Session 7.10 Effect of future climatic variability on agriculture in a Mediterranean region

Wednesday, 25 August 2004: 2:00 PM
Pierpaolo Duce, CNR-Institute of Biometeorology, Sassari, Sardinia, Italy; and A. Arca, S. Canu, D. Spano, and A. Motroni

Presentation PDF (1.3 MB)

Considerable effort has been expended in studies on a large-scale and general circulation modeling to assess climatic risk. However, downscaling to the local level to assess climate risk for agricultural areas and crops has proven difficult. In addition, there have been few attempts to account for inter-annual climate variability. To overcome these deficiencies, a project to incorporate year-to-year climate variability with land evaluation (LE) was conducted. Land evaluation provides qualitative information about land, such as its cropping potential or land vulnerability risk, based on bio-physical and socio-economic characteristics. In general, land qualities derived from measurements of dynamic variables (e.g. climate data) are converted to static variables (means) for the purposes of LE. Observed and future climate variability data were combined with geographic and soil information from Sardinia, Italy using a Land Capability for Agriculture (LCA) classification system, which classifies agricultural land into a range of quality and potential productivity. The climatic variability LCA classification was based on maximum soil moisture deficit and heat accumulation from the reference period 1961-2000. A climatic risk index was developed analyzing the inter-annual variability of maximum soil moisture deficit for the period 1961-2000 and the results were compared with climatic risk index values calculated for the period 2000-2099 from two levels of CO2 increase using a general circulation model. The analysis showed that future temperature and rainfall regimes could cause an increase in climatic risk affecting mainly the most important agricultural areas of Sardinia.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner