13.2.5 Current adaptation

Weather and climate variability forecast

The mega 1982/83 El Niño set in motion an international effort (the Tropical Ocean-Global Atmosphere (TOGA) programme) to understand and predict this ocean-atmosphere phenomenon. The result was the emergence of increasingly reliable seasonal climate forecasts for many parts of the world, especially for Latin America. These climate forecasts became even more reliable with the use of TOGA observations of the Upper Tropical Pacific from the mid-1990s, although they still lack the ability to correctly predict the onset of some El Niño and La Niña events (Kerr, 2003). Nowadays such forecasting systems are based on the use of coupled atmospheric-ocean models and have lead times of 3 months to more than 1 year. Such climate forecasts have given rise to a number of applications and have been in use in a number of sectors: starting in the late 1980s for fisheries in the Eastern Pacific and crops in Peru (Lagos, 2001), subsistence agriculture in north-east Brazil (Orlove et al., 1999), prevention of vegetation fires in tropical South America (Nepstad et al., 2004; http://www.cptec.inpe.br/), streamflow prediction for hydropower in the Uruguay river (Tucci et al., 2003; Collischonn et al., 2005), fisheries in the south-western Atlantic (Severov et al., 2004), dengue epidemics in Brazil (IRI, 2002), malaria control (Ruiz et al., 2006) and hydropower generation in Colombia (Poveda et al., 2003).

Agriculture is a key sector for the potential use of ENSO-based climate forecasts for planning production strategies as adaptive measures. Climate forecasts have been used in the north-east region of Brazil since the early 1990s. During 1992, based on the forecast of dry conditions in Ceara, it was recommended that crops better suited to drought conditions should be planted, and this led to reduced grain production losses (67% of the losses recorded for 1987, a year with similar rainfall but without climate forecasting). However, this tool has not yet been fully adopted because of some missed forecasts which eroded the credibility of the system (Orlove et al., 1999). Recently, in Tlaxcala (Mexico), ENSO forecasting was used to switch crops (from maize to oats) during the El Niño event (Conde and Eakin, 2003). This successful experience was based on strong stakeholder involvement (Conde and Lonsdale, 2005). Recent studies have quantified the potential economic value of ENSO-based climate forecasts, and concluded that increases in net return could reach 10% in potato and winter cereals in Chile (Meza et al., 2003); 6% in maize and 5% in soybean in Argentina (Magrin and Travasso, 2001); more than 20% in maize in Santa Julia, Mexico (Jones, 2001); and 30% in commercial agricultural areas of Mexico (Adams et al., 2003), when crop management practices are optimised (e.g., planting date, fertilisation, irrigation, crop varieties). Adjusting crop mix could produce potential benefits close to 9% in Argentina, depending on site, farmers’ risk aversion, prices and the preceding crop (Messina, 1999). In the health sector, the application of climate forecasts is relatively new (see Section 13.2.5.5). Institutional support for early warning systems may help to facilitate early, environmentally-sound public health interventions. For instance, the Colombian Ministry of Health developed a contingency plan to control epidemics associated with the 1997/98 El Niño event (Poveda et al., 1999).

In some countries of Latin America, improvements in weather-forecasting techniques will provide better information for hydrometeorological watching and warning services. The installation of modern weather radar stations (with Doppler capacity) would improve the reliability of these warnings, but the network is still very sparse (WMO, 2007). Furthermore, the deficiencies in the surface and upper air networks adversely affect the reliability of weather outlooks and forecasts. Nevertheless, the exacerbation of weather and climate conditions and the problems arising from extreme events have led to planning and implementation actions to improve the observation, telecommunications and data processing systems of the World Weather Watch (WWW). Moreover,the participation of Latin American countries in the UN-IDSR would lead to the implementation of new (and further development of existing) monitoring and warning services in the region. Examples of networks that predict seasonal climate and climate extremes are the Regional Disaster Information Centre-Latin America and Caribbean (CRID), the International Centre for Research on El Niño Phenomenon (Ecuador), the Permanent Commission of South Pacific (CIIFEN; CPPS) and the Andean Committee for Disaster Prevention and Response (CAPRADE). Some networks set up to respond to and prevent impacts are, for example, the multi-stakeholder decision-making system developed in Peru (Warner, 2006), the National Development Plan and the National Risk Atlas implemented in Mexico (Quaas and Guevara, 2006) and the communication programme for indigenous populations, based on messages in the local language (Alcántara-Ayala, 2004).