ABSTRACT
Climate predictions are notoriously low in resolution, with cells commonly containing numerous countries, hundreds of agroecological zones, and countless variations in topography and surface type. The Intergovernmental Panel on Climate Change (IPCC) predictions for areas such as Southeast Asia or north Asia, for example, give single generic values for temperature and rainfall changes across areas from hundreds up to thousands of kilometers across. These gures have become soundbite statistics that we have often seen, but which are so nonspecic and so unrelated to the diverse agricultural areas they represent such that their continued citing in the literature is perhaps becoming more of a hindrance than a help. For scientists whose work relies on locationspecic climate information, it is simply impossible to accurately base agricultural research or choose interventions for specic locations based on these predictions.
