The Value of Seasonal Productivity Forecasting in Biodiesel Plans

摘要

Crop productivity is commonly assumed as a deterministic function when developing udagricultural plans. Actual data prove however that, even for the same soil at the same location, crop udproductivity can be better interpreted as a random variable due to the meteorological conditions of the udspecific year. For the production of biodiesel, crops are easily substitutable and the farmer can chose udevery year between various alternatives. Without information on the seasonal meteorology, the udfarmers select the crop to cultivate mainly on the basis of the expected productivity. However, udchanges in the meteorological situation may result in a reduction in crop profitability. As a result, uda crop, that on average is less interesting, may become the best choice in a specific year. Given that udseasonal forecasts based on long range climatic variables, such as ENSO, are becoming available, udthe paper examines their effectiveness in biodiesel production plans, with reference to an area in Mato Grosso, Brazil. We formulate and solve a mathematical programming problem to determine the most efficient crop plan under different scenarios: (i) no information about the seasonal meteorology, (ii) perfect information and (iii) meteorological forecasts with different precision. This allows us to udquantitatively evaluate how important the availability of seasonal productivity forecasting might be and udshows that even a rough seasonal forecast, if systematically applied, may improve the average udproduction and reduce the risk of traditional agricultural decisions.