Projecting yield changes of spring wheat under future climate scenarios on the Canadian Prairies

Abstract

The potential impact of the rise in atmospheric CO2 concentration and associated climatic change on agricultural productivity needs assessment. Projecting crop yield changes under climate change requires future climate scenarios as in- put to crop yield models. It is widely accepted that downscal- ing ofclimate data isrequiredtobridgethe gap between large- scale global climate models (GCMs) and climate change im- pact models, such as crop growth models. Regional climate models (RCMs) are often used to dynamically downscale GCM simulations to smaller regional scales, while statistical methods, such as regression-based transfer functions and sto- chastic weather generators, are also widely employed to de- velop future climate scenarios for this purpose. The methods usedindeveloping future climate scenarios often contributeto uncertainties in the projected impacts of climate change, in additiontothoseassociatedwithGCMsandforcingscenarios. We employed climate scenarios from the state-of-the-art RCMs in the North American Regional Climate Change As- sessmentProgram(NARCCAP), along with climatescenarios generated by a stochastic weather generator based on climate change simulations performed by their driving GCMs, to drive the CERES-Wheat model in DSSAT to project changes inspringwheatyieldontheCanadianPrairies.Thefuturetime horizon of 2041-2070 and the baseline period of 1971-2000 were considered. The projected changes showed an average increase ranging from 26 to 37 % of the baseline yield when the effects of the elevated CO2 concentration were simulated, but only up to 15 % if the elevated CO2 effect was excluded. In addition to their potential use in climate change impact assessment, the results also demonstrated that the simulated crop yield changes were fairly consistent whether future cli- mate scenarios were derived from RCMs or they were gener- ated by a stochastic weather generator based on the simulated climate change from the GCMs that were used to drive the RCMs, in this case, when they were compared for regional averages.