Global Assessment Study

Figure 12.1 summarizes the overall approach used in the integrated global analysis of the impacts of climate change and climate variability on the availability and use of water and the production and consumption of food. At the core of the approach is the IMPACT model, which is a representation of a competitive world agricultural market for crops and livestock. It is specified as a set of country or regional submodels, within each of which supply, demand and prices for agricultural commodities are determined. The country and regional agricultural submodels are linked through trade, a specification that highlights the interdependence of countries and commodities in the global agricultural markets. World agricultural commodity prices are determined annually at levels that clear international markets.

In this study, IMPACT has been supported by two additional models that provided forcing data to drive it. The first model is a global hydrology model 'IMPACT-WATER' that uses first-order data (climate, land cover, soil type, etc.) specified at a 0.5° X 0.5° spatial grid for all global land points (excluding the Arctic and Antarctica) to produce estimates of river basin runoff over a 30-year time horizon and on a monthly time step. In the IMPACT-WATER model, water is represented as a scarce resource needed for agricultural production. Water demands evaluated in the water model include irrigation, livestock, domestic, industrial and committed flow for environmental, ecological and navigational uses. The river basin runoff data are used by ADAPT to define the water supply to each of the hydro-economic zones. A second external model, the Global Agro-Ecological Zones (GAEZ) model is used to determine crop potential yield. The GAEZ provides a standardized framework for the characterization of climate, soil and terrain conditions relevant to agricultural production, most notably the estimate of maximum potential crop yield in a gridded format that can be used by IMPACT. In GAEZ, crop modelling and environmental matching procedures are used to identify crop-specific limitations due to climate, soil and terrain, under assumed levels of inputs and management conditions. The GAEZ model was derived from Fischer et al. (2003).

Since a keen interest of this study is adaptation, the lower right corner of Fig. 12.1 expresses the implicit adaptation that can be accounted for in the IMPACT model. IMPACT is a representation of a competitive world agricultural market for tradable crops and livestock, which determines supply, demand and prices for these commodities and determines their price such that international markets clear. Thus, different climate or socio-economic scenarios trigger autonomous adaptation implicit in the IMPACT model results. Autonomous adaptation implies either gradual or abrupt changes in food production, such as increases in acreage, changes in water use, changes in cropping patterns over time, that are result of the dynamics of a competitive food market, constrained by production factors such as a limited water supply or limited agricultural capital (labour, land, mechanization, etc.).

Although the IMPACT model will account for autonomous adaptation, IMPACT is also capable of describing the response of global water use and food production to exogenous adaptation strategies such as improved irrigation efficiency, changes in farm subsidies, capital investments, etc. These external or exogenous variables would be prescribed, and would comprise individual scenarios whose results would be examined relative to baseline, non-exogenously driven scenarios. For this study, we have focused on autonomous adaptations, with the hopes of extending the work to include scenarios of explicit, exogenous adaptation strategies in the future.

The IMPACT model divides the world into 69 broad hydro-economic regions, with three of these regions (the USA, China and India) divided more finely into their major river basins. For each basin, the mean monthly runoff derived for the period 1961-1990 was compared with an estimate of monthly mean runoff from Alcamo et al. (2000), which was used to calibrate the global water balance used in this study (Yates, 1996). The runoff model was calibrated by comparing the 'observed' monthly average discharge from the Alcamo dataset, to model estimates of runoff, where

Table 12.1. HadCM3 and ECHAM4 global temperature projections (A1 and B2).

A1 Global mean temperatures relative to 1990 (°C)

2000

2010

2020 2030 2040 2050 2060 2070 2080

2090

2100

HadCM3 ECHAM4

0.570 0.379

0.730 0.488

0.989 1.385 1.863 2.253 2.695 3.081 3.383 0.671 0.962 1.315 1.595 1.923 2.209 2.434

3.621 2.614

3.823 2.770

B2 Global mean temperatures relative to 1990 (°C)

2000

2010

2020 2030 2040 2050 2060 2070 2080

2090

2100

HadCM3 ECHAM4

0.570 0.379

0.839 0.577

1.163 1.468 1.768 2.070 2.362 2.653 2.943 0.813 1.033 1.251 1.472 1.686 1.900 2.115

3.229 2.328

3.510 2.538

monthly runoff is simply the accumulation of the surface and subsurface components. Calibration of the regions and basins that comprise the 69 hydro-economic zones was done by enumerating values that would minimize the RMSE error between the 'observed' mean monthly discharge, and the estimated discharge from the model for each of the 69 basins. Thus calibrations for each basin result in a unique correlation value for each land cover within that basin.

0 0

Post a comment