How well do GCMs simulate high impact measures of climate change? An assessment of past trends and future projections of agro-climate indices. (Invited)

There is developing scientific consensus that humans are responsible for most of the warming on earth over the last century as measured in the mean global temperature record. What is not clear is how this anthropogenic signal has manifested itself in other aspects of the climate system at regional scales that are more pertinent to human-environment systems than simple measures of mean surface temperatures. Agro-ecosystems are perhaps the most important human-environment system and responses to various climate indices such as growing-degree days are more strongly correlated with biomass production compared to mean temperature. Recent changes in three agro-climate indices (frost days, thermal time, and heat stress index) in North America are examined using observations from the US Historical Climate Network and an ensemble of 17 global climate models (GCMs) from the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Observations from the last 50 years (1956 - 2005) confirm conclusions of previous studies showing continuing declines in the number of frost days and increases in accumulated thermal time. Increases in heat stress are largely confined to the western half of the continent. GCM skill, defined as the ability to reproduce observed patterns (i.e. correlation and error) and variability, is highest for frost days and lowest for heat stress patterns. The lack of agreement between simulated and observed heat stress is relatively robust with respect to how the heuristic is defined and appears to reflect a weakness in the ability of GCMs to reproduce this impact-relevant aspect of the climate system. The results show the extent of the anthropogenic signal throughout the climate system and the potential for future changes in the climate system to significantly affect food production.