The benefits to Mexican agriculture of an El Nino-southern oscillation (ENSO) early warning system

Weather agencies worldwide are attempting to determine if systematic disturbances in climate, such as the El Nino-southern oscillation (ENSO), can be detected far enough in advance so that decisions can be altered to better accommodate these disturbances. Mexico is one country where ENSO-related climatic disturbances have been observed. If climate forecasters were able to disseminate information on upcoming ENSO-induced weather patterns with sufficient lead time, Mexican farmers could adjust by altering a variety of crop decisions, such as growing less (or more) water consumptive crops, planting drought resistant varieties, or altering planting times. This could have a positive impact on crop production, enhancing food security, farmers' incomes, and social welfare. The purpose of this paper is to value such forecasts in a Mexican agricultural setting. To assess the economic consequences of climate arising from various ENSO phases, estimates of regional crop yield sensitivity for key crops were modeled using a crop biophysical simulator. The value of a forecast is then measured by the expected increase in economic benefits due to changes in cropping patterns, production and consumption arising from the yield changes under each ENSO phase forecast. These economic estimates are derived from an economic model of Mexican agriculture. The value of the ENSO information will depend on its accuracy in terms of predictions of the weather consequences of each phase. The economic model is a stochastic, price endogenous, mathematical programming model that represents agronomic and economic conditions in a five-state Mexican region. This model depicts agricultural behavior across the three ENSO phases and provides the basis for calculating the value of information. The benefits of an ENSO early warning system for Mexico is approximately US$ 10 million annually, based on a 51-year time period of ENSO frequencies and when a forecast skill of 70% is assumed. This value translates into an internal rate of return for such an early warning system of approximately 30%. The values for higher skill levels are correspondingly higher. The values estimated here should be viewed as lower bound estimates of the value of an ENSO early warning system because benefits are not estimated for other parts of Mexican agriculture, such as non-commercial (subsistence) agricultural areas, areas where there is only a weak ENSO signal that is not very predictable, and the livestock sector. Also, benefits here do not include benefits that could occur with adjustments in energy generation, water management, or any other economic sectors that may be positively affected by the existence of an ENSO early warning system. (C) 2002 Elsevier Science B.V. All rights reserved.