Introduction to GEM-E3 (with special emphasis on impacts assessments)
(JRC PESETA II project)
1. Juan-Carlos Ciscar *2. Potsdam, September 18, 2013
* The views expressed are purely those of the authors and may not in any circumstances be regarded as stating an official position of the European Commission
Outline
1. GEM-E3 Model
2. JRC PESETA II, EU assessment
3. Conclusions
The GEM-E3 Model:General Equilibrium Model for Energy-Economics-Environment
interactions
www.GEM-E3.net
The GEM-E3 Model: The Development
The model is a result of a collaborative effort by a consortium involving:
Core modelling team• National Technical University of Athens (Coordinator)• Catholic University of Leuven (Centre for Economic Studies)• University of Mannheim and Centre of European Economic Research
(ZEW)
Contributors• University of Toulouse (IDEI)• University of Strathclyde• Stockholm School of Economics• Ecole Centrale de Paris (ERASME)• Catholic University of Leuven (CORE)• Middlesex University
Partially funded by the European Commission, Programme JOULE, DG-XII/F1
The GEM-E3 Model: Overview (1/3)
• Computable General Equilibrium model• representing multiple production sectors and countries• integrating energy and environment in the economy
• to simulate• the simultaneous competitive equilibrium of all markets: goods,
labour, energy, pollution permits etc.; • the endogeneous response of producers/consumers to
environmental/energy policies (abatement, policy limits, taxes);
• to evaluate• costs and benefits (including the environmental dimension)• distributional effects of policy instruments (different taxes and
subsidies, auctioning, permits, command-and-control)
• The model simulates an economy with:• multiple sectors, each producing a homogeneous commodity• a single representative firm operates in each sector
• minimizing cost under CRTS technology• deriving optimal demand for production factors (including all
other commodities, labour and capital)• a single representative household
• maximizing utility• allocating revenues to consumption of commodities and
savings• determining labour supply
• and a Government ensuring transfer distribution and applying policy through
• taxes, consumption, investments etc.
The GEM-E3 Model: Overview (2/3)
• The economy is open to foreign competition• Imported goods combine with domestic production to form
supply• Consumers (final, intermediate, government etc.) may
substitute domestic and foreign commodities to form demand
• The stock of capital is fixed within the period (either sectorally or for the country or World wide)• Constraining production possibilities in static terms
• while dynamically accumulating (through investments)
• Labour market includes unemployment
The GEM-E3 Model: Overview (3/3)
The GEM-E3 model: Framework
• Data:• Input/Output tables (GTAP-8)• Bilateral Trade matrices (GTAP-8)• Employment• Environment
• • Mixed Complementarily formulation using GAMS/PATH solver
• Templates for scenario building and result reporting
• Baseline scenario
The GEM-E3 model sectors
19 Sectors
• Agriculture• Coal• Refined Oil and coke• Crude oil• Gas • Electricity transmission and
distribution• Ferrous & non-ferrous, ore,
metals• Chemical Products• Other Energy-Intensive
Industries• Electrical Goods• Transport Equipment• Other Equipment Goods
Industries• Consumer Goods Industries• Building And Construction• Land Transport• Air Transport• Water Transport• Other Market Services• Non-Market Services
Power technologies
• 10 Power Technologies
• Coal Conventional thermal• Oil Conventional thermal• Gas Conventional thermal• Nuclear• Biomass• Hydro• Wind• PV & Solar• Coal CCS• Gas CCS
The GEM-E3 regional detail
•All EU27 member states individually represented
•USA •Canada •Japan •Oceania •China •India•Russian Federation•Brazil •Rest of Annex I•RoW
The GEM-E3 variables and parameters
•Population•Government consumption
and investment•Government tax, subsidy
and social benefit policies•Technical Progress
• Total Factor Productivity• Technical Progress
Embodied in Production Inputs
•Elasticity Parameters•Technical coefficients in
investment and consumption matrices
•All the elements of the Social Accounting Matrix
•Consumption by purpose•Investment and Capital•Bilateral Trade•Labour market
participation employment and unemployment
•Basic Interest Rate•Emissions and damages•Welfare and GDP Index
Exogenous Endogenous (volumes values and deflators)
The GEM-E3 Model: Simulation
GEM-E3 SAM
Economic circuit
•The model computes the price vector that simultaneously clears the product, capital and labor markets
Advantages of CGE modelling
• Consistency• Theory (microeconomics foundations, within a consistent
macroeconomic framework)• Data (National Accounts, SAM)
• Structural model (versus reduced-form models): explain behavior of agents in markets, taking into account institutions
• Takes fully into account the spill-over effects across sectors, consumers, government and other countries
• Transparency• Systematic analysis; not mechanical• Flexibility• Can address a broad range of policy issues: climate and energy,
taxation, trade, agriculture, etc.
Criticisms / disadvantages of CGE modelling
• Weak empirical validation (calibration versus econometric estimation)
• The critical role of functional forms• Simplification of exogenous elements of the model• Data requirements• Heavy computational load
2. 2. JRC PESETA II project
EU Adaptation Strategy
• Following the Green Paper (2007) and White Paper (2009) on adaptation, the EU Strategy on Adaptation to Climate Change was adopted in April 2013 (European Commission Communication)
• The JRC PESETA II project provides background evidence on climate impacts in the Impact Assessment of the Communication.
Questions of interest
• What are the climate impacts (reference and 2ºC)
• What are the distributional implications of climate impacts? Fairness and equity issues
• How much adaptation can reduce climate impacts?
• Are spatial (cross-country) spillovers significant?
Integrated, granular modelling
• Multi-disciplinary impact assessment
• Soft-link of models
• High space-time resolution of climate data (T, P, other), common to all impacts (considers spatial correlation)
• Run detailed physical impact models for each impact category
• Integration of market impact results under a Computable General Equilibrium (CGE) model: overall economic effects, direct + indirect; trade effects
Socioeconomic scenario: GDP, population assumptions
Agriculturemodel
Coastal Systems
model
RiverFlooding
model
Tourismmodel
Stage 1:Modelingfuture climate
Physical impacts
agriculture
Physical impacts coasts
Physical impacts floods
Physical impacts tourism
Stage 2:Modelingphysicalimpacts
Stage 3:Modelingeconomicimpacts
Climate model
General Equilibrium model
Climate data(T, P, SLR)
Economicimpacts
Valuationagriculture
impacts
Valuationcoasts impacts
Valuationfloods impacts
Valuationtourism impacts
Socioeconomic scenario: GDP, population assumptions
Agriculturemodel
Coastal Systems
model
RiverFlooding
model
Tourismmodel
Agriculturemodel
Coastal Systems
model
RiverFlooding
model
Tourismmodel
Stage 1:Modelingfuture climate
Physical impacts
agriculture
Physical impacts coasts
Physical impacts floods
Physical impacts tourism
Physical impacts
agriculture
Physical impacts coasts
Physical impacts floods
Physical impacts tourism
Stage 2:Modelingphysicalimpacts
Stage 3:Modelingeconomicimpacts
Climate model
General Equilibrium model
Climate data(T, P, SLR)
Economicimpacts
Climate model
General Equilibrium model
Climate data(T, P, SLR)
Economicimpacts
Valuationagriculture
impacts
Valuationcoasts impacts
Valuationfloods impacts
Valuationtourism impacts
Valuationagriculture
impacts
Valuationcoasts impacts
Valuationfloods impacts
Valuationtourism impacts
Valuationagriculture
impacts
Valuationcoasts impacts
Valuationfloods impacts
Valuationtourism impacts
3 stages in the integration
PESETA II Project strategy
• Building climate impact modeling capabilities at JRC
• Existing data and resources within JRC • Operational and research models • Learning-by-doing within JRC
• To support the EC services on adaptation policy
EU adaptation strategy DG AGRI, CLIMA, ENER, ENV, MOVE, REGIO, Others
Climate models (A1B)
Acronym RCM GCM
C4I-RCA-HadCM3 RCA HadCM3
CNRM-ALADIN-ARPEGE ALADIN ARPEGE
DMI-HIRHAM5-ARPEGE HIRHAM5 ARPEGE
DMI-HIRHAM5-BCM HIRHAM5 BCM
DMI-HIRHAM5_ECHAM5 HIRHAM5 ECHAM5
ETHZ-CLM-HadCM3Q0 CLM HadCM3Q0
KNMI-RACMO2-ECHAM5 RACMO2 ECHAM5
METO-HadRM3Q0-HadCM3Q0 HadRM3Q0 HadCM3Q0
MPI-REMO-ECHAM5 REMO ECHAM5
SMHI-RCA-BCM RCA BCM
SMHI-RCA-ECHAM5 RCA ECHAM5
SMHI-RCA-HADCM3Q3 RCA HADCM3Q3
ensemble of 12 RCM/GCM combinations spatial resolution of 25 km
Climate models (E1)
Acronym RCM GCM
MPI-REMO-ECHAM5-r1 REMO ECHAM5 (BC r1)
MPI-REMO-ECHAM5-r2 REMO ECHAM5 (BC r2)
MPI-REMO-ECHAM5-r3 REMO ECHAM5 (BC r3)
spatial resolution of 50 km 3 runs with same RCM-GCM combination different boundary conditions GCM captures much less uncertainty in future climate for
E1
Reference
Reference variant 1
Reference variant 2
2°C
Northern Europe 3,8 4,8 3,4 3,2UK & Ireland 2,1 2,9 1,7 1,4Central Europe north 2,8 3,7 2,0 2,1Central Europe south 3,0 3,8 2,0 2,1Southern Europe 3,2 3,7 2,4 2,3EU 3,1 3,9 2,4 2,4
Temperature change (°C) in climate runs for 2071-2100, compared to 1961-1990
PESETA II Impact categories
• Sectoral impact categories teams
• - Agriculture physical modelling (IES)• - Agriculture economic modelling (IPTS)• - Forest fires (IES)• - Tree species habitat suitability (IES)• - River flood (IES)• - Tourism (IPTS)• - Energy (IPTS)• - Transport (IPTS)• - Human health (IPTS)
• - Climate tipping points (IPTS)
• CGE modeling for the integration
Sector Input variables Time resolution Spatal ResolutionAverage Temperature Maximum TemperatureAverage PrecipitationMaximum Temperature (June-September)Average Temperature
Tourism Average Temperature Daily NUTS 2 RegionsMaximum air temperatureMinimum air temperatureTotal PrecipitationGlobal solar radiation Air relative humidity maximum and minimumWind speedReference evapotranspirationVapour pressure deficitMaximum and Average TemperaturePrecipitationHumidityWindspeedSolar + thermal radiationAlbedoDewpoint temperatureAverage TemperatureAverage PrecipitationWind SpeedAverage Air Temperature Relative HumidityWind SpeedAverage PrecipitationAverage Temperature Annual; MonthlyMaximum Temperature MonthlyMinimum Temperature MonthlyAverage Precipitation Annual; Monthly
25x25, 50x50 Km
25x25, 50x50 Km
25x25, 50x50 Km
NUTS 2 Regions
25x25, 50x50 Km
25x25, 50x50 Km
Country
Daily
Daily
Daily
Daily
Daily
Annual
Transport
Agriculture
River Floods
Energy
Forest Fires
Forest Species Habitat Suitability
Human Health
Climate data input per impact category
Preliminary economic results
(Impact Assessment of 2013 Adaptation Strategy)
(based on JRC PESETA II, and FP7 ClimateCost results for Agriculture and Coasts)
Shock implementation into GEM-E3
Impact Model implementation Agriculture Yield change Productivity change for crops
Coastal areas
Migration cost Additional obliged consumption
Sea floods cost Capital loss
River
floods
Residential buildings damages
Additional obliged consumption
Production activities losses Capital loss
Energy Heating and cooling demand changes
Households: Change in subsistence energy demand
Service sector: Change in electricity and fuel (per unit of product)
Transport infrastructure
Changes in cost of road asphalt binder application & bridge scouring
Additional obliged consumption
Net change in costs related to extreme flooding & -winter conditions
Capital loss
Forest Fires Burned area damage Capital loss
Reconstruction costs Obliged consumption
Reference run, 2080s(Welfare change, million €)
Source: JRC PESETA II, ClimateCost (agriculture, coasts)
Coastal areas
Energy Agriculture Forest Fires
River Floods
Transport million € €/person
Northern Europe -2,485 2,284 712 1 212 -801 -78 -3UK & Ireland -7,616 8,050 -1,100 5 -2,965 -434 -4,060 -61Central Europe north -21,483 18,762 -4,379 5 -469 -748 -8,310 -56Central Europe south -6,011 6,427 -2,541 -435 -3,210 -874 -6,644 -54Southern Europe -4,659 -31,258 -10,491 -2,419 -1,037 -194 -50,057 -369EU -42,253 4,266 -17,799 -2,844 -7,469 -3,050 -69,149 -138
Impact categories (million €) Sum of impacts
Reference run, 2080s(Welfare change, % of GDP)
Coastal areas
Energy AgricultureForest Fires
River Floods
TransportSum of impacts
Northern Europe -0.4% 0.3% 0.1% 0.0% 0.0% -0.1% 0.0%UK & Ireland -0.4% 0.4% -0.1% 0.0% -0.2% 0.0% -0.2%Central Europe north -0.7% 0.6% -0.1% 0.0% 0.0% 0.0% -0.3%Central Europe south -0.3% 0.3% -0.1% 0.0% -0.1% 0.0% -0.3%Southern Europe -0.2% -1.2% -0.4% -0.1% 0.0% 0.0% -1.9%EU -0.4% 0.0% -0.2% 0.0% -0.1% 0.0% -0.7%
Regional welfare change (%GDP), Reference and 2ºC
-2.00
-1.50
-1.00
-0.50
0.00
0.50
1.00
Re
fere
nce
2°C
Re
fere
nce
2°C
Re
fere
nce
2°C
Re
fere
nce
2°C
Re
fere
nce
2°C
Re
fere
nce
2°C
Northern Europe UK & Ireland Central Europenorth
Central Europesouth
Southern Europe EU
Transport
Coastal areas
Energy
Agriculture
Forest Fires
River Floods
Coastal impacts, 2080s, adaptation(Welfare change, million €)
No Adaptation
Adaptation
Northern Europe -2,485 -43UK & Ireland -7,616 -181Central Europe north -21,483 -844Central Europe south -6,011 -378Southern Europe -4,659 -132EU -42,253 -1,577
Uncertainty: Range of impacts for River Floods (Welfare change, million €)
Worst case Reference Best case
Northern Europe -493 212 -26UK & Ireland -13,462 -2,965 110Central Europe north -3,702 -469 -383Central Europe south -9,818 -3,210 -57Southern Europe -4,489 -1,037 -2,603EU -31,965 -7,469 -2,958
Coast / Central Europe North
Agriculture / Southern Europe
Northern Europe -491 -173UK & Ireland -1,677 -798Central Europe north -20,518 -1,380Central Europe south -1,966 -1,209Southern Europe -1,530 -14,979EU -26,181 -18,540
Transboundary effects(Welfare change, million €)
- JRC PESETA II as a pilot study: soft-linking of JRC models
- Fruitful cooperation within JRC
-Contribution to IA of Adaptation Strategy
Next
1.Water, land use2.Non-market impacts, extremes3.Dynamic perspective: climate change and growth4.Damage function derivation?
Conclusions