Main content area

A climate generator for agricultural planning in southeastern South America

Greene, Arthur M., Goddard, Lisa, Gonzalez, Paula L.M., Ines, Amor V.M., Chryssanthacopoulos, James
Agricultural and forest meteorology 2015 v.203 pp. 217-228
climate, climate models, crop models, crop yield, markets, planning, summer, uncertainty, South America
A method is described for the generation of climate scenarios in a form suitable for driving agricultural models. The scenarios are tailored to the region in southeastern South America bounded by 25–40° S, 45–65° W, denoted here as SESA. SESA has been characterized by increasing summer precipitation, particularly during the late 20th century, which, in the context of favorable market conditions, has enabled increases in agricultural production. Since about year 2000, however, the upward tendency appears to have slowed or possibly stopped, raising questions about future climate inputs to regional agricultural yields.The method is not predictive in the deterministic sense, but rather attempts to characterize uncertainty in near-term future climate, taking into account both forced trends and unforced, natural climate fluctuations. It differs from typical downscaling methods in that GCM information is utilized only at the regional scale, subregional variability being modeled based on the observational record. Output, generated on the monthly time scale, is disaggregated to daily values with a weather generator and used to drive soybean yields in the crop model DSSAT-CSM, for which preliminary results are discussed. The simulations produced permit assessment of the interplay between long-range trends and near-term climate variability in terms of agricultural production.