Scenarios for Amazon future
Eustáquio J. Reis IPEA
18th LBA-SSC MeetingSão Paulo, 14-15 November 2005
Modeling issues:environmental consequences (Bosello and Zang 2005)
Predicted changes in climate variables: GHG and CO2 emissions, temperature and precipitation
Temperature affects production decisions directly or through changes in water availability and in biodiversity (insects variety, plant diseases, weed infestation)
Evapotranspiration and precipitation affects soil moisture and erosion
Biophysical reactions of crops to climate changes: C02 fertilization (smaller in tropical crops)
Feedback of agriculture on GHG emissions small (except for NOx CH4 in rice and cattle raising and CO2 emission from deforestation)
Modeling issues: Socio-economic adaptations
Microeconomic adaptation Choice of techniques: factor intensity (land/labor),
cultivation timing, mix and location, irrigation Technologies (R&D): development of cultivar adapted
to new climate Macroeconomic adjustment
Price, income and wealth (land price) effects Changes in production, consumption and trade Migration and capital flows: regional, national and
international
Role of policies
Modeling approaches: Structural, experimental-simulation or bottom-up
Experimental model of crop response (plant physiology and vegetation distribution)
Extrapolation to the relevant universe Model of human or socio-economic
adaptations and responses appended
Modeling approaches: Spatial analogue, top-down or Ricardian Statistical inferences based upon geographical
cross-section of climate conditions Adaptations and responses, both natural and
socioeconomic, are automatically incorporated Criticisms (Schneider 1997):
equivalence of time and space sufficient information on climate data as well as other
variables across space (edaphic, technology, infrastructure, etc.)
extreme events and abrupt changes unicity of steady states
Empirical results: crucial issues and assumptions Dumb farmer hypothesis x socioeconomic
adaptation (Rosenweig et al. 1993, Reilly 1994, Mendelsohn et al. 1994, 1999, 2005)
CO2 fertilization accounted General/global x partial/regional equilibrium
models: substitution and different agents Time of evaluation (higher GDP in the future) Sustainability, vulnerability and uncertainty:
effects of extreme events and abrupt changes in climate (Reilly 1999, Schimmelpfenning et al. 1996)
Empirical results: main findings (Bosello and Zang 2005, Mendelshon 2005) Small impact of 2 x CO2 on world agriculture
Food production (-2.5% to –0.07%) Welfare (-0.047% to 0.01%)
Higher value for specific regions: Welfare (-5.48% to + 2.73%)
Crucial role of adaptation: hill shaped damage functions (CO2 fertilization)
Equity: vulnerability of low latitude developing countries: geography and lower capacity to adapt
Extreme events: effects become negative above Δ3% C.
Damage functions are hill shaped: increases in temperature have positive benefits at first (Mendelsohn, 2005)
Empirical results:the Brazilian case
RegionCrops/Sector
Effects of a doubling of CO2 (+Δ 2.5ºC and +7%
precip. ) on: ReferenceΔ% land value
Brazil (1970-85)Cerrado GO, MT, TO, RO MGRGS, SC
-2.6% to –7.4%
-3.67% to – 18.44%-2,99% to –16.58%+0.80% to +4.66%
Sanghi et al. 1997
Modeling deforestation:basic assumptions Exogenous drivers (or structural causes)
of deforestation Population Roads
Agricultural land uses are sources (or proximate cause) of deforestation
Logging caused or induced by deforestation and thus plays a subsidiary role in the model (very questionable assumption)
Modeling deforestation:crucial issues Demographic transition and urbanization
smaller long run rates of population growth Population density higher price of land
intensification of land use saturation effects in deforestation
Roads lower transport cost higher price of land intensification of land use
Feedback of climate on land yield, uses of land and settlement in AML
Broader geographical perspective of models
EXOGENOUSVARIABLES================
POPULATIONUrbanization___________________
TECHNOLOGYAgric. productivity___________________
INFRASTRUTURERoadsPortsHealthEducationEnergy==================
AGROCLIMATICCONDITIONSVegetationSoil qualityClimate
ENDOGENOUS VARIABLES==============
FACTOR PRICESWagesLand pricesTransport costs________________ FACTOR USESLabor employment
Land use:Crop areaPasture areaFallow landsLogging
POPULATION GROWTH AND
INFRASTRUCTEINVESTMENT
DEFORESTATIONBiomass contentcarbon stocks in
soil and vegetation
CO2 EMISSIONS
Production fuction: 2nd law of thermodynamics applied to
economics: conservation of economic value
Output = F(Land, Labor, Roads, Temperature, Precipitation, etc.)
Land = G(Output, Labor, Roads, Temperature, Precipitation, etc.)
Land Yield = G(Output, Labor, Roads, Temperature, Precipitation, etc.)
Scenarios Labor Population scenario Land agro-ecological zoning Roads infrastructure policies Precipitation and temperature from
climate models
Methodological strategy Lack of historical data on relationship
between climate and economic activity Spatial analogue spatial cross-section
at municipal or Census tract Panel data 260 municipios in AML and
3660 in Brazil from 1960 to 2000 Census tract 1995 and 1985 (non-
georeferenced)
Modeling deforestation:building blocks Investment in municipal roads (proxy for
infrastructure) is a policy decision Dynamics of municipal (rural and urban)
population is determined by agro-ecological and socio-economic conditions in previous periods
Agricultural land use and yields are determined by profit maximization in hirearchical model
Logging is a function of deforestation Distribution of deforestation according to
vegetation types Dynamics of land use (includeing fallow areas)
and carbon stocks
Infrastructure investment and population dynamics
Period t
Population
Population
InfrastructureRoads
Investments in infrastructure
(policy decision)
InfrastructureRoads
Population growth
•Socio-economic conditions•Land availability (zoning policy)•Infrastructure conditions
Period t+1
Demand for agricultural
land in period t
Supply of agricultural
land in period t-1
Clearing is necessary ?( period t )
Yes No
Recovered fallowareas available?
Fallow areas =Recoveredvegetation. Yes
No
Clear fallow areas
Emission of CO2
Deforestation of pristine forest
Absorption of CO2
Economic Socio demographic
Geo-ecologic conditions
Dynamics of carbon stocks
Source: Author´s simulation
Simulation 1985-2010 - Legal Amazonia
Source: Author´s simulation
Simulation 1985-2010 - Legal Amazonia
Source: Author´s simulationObs: The values of CO2 in 1985 are estimated
Simulation 1985-2010 - Legal Amazonia