The Regional Impacts of Climate Change

Other reports in this collection

6.2.4. Climate Change

Annex B of this report provides information on the progress made with transient runs of atmosphere-ocean general circulation models (AOGCMs), which allow-within certain limits and with reliable complementary data-climate projections at a regional scale. Annex B also provides information on different regionalization techniques, which recently have been developed and tested to improve the coarse resolution in regional climate change simulation. However, the regions identified for this type of simulation do not include Latin America.

Whetton et al. (1996) analyzed the performance of models in South America to assess the ability of GCMs to reproduce key climate features of the Southern Hemisphere, as well as the major gaps in their simulation capability. The authors considered several "mixed-layer ocean" experiments and several "coupled" or "full dynamic ocean" experiments. The correlation between modeled and observed annual mean precipitation over the region ranges from 0.35 to 0.70 in mixed-layer GCM runs; slightly higher correlations are found in coupled GCM climate experiments.

In greenhouse climate experiments performed over South America and adjacent oceans (Labraga, 1997; Labraga and Lopez, 1997), mean temperature increases simulated in mixed-layer equilibrium experiments ranged between 1.5C and 4.0C at the time of CO2 doubling; transient experiments with coupled GCMs simulated mean temperature increases of 1.2-1.7C. According to the results of five coupled GCMs, temperatures in the semi-arid zone of central Chile and central-western Argentina during the Southern Hemisphere summer are simulated to increase by 1-3C at the time of CO2 doubling. During the same season, precipitation in the area was projected to decrease by 10-15% per degree of global warming, according to the same set of experiments. There is consensus among models that the semi-arid subtropical zone will experience intensified and extended dry conditions. In addition, the coupled GCMs all project increases in rainfall in the ITCZ in the eastern equatorial Pacific and the northWestern part of the continent, the South Atlantic Convergence Zone (SACZ) in the eastern part of Brazil and the adjacent Atlantic Ocean, and the southern tip of the continent.

Figure 6-2: Seasonal change in surface temperature from 1880-1889 to 2040-2049 in simulations with aerosol effects included. Contours are at every 1C (IPCC 1996, WG I, Figure 6.10).


Figures 6-2 and 6-3 reproduce maps showing expected seasonal changes in surface temperatures and precipitation, as projected by transient coupled AOGCM experiments performed at the Max Planck Institute (MPI, Germany) and the Bureau of Meteorology Research Centre (BMRC, Australia) (IPCC 1996, WG I, Chapter 6).

Figure 6-3: Seasonal changes in precipitation at the time of doubling CO2 following a 1%/yr increase. Contours are at 0.5, 1, 2, and 4 mm/day; negative contours are dashed and areas of increase areas stippled (IPCC 1996, WG I, Figure 6.11).

Because of the great climatic diversity of this region and the present limitations of GCMs in simulating Latin America's regional climate, many studies that have estimated vulnerability have used the GCMs as tools for obtaining reference scenarios, in order to perform sensitivity analyses (e.g., for different crops at specific sites and water resources for specific basins). These analyses have included detailed knowledge of the local climate founded on sufficient and reliable real data and correctly evaluated proxy data. A good example of this approach appears in a study on "Global Warming and Climate Change in Mexico" (Liverman and O'Brien, 1991), in which the authors note the importance of acknowledging differences among climate model projections-as well as between model simulations and observed climates-because these differences underscore the uncertainties involved in assessing the regional impacts of global warming. A range of possible outcomes is captured by using the results of different GCMs and undertaking a sensitivity analysis of the results. Some of the conclusions reached appear in the relevant sections of this chapter, which also provide information on the vulnerabilities of and potential impacts on natural and managed systems and human health in Latin America as a result of global warming. In this context, it is very important to realize that climate change is only one among many causes leading to the changes that have been observed in these systems during recent decades. In fact, many factors-such as limited access to technology, poor research capacity, economic crises, social inequality, and population growth-could give rise to larger effects than those stemming from atmospheric warming.

Other reports in this collection