No description available.

This study utilized (a) actual measured agricultural water use along with (b) geostatistical techniques, (c) crop simulation models, and (d) general circulation models (GCMs) to assess irrigation demand and the uncertainty associated with demand projections at spatial scales relevant to water resources management. In the first part of the study, crop production systems in Southwest Georgia are characterized and the crop simulation model error that may be associated with aggregated model inputs is estimated for multiple spatial scales. In the second portion of this study, a methodology is presented for characterizing regional irrigation strategies in the Lower Flint River basin and estimating regional water demand. Regional irrigation strategies are shown to be well represented with the moisture stress threshold (MST) algorithm, metered annual agricultural water use, and crop management data. Crop coefficient approaches applied at the regional scale to estimate agricultural water demand are shown to lack the interannual variability observed with this novel approach. In the third portion of this study, projections of regional agricultural demand under climate change in the Lower Flint River basin are presented. GCMs indicate a range of possible futures that include the possibility of relatively small changes in irrigation demand in the Lower Flint River basin. However, most of the GCMs utilized in this work project significant increases in median water demand towards the end of this century. In particular, results suggest that peak agricultural water demands in July and August may increase significantly. Overall, crop simulation models are shown to be useful tools for representing the intra-annual and interannual variability of regional irrigation demand. The novel approach developed may be applied to other locations in the world as agricultural water metering programs become more common.