demand we
use the average elasticities for hot countries for Southern Europe and the Iberian Peninsula,
and the average for cold regions for the other regions. The elasticities are shown in Table 4.8,
which refer to percent change in temperature measured in Farenheit. The temperature
elasticities are based on seasonal temperatures. To use them directly, we have to assume that
the temperature changes equally throughout the year in each region, without implications for
seasonal variations. Studies which include seasonal variations indicate that temperatures will
increase most during summers in southern parts of Europe, and during winter in northern
parts (PESETA, 2007). If so, our study underestimates the impacts on energy demand.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
24
Region
Temperature elasticities for different energy vectors
4
Electricity
Gas
Oil products
Baltic States
-0.38
-2.60
-1.43
British Islands
-0.38
-2.60
-1.43
Central Europe East
-0.38
-2.60
-1.43
Central Europe North
-0.38
-2.60
-1.43
Southern Europe
1.00
-2.80
-1.46
Central Europe West
-0.38
-2.60
-1.43
Iberian Peninsula
1.00
-2.80
-1.46
Nordic Countries
-0.38
-2.60
-1.43
Table 4.8: Temperature elasticities for the residential sector.
Source: De Cian et al (2007).
Both the Baltic States and Central Europe West are defined as cold regions, and energy
demand is expected to decrease due to the heating effect.
On the British Islands, higher temperatures will decrease demand for space heating in winter,
spring and autumn. A reduction in demand of 5-10 percent for fossil fuels and 1-3 percent for
electricity due to a 2 degree warming by 2050 was estimated (Smith & Hitz, 2003). Likewise,
a 10-15 percent increase in electricity demand due to increased air conditioning was estimated
(LCCP, 2002). Our study uses the elasticities estimated by De Cian et al (2007). An expected
increase in demand for electricity for cooling purposes during summers is outweighed by the
decrease of electricity demand for heating purposes the rest of the year and the total effect are
expected to be negative.
In Central Europe East, wintertime heating is expected to decrease by 6-8 percent on the
average by the period 2021-2050 (Vadja, Venäläinen, Tuomenvirta, & Jylhä, 2004), while the
summer cooling requirements are expected to increase (Arnell et al., 2005). However, the
baseline use of energy for cooling purposes is low, and the expected increase is outweighed
by decreased heating demands. Based on the temperature elasticities estimates in Table 4.8
and expected temperature increases between 12 and 15 percent, the total energy demand is
expected to decrease by approximately 4 percent.
In Central Europe North, energy demand is expected to shift from winter heating demand to
summer cooling demand. In a study of the response of climate changte on energy demand in
buildings in Switzerland, energy demand for heating was projected to decrease by 33-44
percent, while energy for cooling increased by up to 223-1050 percent, depending on type of
building (Frank, 2005). However, baseline cooling demand is not known. The expected
temperature increase varies between 12 to 14 percent, and total energy demand is likely to
decrease by more than four percent by the period 2071-2100. Energy demand is expected to
shift from winter heating demand to summer cooling demand.
In Central Europe West, electricity demand is expected to increase due to higher summer
cooling requirements (Arnell et al., 2005).
4
The elasticities are defined as the percentage change in demand induced at one percent increase in
temperature In this table, temperature is measured in Fahrenheit. Elasticities depend on baseline
temperatures. Hot countries tend to have higher elasticities because one percent represents a higher
absolute increase the warmer the baseline temperature is (Roson, Bosello, & De Cian, 2007).

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
25
Electricity demand in Southern Europe will be affected by an increase in summer space
cooling needs (Arnell et al., 2005; Cartalis, Synodinou, Proedrou, Tsangrassoulis, &
Santamouris, 2001; De Cian et al., 2007). Yet, total energy demand is expected to decrease by
approximately four percent because of reduced demand during wintertime.
As for Southern Europe, electricity demand on the Iberian Peninsula is expected to increase
(Arnell et al., 2005; De Cian et al., 2007). The demand for energy to space heating during
winter is likely to decrease, which implies a decrease in the demand for fossil fuels. The total
effect is a slight decrease, by less than one percent, in total energy demand by 2071-2100.
Despite the low expected change in energy demand, it is important to keep in mind that this is
averaged throughout the year.
In the Nordic Countries, total energy demand is likely to decrease due to decreased heating
demand (De Cian et al., 2007; Tammelin et al., 2002; Venäläinen et al., 2004). Despite the
relatively high temperature increases, varying between 13 and 19 percent, the effect on energy
demand is among the lowest of all regions at no more than a decrease of 1,2 percent. This is
due to the fact that the share of total final consumption by residents is relatively low.
Region
Expected changes
Oil
Gas
Electricity
Total
The Baltic States
- - -
-
The British Islands
-
--
-
-
Central Europe East
-
--
-
-
Central Europe North
-
--
-
-
Southern Europe
-
--
+
-
Central Europe West
-
--
-
-
The Iberian Peninsula
-
-
+
-
The Nordic Countries
-
-
-
-
Table 4.9: Main impacts per region on energy demand 0: no change, single: 1-10
percent, double: 11-20 percent, triple: more than 20 percent.
4.5.2 Electricity generation
Most of the electricity generated in Europe is thermal power, but the share varies from 37
percent in the Nordic region to 95 percent in the Baltics. Among renewable sources, hydro
power dominates in most regions, with contributions from 2 percent on the British Islands to
56 percent in the Nordic region. Bio power contributes up to 5 percent (in the Nordic region)
and wind power up to 7 percent (in Iberia). Generation of all electricity is subject to climate
change, but for highly different reasons and to widely different extent.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
26
Region
Hydro power Bio power Wind Fossil-fired
thermal
Baltic States
4
1
0
95
British Islands
2
3
1
94
Central Europe East
13
1
0
86
Central Europe North
13
3
4
80
Southern Europe
13
2
1
84
Central Europe West
8
2
0
90
Iberian Peninsula
8
1
7
83
Nordic Countries
56
5
2
37
Table 4.10: Shares of total output in the electricity sector in percentages.
Source: IEA(2005).
The hydro power potential is closely linked to changes in precipitation. A study by Lehner et
al (2005) of the impacts of climate change on hydropower potential in Europe found that most
countries will be affected. Our estimates of the real change in hydropower potential are based
on the reported hydro power production statistics from the International Energy Agency (IEA,
2005). Lehner et al (2005) base their estimates of changes on production in the period 1998 to
2000. These changes are assumed to apply similarly for the actual production in 2005, hence
assuming no significant changes in production of electricity from the baseline of the study by
Lehner et al (2005) to the 2005 numbers from IEA. For Europe as a whole, gross hydropower
potential was, then, estimated to decrease by 6 percent by the 2070’s. Iberia is the most
negatively affected region, and is likely to face a reduction in the hydro power potential of 25
percent or more by 2070. For the Nordic countries, on the other hand, a strong increase is
projected (Lehner et al., 2005).
Bio power is becoming increasingly important in Europe, but there are few studies on the
impact of climate change on this source of energy. Generally, the supply of the raw material
can be expected to follow the direction and proportions of the productivity changes in
agriculture and forestry (Kirkinen et al., 2005). On this basis, we adopt the forestry results
from the study by Fronzek and Carter (2007) to biomass for electricity generation as well.
The baseline for this study is the period 1961 – 1990. Similar as for hydropower, we assume
that there has not been a significant change up to 2005, so the distribution of biopower across
regions reported by IEA (2005) apply directly.
Because of the difficulties in obtaining projections in changes of wind-speed under climate
change, it is hard to estimate the effect of climate change on wind energy potential. Moreover,
the share of electricity generated by wind power is relatively small in most regions, with an
exception of Iberia and Central Europe North. Our estimates are based on a study on climate
change impacts on wind energy resources in northern Europe (Pryor, Barthelmie, &
Kjellström, 2005).
Climate change is expected to reduce the availability of cooling water needed at thermal
power plants during summers, but no estimates of costs have been available. Offshore oil and
gas extraction may be negatively affected by climate change if the frequency of extreme
weather events increases (Arnell et al., 2005), while less ice cover opens new opportunities
for finding oil and gas, but possible impacts on extraction are ignored in this study.
The Baltic States get a relatively high share of total output from the electricity sector, but the
sector is not particularly climate sensitive, given that the electricity is mainly generated by the

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
27
use of fossil fuels, such as coal, oil and gas (IEA, 2005). Bio-, hydro- and wind power
potential are all expected to increase under climate change (Fronzek & Carter, 2007; Lehner
et al., 2005; Pryor et al., 2005). Given the small market shares from these sources, the total
impact on the electricity sector in the region is expected to be just above 1 percent. Latvia is
an exception in this region, with a high share of hydropower. Due to an expected increase of
24 percent by 2070 (Lehner et al., 2005), our estimated output increase of the electricity
sector is more than 16 percent.
The electricity sector in the British Islands is highly dominated by gas, coal and nuclear
driven power plants (IEA, 2005), and the expected positive impacts on hydro- and bio power
potential is not expected to have a significant effect on the sector.
Central Europe East has the highest share of total output from electricity, and both bio- and
hydropower potentials are expected to decrease (Fronzek & Carter, 2007; Lehner et al., 2005).
The total effect is a decline of approximately 4 percent, the highest decrease of all regions.
The electricity sector in Central Europe North is expected to be negatively affected. The
hydropower potential has decreased in all countries of the region by 2070 (Lehner et al.,
2005). An expected increase in bio power production in Austria and Switzerland is
outweighed by the decrease in Germany (Fronzek & Carter, 2007). The share of electricity
from wind power is more important in Germany that in most other countries in Europe, with a
4 percent share (IEA, 2005). Due to the expected increase in wind energy potential in
northern Europe (Pryor et al., 2005), we find it reasonable to assume an increase in Germany
as well.
In Southern Europe, the overall effect of climate change on the electricity sector is expected
to be negative. Hydropower potential is expected to decrease in all countries, and as much as
by 22 percent in Italy (Lehner et al., 2005), which is the largest hydropower producer in the
region. The potential for bio power production is also expected to decrease in all countries
except Slovenia (Fronzek & Carter, 2007). As wind speed is expected to decrease (Räisänen
et al., 2004), we assume the impact on wind power potential will be negative. The net effect
on electricity production is estimated to a decrease of 3,5 percent.
Both bio- and hydropower potential is expected to decrease in Central Europe West (Fronzek
& Carter, 2007; Lehner et al., 2005), because of lower precipitation and declining forest
biomass.
On the Iberian Peninsula, precipitation will decrease, which has a negative impact on
hydropower potential (Arnell et al., 2005; Lehner et al., 2005). A decrease in biomass is also
expected, particularly in Spain (Fronzek & Carter, 2007). Mean windiness is expected to
decrease as well in the Mediterranean area (Räisänen et al., 2004), and reduces the wind
power potential in this region, which has the highest share of wind power in Europe.
The Nordic Countries will benefit from enhanced growth of biomass and increased
precipitation (Fronzek & Carter, 2007; Lehner et al., 2005). More than 60 percent of
electricity is from bio- and hydropower in this region. Climate change therefore implies a
positive effect on the electricity sector. Denmark is the country in Europe with the highest
share of electricity generated from wind power (IEA, 2005). It will benefit from an expected
increase of wind in the region, both on-shore and off-shore (Kirkinen et al., 2005; Pryor et al.,
2005). The total effect on the electricity sector is expected to be as much as 15 percent.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
28
Region
Expected changes
Hydro power
Bio power
Wind
Other
Total
Baltic States
++
+
+
0
+
British Islands
+
+
0
0
+
Central Eur East
---
-
0
0
-
Central EurNorth
--
+
+
0
-
Southern Europe
---
-
-
0
-
Central Eur West
--
-
0
0
-
Iberian Peninsula
--
--
-
0
-
Nordic Countries
+++
+++
+
0
++
Table 4.11: Main impacts on renewable electricity generation by sub-sector.
0: no change, single: 1-10 percent, double: 11-20 percent, triple: more than 20
percent
4.6 Manufacturing industries
The impact by climate change on the industry sector is mostly indirectly, such as through the
supply of raw materials, intermediates, transport etc. This will be taken care of by the
GRACE_adapt model, as this includes cross-over supplies.
4.7 Tourism
The contribution of tourism to total economic activity varies throughout Europe, but plays a
more dominating role in the southern parts than elsewhere. A change in tourism will affect
many economic sectors. We have not found detailed information on what sectors and to what
extent they depend on tourism. To keep it simple, we use the numbers from Statistics Finland
(2008) where it is assumed that the contribution to GDP from tourism splits into the transport
and the service sectors, with 30 percent to transport and 70 percent to services. To compare
the size of tourism in all countries, we apply the satellite accounting tables on tourism from
World Travel & Tourism Council (WTTC), where all the countries in our model are
represented.
Region Contribution
to GDP
Share of arrivals Season
Domestic
International
Summer
Winter
Baltic States
4
67
33
65
35
British Islands
7
77
23
60
40
Central Eur East
4
57
43
64
36
Central EurNorth
6
68
32
60
40
Southern Europe
8
54
46
66
34
Central Eur West
7
60
40
86
14
Iberian Peninsula
12
56
44
64
36
Nordic Countries
5
73
27
68
32
Table 4.72: Overview of the tourism sector (in percentages).
Sources: Eurostat (2005b) and TSA (2008).

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
29
The output of tourism is calculated as the sum of direct and indirect gross domestic product
and employment associated with travel and tourism consumption. Then, 12 percent of GDP
on the Iberian Peninsula is generated by tourism, whereas the contributions in Malta and
Cyprus are above 15 (TSA, 2008). The summer season, defined by us as ranging from May to
October, is the most important season in terms of both arrivals and nights spent in different
accommodation facilities, in all regions. On the other hand, the relative number of domestic
and foreign tourists varies, both within and across regions. However, there are more domestic
tourists compared with foreign tourists in terms of arrivals in all regions (Eurostat, 2005b).
The total amount spent on travel and tourism will probably not be affected by climate change,
but there may be significant changes in destinations (Berrittella et al., 2006). Climate is, in
many cases, an important factor when tourists make their choice of destination, including the
climate of the source regions (Giannakopoulos et al., 2005). Changes in the length and quality
of tourist seasons (such as beach holidays or winter sports activities) are likely to have
implications for the competitiveness of destinations and consequently the profitability of
tourism enterprises (Simpson, Gössling, Scott, Hall, & Gladin, 2008). Countries closer to the
poles may attract more foreign tourists during summer. At the same time, their own citizens
may become less attracted to destinations in southern countries (Amelung, Nicholls, & Viner,
2007; Hamilton, Maddison, & Tol, 2005). A shift towards tourist destinations at higher
altitudes is expected for these reasons (Hamilton & Tol, 2007; Simpson et al., 2008). If so,
the countries most dependent on tourism today will “suffer” the most from climate change in
the future. However, an important aspect of tourism is its dominating role in some local
communities. Even though tourism may become beneficial to a country or region, changes
may become negative in smaller tourist based communities with severe negative impacts.
In order to estimate expected changes in tourism due to climate change, we use the predicted
tourism flows by Hamilton et al (2005) for 2025 and a 1 degree global temperature increase
as an indication on change in tourism by 2070 using the PESETA predictions of climate
change. There are several reasons to consider the estimates with caution, as the knowledge
about relationships between climate and tourism are poor. To some tourist activities, the
linkage may seem obvious, but for other, such as event seeking or cultural tourism, it less
clear. Moreover, we interpret changes in departures as changes in domestic tourism. That is, a
five percent decrease in departures implies a five percent increase in domestic tourists, not a
reduction of tourism.
Tourism in all countries in Southern Europe and Iberia is expected to be negatively affected
by climate change. In all other regions, the increase in domestic tourism cancel out or
outweigh the decrease in international arrivals. Countries highly dependent on tourism, like
Malta and Cyprus, are likely to be hit hard, both due to the relatively important tourist sector
and the high share of foreign tourists.
The Baltic States are expected to become more attractive to international tourists. Apart from
Poland, domestic tourism will increase as well. The total regional effect on domestic tourism
is slightly negative due to the relative size of Poland, but the net effect is an increase in the
tourism industry. The negative effect of less international tourists is outweighed by the
expected increase in domestic tourists in the British Islands as well.
In Central Europe North, an increase in international tourists is expected in Austria and
Switzerland. This outweighs an expected decrease in Germany. This is because of the
assumption that localities at higher altitudes will become more attractive (Hamilton & Tol,
2007; Simpson et al., 2008). All countries in the region will also become more attractive to
own citizens, and the total net effect is an expected increase of 12 percent.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
30
Southern Europe is the region with the second highest share of GDP generated by tourism,
andthe share of international tourists is the highest in this region. This region includes the
only two countries where tourism contributes more than 15 percent of GDP, namely Malta
and Cyprus. In these countries 90 and 79 percent consists of international arrivals,
respectively. Climate change is expected to turn the region less attractive to foreigners. The
expected decrease in arrivals for the region is 17 percent. Domestic tourism is expected to
increase, and the net effect is a 5 percent decrease in tourism.
Expected changes
Domestic
International
Net effect
Baltic States ++ - ++
British Islands
++
-
++
Central Europe East
+
-
+
Central Europe North
++
+
++
Southern Europe
+
--
-
Central Europe West
+
-
+
Iberian Peninsula
+
--
-
The Nordic Countries
++
++
++
Table 4.83: Main impacts on tourism flows by region.
0: no change, single: 1-10 percent, double: 11-20 percent, triple: more than 20
percent
The region far most dependent on tourism is the Iberian Peninsula. Climate change is
expected to result in a lengthening and flattening of the season (Maddison, 2001), with a
resulting reduction in number of visitors. Hein (2007) estimated a 20 percent reduction in
tourist flow by 2080 compared to 2004. We use estimates far below this. Based on the
numbers by Hamilton et al (2005) and Eurostat data (2005b), we have estimated a decrease of
5 percent.
Similar to the Baltic States, it is assumed that also the Nordic Countries will become more
attractive to international tourists, especially during summer (DMI, 2006; RossbyCentre,
2006). The effect on winter tourism is, on the other hand, more uncertain, being dependent on
snow. As one of the projected effects of climate change is higher temperatures and less snow,
this might have a negative effect on winter tourism in some areas (Aaheim et al., 2008).
4.8 Transport
The sub-sectors rail and road transport constitute most of the transport sector. Apart from sea
and air transport, other sub-sectors include pipelines, auxiliary transport activities and travel
agencies. Estimating economic impacts on the transport sector is difficult, because little
research has been done. Some research has been conducted on the expected damage on
infrastructure due to climate change, but to the best of our knowledge, very little has been
done on estimating the cost/gains of delays/time saved due to a changing climate. However,
tourism is an integrated part of transport sector and we assume that 30 percent of tourism
contribution to GDP stems from transport. Based on this, we include a share of the expected
changes in tourism due to climate change as expected change on the transport sector.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
31
Region Share
of GDP
Tourism
share of total
transport
Sub-sectors share of sector total output
Sea
Rail and road
Air
Baltic States
7
19
11
84
5
British Islands
6
34
9
76
15
Central Eur East
6
28
9
84
7
Central EurNorth
4
45
14
72
15
Southern Europe
5
44
13
79
7
Central Eur West
5
41
15
67
18
Iberian Peninsula
5
67
7
76
17
Nordic Countries
8
16
31
54
16
Table 4.94: Overview of the transport sector in percentages.
Source: GTAP (Dimaranan, 2006).
Road and rail surfaces might be damaged as a result of temperature increases (Eddowes et al.,
2003; Wooller, 2003). Increased frequency of extreme weather might damage infrastructure
further, and also decrease passability by reducing performance and safety (Wooller, 2003).
Changed precipitation patterns, and intense downpours of rain may increase the probability of
landslides, and thereby undermine the stability of the infrastructure (DS, 2005). In colder
areas, increased temperatures may reduce snow cover. At higher altitudes and latitudes, snow
clearing is a demanding activity, and in some cases snow is not even cleared during winter
season. Less snow as a result of climate change may increase the utilization of roads in
mountainous areas during winter. Maintenance of roads may also be affected in areas where
the periods at which the temperature lies around 0 °C expand, because periods with freezing
and thaw causes extra damages to roads.
Region
Share of total output
The Baltic States
+
The British Islands
+
Central Europe East
+
Central Europe North
+
Southern Europe
-
Central Europe West
+
The Iberian Peninsula
-
The Nordic Countries
+
Table 4.105: Main impacts on the transport sector.
0: no change, single: 1-10 percent, double: 11-20 percent, triple: more than 20
percent.
On the other hand, there are indications that future climate change may increase the number
of landslides, and if this will be the case, this will affect the transport sector negatively
(Aaheim et al., 2008). Changes in passenger preferences and modal shifts as climate changes
are unpredictable. Studies from England predict that public transport may be perceived as

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
32
more uncomfortable compared to private vehicles when temperature increases, and dry
summers could lead to an increase in cycling (LCCP, 2002; Wooller, 2003). A study from
Norway estimates a shift from private to public transport as a result of more precipitation.
Also, in line with the study from England, people are expected to walk and bike more, and the
estimated total effect was a reduction in transportation costs.
Extreme events are also costly to the transport sector. One single flooding of the London
Underground had an estimated cost of £0.74 million due to passenger delays alone. For the
period between September 1999 and March 2004 the total cost amounted to approximately
£14.6 million (LCCP, 2005). Extreme events have, however, implications for many sectors
and activities, and we therefore base our estimate on the expected extreme events to each
sector as a share of the total expected cost of extreme events. This cost is imposed as an
annual “loss of real capital”, allocated to the different sectors.
Sea transport is likely to be affected as a result of less ice cover and more open waters, which
may open new international routes, such as the northern sea route, or make old routes more
effective. On the other hand, less ice cover increases the height of waves, which is
challenging to the sea transport (ACIA, 2004). Possible changes in wind speed, frequency of
storms etc. will also affect this sector.
The net economic effect of the physical expected changes due to climate change is hard to
predict. In our case, transport specific impacts are due only the changes driven by changes in
tourism. Thus, for other possible impacts, the positive and the negative impacts are expected
to be balanced in all regions.
4.9 Services
This sector includes public and private services. Consisting of many sub-sectors, the indirect
effects of climate change to the service sector may be substantial. However, direct effects are
more difficult to single out, except for the abovementioned impacts on tourism. This industry
is mainly dominated by small and medium sized enterprises, and is highly interlinked to the
rest of the economy, particularly the service sectors. On this basis we assume that changes in
the tourism industry will affect the service sector, such as in hotels and restaurants or the the
retail sector. Although changes in the service sector induced by changes in tourism might
appear small, one should keep in mind that the impacts are likely to be substantial for
enterprises and communities highly dependent on tourism.
Region
Share of GDP
Tourism share of total services
Baltic States
21
14
British Islands
32
15
Central Europe East
19
20
Central Europe North
37
11
Southern Europe
33
17
Central Europe West
36
13
Iberian Peninsula
28
29
The Nordic Countries 31 12
Table 4.16: Overview of the service sector in percentages.
Source: GTAP (Dimaranan, 2006).

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
33
4.10 Non sector-specific impacts
Apart from the impacts surveyed in Sections 4.2 through 4.9, climate change will have
economic consequences for a broad range of sectors which are not specific to certain sectors.
The typical example is extreme events, which may destroy capital and infrastructure
depending on where the activities are rather than what these activities are about. These
overarching impacts are often those most frequently associated with climate change impacts,
which besides extreme events, include sea-level rise and health impacts. Despite lots of
attention to these impacts, there are few numerical estimates on country level or regional
levels. We have made attempts to include these impacts in GRACE-adapt, but they are
loosely based on available estimates. Further comments on this are given in Sections 5.7 and
5.8.
5 Implementation of impacts functions
This section presents the impact functions by sector in the GRACE_adapt model implemented
for the eight EU regions. For most sectors, the same function is assumed to apply for all
regions. The parameters are estimated on the basis of the assessment of impacts of climate
change from underlying studies surveyed in Section 4. In some cases the country estimates
are based on a subjective evaluation of reference studies in other countries, applied to the
country in question. In other cases the country estimate is based on several studies of the
country. Thus, the impacts to a sector in each country is represented with only one estimate,
wich we refer to as the “observation”. Differences in impacts across regions arise as a result
of different changes in the climate indicators. In most cases we use changes in mean
temperature and annual precipitation. There are good reasons to question if impacts across
Europe can be described only with reference to changes in a few, simple climate indicators as
if other country specific charactheristics were insignificant. We do not belive so, , but on the
other hand, the alternative would be to apply regionally specific impacts functions without
any reference to studies of that region. Our choice may therefore be considered a result of
relatively scarce literarture. Recall, also, that most integrated models confine themselves to
mean temperature. The indicators we use to represent the climate are, nevertheless, very
rough when considering the information needed to establish reliable relationships between
climate and various impacts
The advantage of estimating a common impact function for each sector in all regions of
Europe is that the impact function becomes based on a broader set of observations than
possible if one impact function is made specifically for each sector in each country or region.
The general functions chosen here may, however, hide important specifics of a country, and
tends to moderate the impacts of climate change in general as compared to the results of
country specific studies. It is, however, difficult to tell whether “outliers” are likely to be real
or due to assumptions, and the choice was taken on the background of the relatively poor
quality of quantitative assessments of impacts.
There are, however, no reasons to hide the fact that the impact functions are extremely
uncertain, partly because of the uncertainty of the “data”, but also because the choices of
functions are taken primarily with the perspective of simplicity rather than realism. The
purpose here is to make clear what has been done, and thereby prepare the ground for a
prioritization of future improvements.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
34
5.1 Agriculture
Impacts are calculated indirectly on the basis of an estimated direct relationship between total
output in the agricultural sector and a combination of temperature, precipitation, income per
capita and population:
where P is precipitation in mm per year, T is temperature in °C, R is GDP per capita and N is
population. The same parameters apply for all regions (and provinces). Hence, impacts is a
result of changes in precipitation and temperature levels P and T. The estimates are shown in
Table 5.1.
To implement this relationship into GRACE, we assume the same sensitivity to the value of
natural resource input as to the estimated total production. Denote by V
s
the input of natural
resources (i.e. the value of land) in province s, and by X
s
total output in agriculture. Define α
s
= V
s
/X
s
and let α
is
= α
s
a
i
(i = 1,2) and α
is
= α
s
b
i
(i = 1,2). Then, the productivity function
for the agricultural sector in can be written as
(5.1)
where C
s
= α
s
(cR
s
+ dN
s
) is the constant referring to the income per capita and population in
each province.
Name
Estimate
Std.dev
a1
-49.275
38.086
a2
0.031025
0.025137
b1
2178.3072
2459.8477
b
2
-93.3839
98.4645
c
285.896
130.2340
d
2.3061
0.0754
Table 5.1. Estimated parameters of the impacts function for agriculture
5.2 Forestry
The forestry sector is the only sector in GRACE_adapt where the management of the natural
resource is addressed. This section discusses both the calibration of equations of the forestry
module, and the estimation of impacts. The calibration of the model is based partly on FAO
statistics (FAO, 2006) and partly on the GTAP database. Impacts are estimated on the basis of
results reported in literature as discussed in Section 4.3.
The point of departure for the forestry module is the Lotka-Volterra equation between the
stock of a resource, S, and its annual rate of growth
:
B and c are positive parameters. In principle, the two parameters of the quadratic Lotka-
Volterra equation can be calibrated by means of an observed combination of standing mass
and the harvest subject to the first-order condition for optimal harvest, which states that the
combination of harvest and mass is at the point where rate of change in growth of mass equals
the social rate of return on capital.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
35
However, observations of the pair of standing mass and corresponding harvest, as described
by the Lotka-Volterra equation, are not available. FAO (2006) provides data for standing
mass either in terms of total mass of forest or as potentially commercial mass. Potentially
commercial mass is defined with reference to physical criteria about the quality of the forest,
and must be separated from mass that is being commercially utilized. An underlying
economic explanation to why forest defined as commercial is not utilized, is that investments
are needed to get access to and utilize the forest.
The model distinguishes between utilized and non-utilized commercial forests by including
infrastructure capital, k
2
, which enables the manager to control the Lotka-Volterra equation
and make a part of it apply only to the utilized part of the standing mass, V
u
.
(5.2)
b and β are positive parameters. The owner then has to decide both on the optimal harvest and
on the investment in infrastructure. The first-order condition for optimal harvest is:
(5.3)
where r is the social rate of return on capital (the discount rate) and wtp is the willingess to
pay for a standing mass of forest. The first order condition for investments in infrastructure is
(5.4)
The parameter β is chosen by assumption. Then B and c and b can be calibrated by (2) – (4)
from observations of the variables V
u
, and k
2
.
What is observed is, however, the total change of the forest,
, the harvest and the total mass
of forest, S. One way out of this problem is to refer all economic behaviour to the utilized part
of the forest, such that harvest equals growth in this part of the forest. Moreover, we take
advantage of the assumption that the utilized and the non-utilized forests are subject to the
same physical conditions. From (2), the utilized forest is controlled by changing the constant
term, B. The “remaining” part of the constant term applies to the non-utilized forest,
(5.5)
The growth of non-utilized forests is not observed directly, but may be deducted under certain
assumptions: As the entire growth in the utilized part is assumed to be harvested, only the
non-utilized part of commercial forests will be subject to change. This is partly due to an
increase in the forested area, for which data are available, and in the total stock of forests,
which is also available. The growth in non-utilized commercial forests,
, can be found by
subtracting an estimate of mass in new forested land from the observed change in total stock.
Currently, the total growth of forest is allocated entirely to new forested area, such that
in all regions. The division between the two catagories of stock is determined by
(5.6)
where S is observed. Now, B, b, c, V
u
and V
n
can be calibrated simultaneously by equations
(5.2) through (5.6). The estimates are shown in Table 5.2.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
36
Capital
infr.
B b β c
Utilized
Non-
utilized
(estimate)
Mill m
3
Mill m
3
Baltic states
122.2
0.128524
2.16101
0.8
-2.54149E-05
1184.1
1819.3
British Islands
79.0
0.296632
4.24978
0.8
-6.00833E-04
142.7
214.5
Central East
51.0
0.049465
0.13920
0.8
-5.81622E-06
1630.3
1495.9
Central North
322.9
0.181488
8.51080
0.8
-2.65899E-05
1502.3
0.0
Southern Europe
117.0
0.075640
0.68452
0.8
-4.18879E-05
1137.3
362.0
Central West
336.2
0.213063
10.07123
0.8
-5.99394E-05
946.5
1599.9
Iberia
367.5
0.152306
8.47029
0.8
-5.08071E-05
921.1
0.0
Nordic
818.4
0.406879
46.25677
0.8
-6.07729E-05
1451.3
3519.3
Table 5.2. Calibrated parameters of the forestry module stocks by region
To divide into districts, the parameter B and the mass variables are divided in accordance with
the key for distribution of economic activity on provinces.
Indicator
Estimate
Std. dev.
Precipitation (mm/year)
0.00129
0.00025
Temperature (°C)
0.00887
0.00595
Table 5.3. Estimated linear impacts of precipitation and temperature on the rate of change in
forest parameters.
The reported impacts of climate change are interpreted as a percentage simultaneous change
in natural growth and productivity of capital. Thus, a 10 percent increase in forests implies a
10 percent increase in the parameters B and b. The impacts function explains the rate of
change in B and b as linear functions of changes in precipitation and temperature. The
estimates are displayed in Figure 5.3
5.3 Fisheries
The complexities of possible connections between the fish stock and air temperature indicate
that single functions to represent such a relationship, as we base the GRACE_adapt model on,
should be interpreted with high caution. An alternative is, clearly, to disregard possible
impacts of climate change on fisheries. On the other hand, the message from the literature is,
that fisheries will, indeed, be affected, but it is difficult to predict in what direction and to
what extent. Impacts of climate change on fisheries are therefore included in a very simple
manner, mainly to clear the ground for an integrated analysis of possible changes to the
sector.
Climate change is assumed to affect the value of the natural resource both in fish commercial
fishing and in aquaculture. The stock of natural resource in commercial fishing can be
interpreted as the stock of fish, whereas the natural resource in aquaculture is the conditions
in the oceans, which include ocean climate. The relationship between air temperature and the
natural resource is based on two more or less arbitrarily chosen variables. One is the ideal air
temperature at which the stock of fish is at its highest, and the second is the rate of change in
stock per °C deviation from this temperature. The ideal temperature as well as the sensitivity

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
37
to air temperature is equal for all regions (provinces) in Europe, but they differ between
commercial fishing and aquaculture.
Let N
s
denote the value of the natural resource in province s (according to the reported GTAP
data), α the share of commercial fishing in the fishery sector, σ
j
the rate of change in stock
value per °C deviation from the ideal temperature in activity (j = c - commercial fishing), a -
aquaculture), T
j
*
the ideal air temperature for activity j, and T
s
the actual air temperature in
province s. Then, the change in the value of the natural resource is written as:
(5.7)
Because of the difficulties in predicting changes in sea temperature from air temperature, we
use the same value for dT for all regions and provinces across Europe. The following
parameters have been chosen:
Parameter Commercial
fishing
Aquaculture
σ
j
0.025
0.020
T
*
10.0 5.0
Table 5.4. Choice of parameters in the impacts function for fisheries
5.4 Electricity supply
The electricity sector is based on different technologies which relate to climatic factors in
different ways: While thermal power may be constrained by the availability of cooling water,
the supply of renewable energy is more directly related to precipitation (hydro power) and
wind (wind power). Although the importance of wind power is increasing across Europe,
hydro power is still the main renewable source of electricity. Roughly speaking, more
precipitation is likely to increase production in both renewable and thermal power production,
although to a widely different extent, whereas temperature may have an influence on the
supply of thermal power, but have little influence on the supply of hydro power.
On the basis of estimates of each region in GRACE_adapt, the sensitivity of climate change
to generation of electricity is explained by a linear relationship between output and change in
precipitation, temperature and the share of hydro power in the system. For seven of the eight
regions, hydro power constitutes between 2 and 13 percent of total electricity supply. In the
Nordic countries, the contribution is nearly 50 percent. One might, therefore, regard the
Nordic countries as an “outlier”, but here, we stick to the same relatiohship for all regions:
(5.8)
where dX is percent change, dT °C change in temperature dP is mm/year change in
precipitation and s is the share of hydro power in total electricity supply. The following
estimates are being used
Indicator
Estimate
Std. deviation
Precipitation (mm/year)
0.1355
0.1127
Temperature (°C)
-0.0122
0.0049
Share of hydro
0.3811
0.0929
Table 5.5. Estimated sensitivity of climatic parameters on generation of electricity.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
38
The observations refer to the change in output from the electricity sector, but it is interpreted
as the percent change in the input of natural resources to the generation of electricity in the
model.
5.5 Energy demand
The temperature adjustment of energy demand is assumed to affect the household sector
(residential energy demand) and the service sector. The changes are based directly on
elasticities suggested in the literature (see chapter 4.5.1), where there is a distinction between
warm and cold regions. In warm regions, electricity demand is assumed to increase with
increasing temperature for cooling, whereas cold regions reduce their electricity demand
because of less heating. The demand for gas and refined products is being used only for
heating, and the elasticity with respect to temperature is therefore negative. Southern Europe
and Iberia are regarded “hot” regions, and the remaining six regions are regarded “warm”.
The following elasticities are being used
Regions
Electricity
Gas
Refined oil
S. Europe and Iberia
1
-2.8
-1.46
Remaining six
-0.38
-2.6
-1.43
Table 5.6. Temperature elasticities in residential and service sectors by carrier and region
5.6 Tourism
The underlying drivers behind a possible change in tourism are complex and diversified.
Little is known about the relationship between climate and tourism, although the emphasis put
on climatic conditions in promotion of tourism is a strong indication of the existence of such a
relationship. But sensitivities depend on what the tourists are attracted to, and one should
distinguish between domestic and foreign tourism, and winter and summer tourism.
Moreover, tourism is not an economic sector, neither in GRACE_adapt nor in the Standard
for National Accounts, but it affects different sectors. In GRACE_adapt, mainly the service
sector and transport are likely to be affected.
The current estimates in GRACE_adapt are based on strong simplifications. First, the
percentage change in total tourism is estimated as a linear function of the changes in
temperature and precipitation and the product of temperature and temperature change. The
possible importance of differences between hot and cold regions is thereby taken into
account:
(5.9)
where dY is the percent change in demand, T is the °C level of temperature dT is the °C
change of temperature and dP is the percent change in precipitation. The parameters are
Variable
Estimate
Std. error
Temperature change (°C)
0.08678
0.01410
Change in precipitation (mm/year)
-0.16909
0.13615
Combination of level and change in temp (°C)
-0.00702
0.00133
Table 5.7. Estimated sensitivities of tourism

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
39
The product of temperature level and temperature change allows the sensitivity to temperature
change to differ across the regions, depending on the temperature level. The estimates
indicate that the impact of climate change on tourism changes from positive to negative as the
regional level of mean temperature increases above 12.5 °C.
In the model, it is assumed that the changes in the tourism sector are estimated as the percent
change of demand for transport and services in the household sector, divided according to the
contribution from tourism in these two sectors, (j = service, transport):
(5.10)
where ε
j
is the contribution from tourism in sector j.
5.7 Extreme events
It is difficult to deal with extreme events in general equilibrium models for three reasons.
First, the events will have to be represented by an expression for expected impacts, unless the
model is run with stochastic shocks. The economic implications of the same loss over ten
years is likely to differ substantially from a loss of ten times this loss in one year and no
losses over nine years. Whether or not to run a model with stochastic shocks depends on the
purpose of the analysis. To examine the real impacts of extreme events, stochastic shocks are
the most relevant. To evaluate implications of present decisions, an expression for the
expectations may be more relevant. The second difficulty relates to the local character of
impacts of some extreme events, such as land-slides. Even the province level suggested in
GRACE_adapt is too aggregated to capture the real socioeconomic costs of many extreme
events.
It is also difficult to predict how climate change will affect future frequencies of extreme
events. It is sometimes made a point out of the fact that if the density functions for variation
in temperature, precipitation and wind are kept constant, but only the mean values change, the
frequencies of what can be labeled extreme events will increase more than proportionally to
the change in the mean values. The assumption of an unaffected density function may even be
considered conservative.
The third difficulty is that the impacts of extreme events are very case-specific, and it is
difficult to use historic evidence to predict impacts, even if the weather events themselves
could be forecasted. It is, therefore, difficult also to draw general lessons about single studies
of impacts of extremes from different places. In large parts of Europe, there are expectations
of an increase of floods, droughts, windstorms and land-slides resulting from climate change.
We have, therefore, included some rough estimates, based on historical records of impacts of
weather related natural hazards, and an illustrative assumption about the increase of
frequency.
In GRACE_adapt, extreme events are represented by the average annual cost, and are based
on historic data for the weather related natural disasters in Europe, which include droughts,
extreme temperature, floods, slides, storms and wild fires.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
40
Number indicators
Cost
Basis
at +3.5
o
C
% of K
A
Baltic
0.497
0.994
0.74
0.060533
British
0.449
0.899
0.61
0.049454
East
0.630
1.259
1.19
0.097073
North
0.418
0.835
0.52
0.042703
South
0.465
0.931
0.65
0.053042
West 0.407 0.815 0.50 0.040667
Iberia
0.638
1.276
1.22
0.099620
Nordic
0.378
0.756
0.43
0.034973
Table 5.8. Indicator for number of extreme events by region by period, associated costs of
damage as percent of total stock of capital and constant term of damage function
We simply assume that the costs of extreme events are quadratic in the number of events to
reflect that the distributions of costs of single events are skewed, with long “tails” to cover
probable disasters. We moreover assume that the average temperature change for Europe at
3.5 °C implies a doubling of the frequency of disasters. The cost function for extreme events
reflects the percent change in the capital stock, K
i
, of region i,:
(5.11)
where A
i
is the parameter shown in the last column in Table 8 and α = 2. Note that measuring
damage in percent of K captures some of the often observed phenomenon that damage of
natural disasters tend to increase without climatic changes as countries grow rich.
5.8 Sea-level rise
A lot of work has been carried out to assess the cost of sea-level rise, and it has been
emphasised that estimates vary considerably depending on whether adaptation is taken into
account or not. There are available estimates for some EU countries, and for EU total, but a
detailed distribution of costs on the EU countries is not available. As a preliminary solution,
the current impacts are based on an estimate of total annual cost for Europe of 1 700 mill
USD for 1 meter rise of sea level from Tol (2002), and distributed according to (preliminary
approximations) of coast-line, low-lying areas and population by province as defined in
GRACE_adapt. The cost function for sea-level rise is quadratic, and is assumed to affect the
change the capital stock. The cost in province s is
(12)
where dk
s
is the percent change in capital stock, a
s
is province s’ share of Europe’s total
coastline, and L
s
is the low lying area (threatened by flooding) in the province. This indicator
is presently indexed from 1 to 5 by province. α and β are parameters that indicate the
importance of coastline and low-lying areas, respectively, for the total cost (each are set equal
to 1). N
s
is the population in province s and γ is the scale parameter determined to make sure

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
41
that the right hand side sums up to the total annual cost of sea level rise of 0.0102, which
corresponds to a capital loss of 1 700 mill USD. Then γ = 0.00001995. Finally, dR is the
increase of sea-level rise in meters.
6 Results
Adaptation is a question of transition and time. Some barriers to adaptation are likely to
become smaller over time, whereas others sustain over long periods. As for the constraints on
mobility addressed in GRACE, the barriers diminish over time at different rates. Natural
resources are hardly mobile at all, whereas the mobility of capital in most cases depends on
the life-time of the capital equipment. Labour is, in principle, fully mobile, but long-term
unemployment in certain areas indicates that the reality is different. Products also are subject
to different degrees of mobility. Commodities can be regarded fully mobile, whereas services
will in most cases have to be supplied at the place where the demand is.
How the time perspective of adaptation is dealt with is therefore utterly important when
running scenarios in integrated models with endogenous adaptation, such as GRACE. In this
section, we leave this issue behind by running only static solutions. That is, we compare the
reference case with three alternatives where the increase in global mean temperature is +2 °C,
+3 °C and +4 °C. The impacts in each province are calculated by linear extrapolations of
mean temperature and precipitation from the PESETA study referred to above, where the
increase of global mean temperature is assumed at +3.5 °C. For example, in a province where
the mean temperature increases by +3 °C according to the PESETA study, we assume that the
increase in the +4 °C scenarios is slightly below +3.5 °C in this province. The changes in
precipitation are adjusted correspondingly in relation to the PESETA scenario.
In principle, we thereby compare two states where the climate shifts immediately, and impose
some restrictions to adaptation. More specifically, we assume no mobility of natural resources
and no mobility of labour. Whereas the restrictions on natural resources are real also in the
long term, a full restriction on the mobility of labour over the period where the mean
temperature increases by +2 °C, +3 °C and even +4 °C is, indeed, an exaggeration. This is to
some extent compensated by the full mobility of capital, however. The choice of putting
restrictions on labour is due to our wish to address possible social inequalities that may arise
from climate change impacts. Moreover, the restrictions apply to the mobility across
provinces, which are relatively large. Within provinces, there is full mobility.
GRACE_adapt is a global model, where the rest of the world is represented by two regions,
one for industrialized countries and one for developing countries. There is full trade between
the European regions and the two world regions, but we assume no impacts of climate change
outside Europe. If negative impacts were imposed on the rest of the world, the negative
impacts in Europe would most likely increase. This is why the level of changes in this study is
of limited importance. What matters are the relative changes between regions, provinces and
across the three cases +2 °C, +3 °C and +4 °C.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
42
Figure 6.1. Percent change in GDP with and without adaptation with increases in global
mean temperature at +2 °C, +3 °C and +4 °C.
Figure 6.1 shows the percent change in GDP in the three cases for the eight regions of
Europe. The changes are calculated both with and without adaptation. The impacts “without
adaptation” are the estimated impacts directly from the impacts functions described in Section
5 expressed in percent of GDP. The impacts “with adaptation” are the percent change in GDP
in the new equilibrium solution. From the figures, we note that the results give some
justification of the +2 °C target advocated by the EU: The +2 °C case shows small economic
impacts, with positive changes in some regions and negative in others. For higher temperature
increases, the impacts are negative in all regions. We also note that the impacts are non-linear:
A +1 °C increase in the global mean temperature results in a doubling of the costs of climate
change.
Adaptation takes place in terms of changes in quantities, but changes in prices also constitutes
an important part of the adaptation process when comparing GDP in the different cases. The
total effect of adaptation is substantial, and reduces the impacts by approximately 80 - 85
percent in many regions. This large effect of adaptation may partly be explained by the
assumption that the world outside Europe is unaffected by climate change. This opens
Europe’s ability to draw advantage of the benefits of climate change and to avoid the
disadvantages, and this process is driven by the changes in world market prices.
The implications of the market effects differ a lot from region to region, though. Somewhat
surprisingly, it is the Baltic and the Eastern part of Central Europe, with the lowest income
among all European regions, which draw most advantage of adaptation. When comparing the
direct impacts of climate change, these two regions are the second and fourth most negatively
affected. However, it is only these two regions that benefit from climate change in the +2 °C
case, when market responses are taken into account. The costs in terms of GDP losses are also
the lowest among all regions in the two other cases.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
43
Figure 6.2. Variation of percent change in GDP among provinces by region in the +2 °C and
the +4 °C cases.
According to our results, Southern Europe and Iberia face the highest costs of climate change,
with more than twice the costs in the northern and western regions of Central Europe, the
British Islands and the Nordic countries. As mentioned, the Baltics and eastern part of Central
Europe face lower impacts in all cases, but the differences between these two regions and the
other regions diminishes as temperature increases. However, when comparing provinces
within the regions, the picture changes somewhat and new patterns occur. Figure 6.2 shows
the ranges (from maximum to minimum) within which the changes in GDP vary within
regions in the +2 °C and the +4 °C cases. The regional averages are marked by dots.
Although the low-income regions lose the least in total, the variations across provinces are
substantial. Both in the +2 °C and the +4 °C cases, there are provinces in these two regions
where the costs exceed the average European level. Also the two southern regions exhibit
large variations, in particular Iberia, where the costs of climate change vary between 0.06 to
0.16 percent in the +2 °C case and between 0.36 and 0.68 in the +4 °C case.
Going further into the details of these variations, it turns out that for most regions, there is a
significant correlation between loss in terms of percent change in GDP and income per capita:
The lower the income per capita is in a province, the higher is the costs of climate change.
The explanation may be that provinces with low income are more dominated by resource
dependent economic activities, such as agriculture.
Figures from Eurostat indicate that provinces with large agricultural sectors tend to exhibit
higher long-term unemployment than other provinces. One may therefore suspect that climate
change can have social implications that have not yet been addressed, because the provincial
dimension is absent in previous economic studies. The GRACE_adapt model has this
dimension, but does not address unemployment directly. It is assumed that all resources are
used, and unemployment is therefore disregarded by assumption. However, the variability of
impacts on the labour market can be indicated by the difference in the changes of wage levels
that arise because of climate change.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
44
in four of the eight regions (Baltic and the eastern, northern and western parts of Central
Europe) there is a tendency that the wages in low income provinces are more negatively
affected. However, we cannot conclude on a general basis whether or not climate change may
enforce present social challenges, such as long-term unemployment, but a closer examination
of the changes in various provinces may be worthwhile.
7 Conclusions and remaining questions
The development of the integrated computable general equilibrium model GRACE_adapt
represents an attempt to bring consistency between adaptation to climate change and
economic behaviour. Adaptation to climate change can thereby be analysed endogenously by
means of the assumptions underlying the behaviour of economic agents. Apart from providing
a documentation of the modelling, this report presents the first results of the model runs,
which addresses one of the two questions that was raised as the background for the ADAM
project, namely to compare the costs of mitigation to aim at +2 °C increase in global mean
temperature with the need for adaptation if the global mean temperature increase by +4 °C.
We found that adaptation indeed contributes to reduce the macroeconomic impacts of climate
change, but that the share of costs that is reduced by adaptation decreases as global mean
temperature increases. Still, the economic impacts at an increase in global mean temperature
at +4 °C are reduced by between 80 and 85 percent European wide as a result of adaptation in
the runs presented here.
Both the level of economic impacts and the degree to which adaptation mitigates the initial
economic costs of climate change vary greatly both between the eight European regions and
wihin regions. The southern regions are most affected by climate change, whereas the low-
income regions seem to take the most advantage from adaptation. This can be explained by
the change in relative prices, which turn out in favour of the low-income regions. The low-
income regions also exhibit relatively large variations within the region. This can also be
considered a result of the changes in relative prices: Some provinces are better capable of
drawing the advantage of this than others. Also the most affected regions exhibit large
variations across provinces. In some regions there is a strong correlation between low income
per capita and losses in GDP or reductions in wages on the province level , but such a pattern
cannot be found in all regions.
It has to be emphasised that the results of the present study are preliminary and subject to
notable weaknesses. First, we present only a static comparison between equilibria under
different assumptions about climatic conditions. To get a more realistic picture of the
economic consequences, one should run scenarios over a given time period. In that case, a
closer examination of the barriers to adaptation would have to be considered, because barriers
will diminish over time. Second, we have assumed that the rest of the world is unaffected by
climate change. As mentioned, the effect of adaptation depends heavily on the change in
relative prices, also on the world market. Implementation of impacts in the rest of the world is
therefore likely to affect the results considerably. Hence, the levels of impacts reported here
are probably biased, which is why we do not want to highlight them. However, we believe
that we have gained some insights when it comes to the comparison of European regions and
variabilities within regions.
Wee would also like to draw the attention again to the uncertainties about the estimates of the
impact of climate change. Numerical assessments of impacts are still scarce and difficult to
compare. We chose to base our estimates on an assumption that the same impacts function
applies across regions in Europe. This is a strong assumption, indeed. Moreover, the impacts

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
45
that were taken into account probably represent moderate estimates, because some effects of
climate change have been disregarded. Health effects, for example, are not included at all in
the numerical analysis, and all impacts are imposed by means of expected values. The most
severe economic impacts of climate change are related to extreme events, which are
stochastic by nature. Stochastic events of certain magnitudes may itself general negative
impacts, which have been disregarded here. A closer study of these impacts was, however,
addressed in Work package A2 of the ADAM project.
Finally, there has been no attempts to include non-market impacts of climate change, which
may turn out to be substantial. Hence the results of this study must be considered a “pure”
economic assesssment of impacts. A full evaluation of the impacts of climate change requires
inclusion of many more aspects.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
46
Literature
Aaheim, A., & Hauge, K. E. (2005). Impacts of climate change on travel habits: CICERO Center for
International Climate and Environmental Research.
Aaheim, A., Helene Amundsen, Synne Solstad, Jonas Vevatne, West, J., & Aasen, M. (2008).
Contributing author(s) to: Betydningen for Norden av 2 grader global oppvarming - Vurdering av
sårbarhet og effekter av klimaendringer: TemaNord 2008:507. Nordisk Ministerråd. Copenhagen,
Denmark. 157 pp.
Aaheim, A., and A. Schjolden (2004): “An Approach to Utilise Climate Change Impacts Studies in
National Assessments”, Global Environmental Change [14], 147-160
Aaheim, A. and N. Rive (2005): “A model for Global Responses to Anthropogenic Changes in the
Environment (GRACE)”, CICERO Report 2005:5, CICERO, Oslo
Aaheim, A. (2007): Macroeconomic modelling, ADAM deliverable from work package A1.3. EU
Commission.
ACIA. (2004). Impacts of a warming Arctic: Cambridge University Press, Cambridge, UK. 144pp.
Alcamo, J., Moreno, J. M., Nováky, B., Bindi, M., Corobov, R., Devoy, R. J. N., et al. (2007). Europe.
Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In M. L. Parry,
O. F. Canziani, J. P. Palutikof, P. J. van der Linden & C. E. Hanson (Eds.), (pp. 541-580):
Cambridge University Press, Cambridge, UK.
Amelung, B., Nicholls, S., & Viner, D. (2007). Implications of Global Climate Change for Tourism
Flows and Seasonality. Journal of Travel Research, 45(3), 285-296.
Arnell, N. W., Tompkins, E., Adger, N., & Delaney, K. (2005). Vulnerability to Abrupt Climate
Change in Europe. : Tyndall Centre for Climate Change Research.
Beaugrand, G., & Reid, P. C. (2003). Long-term changes in phytoplankton, zooplankton and salmon
related to climate. Global Change Biology, 9(6), 801-817.
Berrittella, M., Bigano, A., Roson, R., & Tol, R. S. J. (2006). A general equilibrium analysis of climate
change impacts on tourism. Tourism Management, 27(5), 913-924.
Bigano, A., Bosello, F., & Marano, G. (2006). Energy Demand and Temperature: A Dynamic Panel
Analysis: SSRN.
Broadmeadow, M. S. J., Ray, D., & Samuel, C. J. A. (2005). Climate change and the future for
broadleaved tree species in Britain. Forestry, 78(2), 145-161.
Broadmeadow, M. S. J., Ray, D., Sing, L., & Poulson, L. (2003). Climate Change and British
woodland: what does the future hold? : Forest Research Annual Report and Accounts 2002-2003.
Cartalis, C., Synodinou, A., Proedrou, M., Tsangrassoulis, A., & Santamouris, M. (2001).
Modifications in energy demand in urban areas as a result of climate changes: an assessment for the
southeast Mediterranean region. Energy Conversion and Management, 42(14), 1647-1656.
Carter, T. R., Saarikki, R. A., & Joukainen, S. K. H. (2000). Modelling climate change impacts on
wheat and potatoes in Finland. Chapter 16. In T. E. Downing, P. A. Harrison, B. R. E & K. G.
Lonsdale (Eds.), Climate change, climatic variability and agriculture in europe: An integrated
assessment . Research Report No. 21 (pp. 289-312): Environmental change unit, University of
Oxford, Oxford, UK.
Clark, R. A., Fox, C. J., Viner, D., & Livermore, M. (2003). North Sea cod and climate change -
modelling the effects of temperature on population dynamics. Global Change Biology, 9(11), 1669-
1680.
De Cian, E., Lanzi, E., & Roson, R. (2007). The Impact of Temperature Change on Energy Demand: A
Dynamic Panel Analysis: FEEM Working Paper No. 46.2007, Fondazione Eni Enrico Mattei.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
47
Dimaranan, B. V. (2006). Global trade, assistance and production: The GTAP 6 database: Center for
global trade analysis, Purdue university.
DMI. (2006). Virkninger av klimaedringer: Danmarks Meteorologiske Institut. Transport og
Energiministeriet. Accessed 2nd of October 2006 at
http://www.dmi.dk/dmi/virkninger_af_klimaaendringer.pdf.
DS. (2005). Sweden's fourth national communication on climate change: Ministry of Sustainable
Development, Sweden.
Drinkwater, K.F. (2005). The response of Atlantic Cod (Gadus morhua) to future climate change, ICES
Journal of Marine Science [62] 1327-1337
Dutil, J.-D., & Brander, K. (2003). Comparing productivity of North Atlantic cod (Gadus morhua)
stocks and limits to growth production. Fisheries Oceanography, 12(4-5), 502-512.
EC. (2006). Facts and figures on the CFP. Basic data on the Common Fisheries Policy: Luxembourg,
European Communities. 40 pp.
Eddowes, M. J., Waller, D., Taylor, P., Briggs, B., Meade, T., & Ferguson, I. (2003). Safety
Implications of Weather, Climate and Climate Change: Final Report (No.
AEAT/RAIR/76148/R03/005 Issue 2): AEAT.
Eurostat. (2005a). Fisheries. Data accessed 23rd of January 2008, from
http://epp.eurostat.ec.europa.eu/portal/page?_pageid=1996,45323734&_dad=portal&_schema=POR
TAL&screen=welcomeref&open=/fish/fish_aq&language=en&product=EU_MASTER_fisheries&r
oot=EU_MASTER_fisheries&scrollto=0.
Eurostat. (2005b). Tourism demand. Number of Tourist. Data from 2005. Data accessed 18th of April
2008, from
http://epp.eurostat.ec.europa.eu/portal/page?_pageid=1996,45323734&_dad=portal&_schema=POR
TAL&screen=welcomeref&open=/tour/tour_dem/tour_dem_to&language=en&product=EU_MAST
ER_tourism&root=EU_MASTER_tourism&scrollto=0.
FAO. (2005). Global Forest Resources Assessment 2005. Retrieved January 25th, 2008, from
http://www.fao.org/forestry/site/43035/en/
FAO (2006): Global Forest Resources Assessment 2005, FAO Forestry Paper 147, Food and
Agricultural Organization of the United Nations, Rome.
FAO. (2007). FAO Statistical Yearbook 2004-2005. Capture Fisheries and Aquaculture Production
(2004): FAO Statistics Division.
Frank, T. (2005). Climate change impacts on building heating and cooling energy demand in
Switzerland. Energy and Buildings, 37(11), 1175-1185.
Fronzek, S., & Carter, T. (2007). Assessing uncertainties in climate change impacts on resource
potential for Europe based on projections from RCMs and GCMs. Climatic Change, 81(0), 357-371.
Giannakopoulos, C., Bindi, M., Moriondo, M., LeSager, P., & Tin, T. (2005). Climate change impacts
in the Mediterranean resulting from a 2°C global temperature rise: WWF.
Hamilton, J. M., Maddison, D. J., & Tol, R. S. J. (2005). Climate change and international tourism: A
simulation study. Global Environmental Change Part A, 15(3), 253-266.
Hamilton, J. M., & Tol, R. (2007). The impact of climate change on tourism in Germany, the UK and
Ireland: a simulation study. Regional Environmental Change, 7(3), 161-172.
Hein, L. (2007). The Impact of Climate Change on Tourism in Spain. Oslo, Norway: CICERO
Working Paper 2007:02.
Hildén, M., Lehtonen, H., Bärlund, I., Hakala, K., Kaukoronta, T., & Tattari, S. (2005). The practice
and process of adaptationi in Finnish agriculture. .
Hiscock, K., Southward, A., Tittley, I., & Hawkins, S. (2004). Effects of changing temperature on
benthic marine life in Britain and Ireland. Aquatic Conservation: Marine and Freshwater
Ecosystems, 14(4), 333-362.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
48
IEA. (2005). IEA Energy Statistics.
Jones, G., White, M., Cooper, O., & Storchmann, K. (2005). Climate Change and Global Wine Quality.
Climatic Change, 73(3), 319-343.
Kammann, C., Grünhage, L., Grüters, U., Janze, S., & Jäger, H. J. (2005). Response of aboveground
grassland biomass and soil moisture to moderate long-term CO2 enrichment. Basic and Applied
Ecology, 6(4), 351-365.
Karing, P., Kallis, A., & Tooming, H. (1999). Adaptation principles of agriculture to climate change.
Climate Research, 12(2-3), 175-183.
Kellomäki, S., Karjalainen, T., Mohren, F., & Lapveteläinen, T. (2000). Expert Assessments of the
Likely Impacts of Climate Change on Forests and Forestry in Europe: European Forest Institution.
University of Wageningen, The Netherlands. University of Joensuu, Finland.
Kirkinen, J., Martikainen, A., Holttinen, H., Savolainen, I., Auvinen, O., & Syri, S. (2005). Impacts on
the energy sector and adaptation of the electricity network business under a changing climate in
Finland. FINADAPT Working Paper 10 (pp. 36 pp). Finnish Environment Institute Mimeographs
340, Helsinki.
Lasch, P., Lindner, M., Erhard, M., Suckow, F., & Wenzel, A. (2002). Regional impact assessment on
forest structure and functions under climate change--the Brandenburg case study. Forest Ecology
and Management, 162(1), 73-86.
LCCP. (2002). The Impacts of Climate Change in London. Technical Report: London Climate Change
Partnership. A Climate Change Impats in London Evaluation Study. Final Report.
LCCP. (2005). Climate Change and London's transport systems. Summary Report: Greater London
Authority.
Lehner, B., Czisch, G., & Vassolo, S. (2005). The impact of global change on the hydropower potential
of Europe: a model-based analysis. Energy Policy, 33(7), 839-855.
Lexer, M. J., Hönninger, K., Scheifinger, H., Matulla, C., Groll, N., Kromp-Kolb, H., et al. (2002). The
sensitivity of Austrian forests to scenarios of climatic change: a large-scale risk assessment based on
a modified gap model and forest inventory data. Forest Ecology and Management, 162(1), 53-72.
Long, S. P., Ainsworth, E. A., Leakey, A. D. B., Nosberger, J., & Ort, D. R. (2006). Food for Thought:
Lower-Than-Expected Crop Yield Stimulation with Rising CO2 Concentrations. Science,
312(5782), 1918-1921.
Lüscher, A., Daepp, M., Blum, H., Hartwig, U. A., & Nösberger, J. (2004). Fertile temperate grassland
under elevated atmospheric CO2--role of feed-back mechanisms and availability of growth
resources. European Journal of Agronomy, 21(3), 379-398.
Maddison, D. (2001). In Search of Warmer Climates? The Impact of Climate Change on Flows of
British Tourists. Climatic Change, 49(1), 193-208.
Maracchi, G., Sirotenko, O., & Bindi, M. (2005). Impacts of Present and Future Climate Variability on
Agriculture and Forestry in the Temperate Regions: Europe. Climatic Change, 70(1), 117-135.
Moen, J., Aune, K., Edenius, L., & Angerbjörn, A. (2004). Potential Effects of Climate Change on
Treeline Position in the Swedish Mountains. Ecology and Society 9(1), 16.
Nakicienovic, N. and Swart, R. (ed.) (2000) Special Report on Emission Scenarios. Intergovernmental
Panel of Climate Change, Cambridge University Press, Cambridge
Olesen, J. E., 2006. Climate change as a driver for European agriculture. Danish Institute of
Agricultural Sciences. SCAR Standing Committee on Agriculture Research portal:
http://ec.europa.eu/research/agriculture/scar/index_en.cfm?p=3_foresight
Olesen, J.E., Carter, T., Díaz-Ambrona, C., Fronzek, S., Heidmann, T., Hickler, T., et al. (2007).
Uncertainties in projected impacts of climate change on European agriculture and terrestrial
ecosystems based on scenarios from regional climate models. Climatic Change, 81(0), 123-143.

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
49
Parsons, D. J., Armstrong, A. C., Turnpenny, J. R., Matthews, A. M., Cooper, K., & Clark, J. A.
(2001). Integrated models of livestock systems for climate change studies. 1. Grazing systems.
Global Change Biology, 7(1), 93-112.
Perry, A. L., Low, P. J., Ellis, J. R., & Reynolds, J. D. (2005). Climate Change and Distribution Shifts
in Marine Fishes. Science, 308(5730), 1912-1915.
PESETA. (2007). Comission staff working document. Accompanying document to the "Limiting Global
Climate Change to 2 degrees Celsius. The way ahead for 2020 and beyond": Commission of the
European Communities.
Pryor, S., Barthelmie, R., & Kjellström, E. (2005). Potential climate change impact on wind energy
resources in northern Europe: analyses using a regional climate model. Climate Dynamics, 25(7),
815-835.
Roessig, J. M., Woodley, C. M., Cech, J. J., & Hansen, L. J. (2004). Effects of global climate change
on marine and estuarine fishes and fisheries. Reviews in Fish Biology and Fisheries, 14(2), 251-275.
Reilly, J., Paltsev, S., Felzer, B., Wang, X., Kicklighter, D., Melillo, J., et al. (2007).
Global economic effects of changes in crops, pasture, and forests due to changing
climate, carbon dioxide, and ozone. Energy Policy, 35(11), 5370-5383
Roson, R., Bosello, F., & De Cian, E. (2007). Climate Change, Energy Demand and Market Power in a
General Equilibrium Model of the World Economy. FEEM Working Paper No. 71.2007;
Department of Economics, University Ca' Foscari of Venice Research Paper No. 09-09. Available at
SSRN:
http://ssrn.com/abstract=999920.
RossbyCentre. (2006). Konsekvenser av klimatförändringen (Consequences of climate change).
Retrieved 26th of April, 2008, from
http://www.smhi.se/sgn0106/klimat/konsekvenser.htm
Räisänen, J., Hansson, U., Ullerstig, A., Döscher, R., Graham, L. P., Jones, C., et al. (2004). European
climate in the late twenty-first century: regional simulations with two driving global models and two
forcing scenarios. Climate Dynamics, 22(1), 13-31.
Rötter, R., & van de Geijn, S. C. (1999). Climate Change Effects on Plant Growth, Crop Yield and
Livestock. Climatic Change, 43(4), 651-681.
Simpson, M. C., Gössling, S., Scott, D., Hall, C. M., & Gladin, E. (2008). Climate Change Adaptation
and Mitigation in the Tourism Sector: Frameworks, Tools and Practices: UNEP, University of
Oxford, UNWTO, WMO: Paris, France.
Smith, J., & Hitz, S. (2003). Background Paper: Estimating Global Impacts from Climate Change.,
OECD Workshop on the Benefits of Climate Policy: Improving Information for Policy Makers (pp.
101 pp). OECD, France.
Sonesson, J. (2004). Climate change and forestry in Sweden - a litterature review. Report from Climate
and the Forest Committee. Kungl. Skogs-och Lantbruksakademiens Tidskrift 143(18).
StatisticsFinland. (2008). Tourism account 1995-2006. National Accounts
Retrieved 14th of September, 2008, from
http://www.stat.fi/til/matp/2006/matp_2006_2008-04-
03_tie_001_en.html
Stuczyiñski, T., Demidowicz, G., Deputat, T., Górski, T., Krasowicz, S., & Kuœ, J. (2000). Adaptation
Scenarios of Agriculture in Poland to Future Climate Changes. Environmental Monitoring and
Assessment, 61(1), 133-144.
Sykes, M. T., & Prentice, I. C. (1996). Climate change, tree species distributions and forest dynamics:
A case study in the mixed conifer/northern hardwoods zone of northern Europe. Climatic Change,
34(2), 161-177.
Tammelin, B., Forsius, J., Jylhä, K., Järvinen, P., Koskela, J., Tuomenvirta, H., et al. (2002).
Ilmastonmuutoksen vaikutuksia energiantuotantoon ja lämmitysenergian tarpeeseen (Effect of
climate change on energy production and heating power demand): Finnish Meteorogical Institute,
Helsinki (in Finnish).

CICERO Report 2009:06
A macroeconomic assessment of impacts and adaptation to climate change in Europe
50
Tol, R. S. J. (2002). Estimates of the Damage Costs of Climate Change, Part II. Dynamic Estimates.
Environmental and Resource Economics, 21(2), 135-160.
Toresen, R. (2000). Sildefiske og klimaendringer, 2000. (Herring fisheries and climate change - in
Norwegian) Havets miljø 2000. Havforskningsinstituttet, Bergen
TSA. (2008). Tourism Satelite Accounts. Country Reports: Accessed 15th of April from
http://www.wttc.org/eng/Tourism_Research/Tourism_Satellite_Accounting/TSA_Country_Reports/.
WTTC.
Turnpenny, J. R., Parsons, D. J., Armstrong, A. C., Clark, J. A., Cooper, K., & Matthews, A. M.
(2001). Integrated models of livestock systems for climate change studies. 2. Intensive systems.
Global Change Biology, 7(2), 163-170.
Vadja, A., Venäläinen, A., Tuomenvirta, H., & Jylhä, K. (2004). An estimate of the influence of
climate change on heating energy demand in regions of Hungary, Romania, and Finland. Idõjárás.
Quarterly Journal of the Hungarian Meteorological Service, 108(2), 123-140.
Venäläinen, A., Tammelin, B., Tuomenvirta, H., Jylhä, K., Koskela, J., Turunen, M. A., et al. (2004).
The Influence of Climate Change on Energy Production & Heating Energy Demand in Finland.
Energy & Environment, 15, 93-109.
Viner, D., Sayer, M., Uyarra, M. C., & Hodgson, N. (2006). Climate Change and the European
Countryside: Impacts on Land Management and Response Strategies. Report Prepared for the
Country Land and Business Assiciation., UK. Publ., CLA, UK.
White, A., Cannell, M. G. R., & Friend, A. D. (2000). The high-latitude terrestrial carbon sink: a model
analysis. Global Change Biology, 6(2), 227-245.
Wolf, J. (2002). Comparison of two potato simulation models under climate change. II. Application of
climate change scenarios. Climate Research, 21(2), 187-198.
Wooller, S. (2003). The changing climate: its impact on the Department for Transport. In D. f.
Transport (Ed.).