are also used for some of the other analyses, in the spatial operational unit of IMPACT-WATER. Projections for water requirements, infrastructure capacity expansion, and water use efficiency improvement are conducted by IMPACT-WATER. These projections are combined with the simulated hydrology to estimate water use and consumption through water system simulation by IMPACT-WATER (Rosegrant et al., 2002).

5.3.2.7Energy use and production

As discussed in Chapter 4, the energy sector may develop in very different ways. For the reference projection, we have chosen to loosely couple future outcomes to IEA reference scenario—a scenario that lies central in the range of avail­able energy projections. The policy variant has been devel­oped using the IMAGE/TIMER model and incorporates the specific assumptions of the IAASTD reference projec­tion with respect to economic growth and land use change. In terms of energy demand growth the IEA scenario is a mid-range scenario compared to full range of scenarios pub­lished in literature. For the development of the energy mix, it is a conventional development scenario assuming no ma­jor changes in existing energy policies and/or societal prefer­ences. These assumptions are also included in the IAASTD reference projection.

5.3.2.8 Climate change

Climate change is both driving different outcomes of key variables of the reference run (like crop productivity and water availability) and is an outcome of the agricultural projections of the reference run, through land-use changes and agricultural emissions, mainly from the livestock sec­tor (FAO, 2006b). Given the medium energy outcomes in the reference run (see 5.3.3.3), results from the B2 scenario are directly used in most of the modeling tools. From the available B2 scenario, the ensemble mean of the results of the HadCM3 model for B2 scenario was used. The pattern scaling method applied was that of the Climate Research Unit, University of East Anglia. The "SRES B2 HadCM3" climate scenario is a transient scenario depicting gradually evolving global climate from 2000 through 2100. In the IM­AGE model, climate change is an output of the model. The IMAGE model uses a global climate model (MAGICC) to calculate global mean temperature change—and uses down-scaling techniques to downscale this data to a 0.5 x 0.5 grid. Through this approach, different GCM results can be used to assess the consequences of the uncertainty in local cli­mate change. For the reference run, the pattern of Hadley Centre's HadCM2 is used for the downscaling approach, which is consistent with the pattern used in the other model­ing tools. For the simulations of the reference world, the me­dium climate sensitivity value is used of the Third Assessment Report (2.5°C), which has been adjusted slightly in the latest IPCC report. According to IPCC, the climate sensitivity is likely to be in the range of 2 to 4.5°C with a best estimate of about 3°C, and is very unlikely to be less than 1.5°C (IPCC, 2007). Climate sensitivity is not a projection but is defined as the global average surface warming following a doubling of carbon dioxide concentrations (IPCC, 2007). The uncertain­ties in the climate sensitivity are not assessed in the reference world. Specific sensitivity analyses will show the importance of the uncertainties in values of the climate sensitivity.

5.3.3 Description of reference world outcomes

5.3.3.1 Food sector

Food supply and demand. In the reference run, global food production increases  1.2%  per year during 2000-2050. This growth is a result of rapid economic growth, slowing population growth, and increased diversification of diets. Growth of demand for cereals slows during 2000-2025 and again from 2025-2050, from 1.4% per year to 0.7% per year. Demand for meat products (beef, sheep, goat, pork, poultry) grows more rapidly, but also slows somewhat after 2025, from 1.8% per year to 0.9% annually.

     Changes in cereal and meat consumption per capita vary significantly among IAASTD regions (Figures 5-2 and 5-3). Over the projection period, per capita demand for cereals as food declines in the LAC region and in the ESAP region. On the other hand demand is projected to considerably in­crease in the sub-Saharan Africa region and also increase in the NAE and CWANA regions. Recovery and strengthen­ing of economic growth in sub-Saharan Africa will drive relatively fast growth in regional demand for food. In de­veloping countries and particularly Asia, rising incomes and rapid urbanization will change the composition of cereal demand. Per capita food consumption of maize and coarse grains will decline as consumers shift to wheat and rice. As incomes rise further and lifestyles change with urbaniza­tion, there will be a secondary shift from rice to wheat. In the SSA region, growing incomes are expected to lead to a shift from roots and tubers to rice and wheat. Per capita food demand for roots and tubers in SSA is projected to de­cline from 171 kg to 137 kg between 2000 and 2050, while rice and wheat demand are expected to grow from 18-20 kg to 30-33 kg (Table 5-4). Under the reference run, the composition of food demand growth across commodities is expected to change considerably. Total cereal demand is projected to grow by 1,305 million tonnes, or by 70%; 50% of the increase is expected for maize; 23% for wheat; 10% for rice; and the reminder, for sorghum and other coarse grains.

     Demand for meat products continues to grow rapidly across all six IAASTD regions, by 6-23 kilograms per per­son. The increase is fastest in the LAC and ESAP regions and slowest in the SSA and NAE regions. Rapid growth in meat and milk demand in most of the developing world will put strong demand pressure on maize and other coarse grains as feed. Globally, cereal demand as feed increases by 553 million tonnes during 2000-2050, a staggering 42% of total cereal demand increase (Figure 5-4).

     Tables 5-5, 5-6, 5-7 and 5-8 present results for changes in livestock numbers for beef, sheep and goats, pigs, and poultry, respectively, for the IAASTD regions. The global population of bovines is projected to increase from some 1.5 billion animals in 2000 to 2.6 billion in 2050 in the refer­ence run. Substantial increases are projected to occur in all regions except NAE: the number of bovines is projected to double in CWANA and ESAP, and to increase by 50% in SSA, for example. Cattle numbers are projected to peak in SSA in about 2045. Bovine populations are relatively stable in NAE to 2050 in the reference run.

     Similar patterns are seen for changes in sheep and goat populations. The global population is expected to increase