Page 100 - Climate Change and Food Systems

climate change and food systems: global assessments and implications for food security and trade
5. Economic models
Economic models build on the results of climate and pathway models to describe the economic impacts of climate change. Different modelling types are applied depending on the research question, the scale (farm, sector, national or global) and the actors targeted (farmers, resource managers, firms, consumers, governments). For expositional purposes, we separate the economic models between sectoral and household (or
farm) types. We focus on agriculture and derive the implications for food security, trade and other important drivers.
5.1 Market- and sector-level models
Partial equilibrium models
Partial equilibrium models describe one or more sectors of an economy in detail, while holding price and quantities in the rest of the economy constant. Their main advantage in climate change analysis
is that they typically include a finely disaggregated set of agricultural sectors, which facilitates their integration with detailed crop yield models. The chief limitation of partial equilibrium models is
that they do not account for linkages between agriculture and the rest of an economy and so they miss potentially important economy-wide impacts of climate change that may originate from agriculture or feedback to agriculture.
■ Crops
Multi-market models are a type of partial equilibrium model that focuses on a subset of interdependent markets in an economy; for example, crop and livestock markets that are linked through acreage allocations and feed demand. Two examples illustrate the type of analysis with this model category. We start with a widely cited multi-market model used to assess climate
change impacts on agriculture and food security – the International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT), developed at the International Food Policy
Research Institute (IFPRI), (Rosegrant, et al.2002; Nelson, et al. 2010). The model describes 32 crop and livestock commodities in 281 food-producing units (FPUs) over a 40-year projection horizon. The FPUs are major river basin areas that are disaggregated from the 115 countries/regions
of the model, based on climate and hydrological variations. Crop production is aggregated up from FPU to national and regional levels at which demand and trade are modeled.
Crop production in the IMPACT model is the product of yield times area. Changes in yields over time are a function of exogenous yield growth rates, which vary by crop and location, water availability for irrigated crops, and prices. Area is a function
of exogenous area growth rates, water availability for irrigated crops and prices. Domestic demand for food, feed, biofuels and other uses is a function of prices, per capita income and population. Countries and regions are linked through trade as the model iteratively solves for each annual set of equilibrating world prices that achieves zero net global trade; these prices are then transmitted back to producers and consumers.
To develop global projections of climate change impacts on agricultural supply, Nelson et al. (2010) link the IMPACT model with DSSAT crop models calibrated to site-specific locations. To address uncertainty on the supply side, they simulate
crop growth based on four climate scenarios
that describe 2000 and 2050 climates using
the SRES A1B and B1 outputs of the Model for Interdisciplinary Research on Climate (MIROC) and Commonwealth Scientific and Industrial Research Organisation (CSIRO) GCMs, plus a fifth scenario in which historical climate continues (“perfect mitigation”). They also run 30 iterations of DSSAT models for each location, based on stochastically generated daily weather data for 2000 and 2050 climates. The mean yield outcomes are then converted to smooth, linear growth rates that are used to adjust the productivity growth coefficient in the yield equations over the 2010-2050 period. DSSAT crop simulation models are carried out
for five crops (rice, wheat, maize, soybeans and groundnuts), and yield results are mapped to other
80