The analyses presented in this chapter should be treated as exploratory. Nevertheless, they indicate that climate change probably did affect growing seasons in many low-income countries during the 20th century. They indicate that these changes probably affected food security in these countries as
Figure 4.7 Estimated contribution of climate-induced changes in average growing season to 2012 projections of food distribution gaps in selected low-income countries (percent). 7A. Estimated from regression models using the Southern Oscillation Index for the El Niño/Southern Oscillation cyc le. 7B. Estimated from regression models using the Cold Tongue Index for the El Niño/Southern Oscillation cycle. g
Figure 4.7 Estimated contribution of climate-induced changes in average growing season to 2012 projections of food distribution gaps in selected low-income countries (percent). 7A. Estimated from regression models using the Southern Oscillation Index for the El Niño/Southern Oscillation cyc le. 7B. Estimated from regression models using the Cold Tongue Index for the El Niño/Southern Oscillation cycle. g well. In most of the countries evaluated, climate change is estimated to reduce food security as early as 2012, and in some countries the impacts may be relatively large. The total contribution of climate change to the total food distribution gap for low-income countries, however, is on average probably very small. The analyses suggest that climate-induced changes in growing season probably interacted with the ENSO cycle as well. Some of these interactions would tend to exacerbate extreme events such as droughts in some countries. However, their potential impacts on food security are not explicitly quantified in this analysis.
Several limitations to this analysis merit attention. The scope of the analysis only considers climate change and a subset of extreme weather events. Other potential impacts of rising atmospheric concentrations of greenhouse gases, such as CO2 fertilization and rising sea level, are ignored. In addition, climate change impacts are restricted to long-run trends in length of growing season and their potential interaction on length of growing season during ENSO cycles. Finally, only the relationship between average growing season length and food security was quantified.
Other limitations pertain to the methods used to simulate the agricultural and food security responses to global climate change. First, this analysis only focused on cropland, while climate-induced impacts on growing seasons on permanent pasture were ignored. Second, it was assumed that the relationship between length of growing season and agricultural productivity was the same in all countries included in the analysis. More precise results could be obtained with country-specific relationships. Third, the FSA model does not capture all the potential economic and policy responses available to countries that are vulnerable to global climate change. When climate change reduces agricultural productivity, for example, more land is typically devoted to agriculture than otherwise would be the case (Darwin, 2003). FSA does not take into account feedback loops that generate this behavior.
Projections based on historical trends appear to be promising alternatives for projections based on GCMs, at least for assessing impacts in the immediate future. They do not, however, reduce uncertainty to zero. As indicated in our results, models using different measures of the ENSO cycle generate slightly different results. And projections based on trends during the latter half of the 20th century instead of the whole century may yield different results. However, the prospects for resolving some of these issues are good and once addressed, researchers will have a more reliable method for simulating some important extreme weather-related events.
Was this article helpful?