Ensemble-based probabilistic projection is an effective approach to deal with the uncertainties in climate change impact assessments and to inform adaptations. Here, the crop model MCWLA-Wheat was firstly developed by adapting the process-based general crop model, MCWLA [Tao, F., Yokozawa, M., Zhang, Z., 2009a. Modelling the impacts of weather and climate variability on crop productivity over a large area: a new process-based model development, optimization, and uncertainties analysis. Agric. For. Meteorol. 149, 831-850], to winter wheat. Then the Bayesian probability inversion and a Markov chain Monte Carlo (MCMC) technique were applied to the MCWLA-Wheat to analyse uncertainties in parameters estimations, and to optimize parameters. Ensemble hindcasts showed that the MCWLA-Wheat could capture the interannual variability of detrended historical yield series fairly well, especially over a large area. Finally, based on the MCWLA-Wheat, a super-ensemble-based probabilistic projection system was developed and applied to project the probabilistic responses of wheat productivity and water use in the North China Plain (NCP) to future climate change. The system used 10 climate scenarios consisting of the combinations of five global climate models and two greenhouse gases emission scenarios (A1FI and B1), the corresponding atmospheric CO2 concentration range, and multiple sets of crop model parameters representing the biophysical uncertainties from crop models. The results showed that winter wheat yields in the NCP could increase with high probability in future due to climate change. During 2020s, 2050s, and 2080s, with (without) CO2 fertilization effects, relative to 1961-1990 level, simulated wheat yields would increase averagely by up to 37.7% (18.6%), 67.8% (23.1%), and 87.2% (34.4%), respectively, across 80% of the study area; simulated changes in evaportranspiration during wheat growing period would range generally from -6% to 6% (-0.6% to 10%), from -10% to 8% (-1.0% to 17%), and from -17% to 4% (7-12%), respectively, across the study area. Further analyses suggested that the improvements in heat and water resources and rising atmospheric CO2 concentration ([CO2]) could contribute notably to wheat productivity increase in future. Climate change could enhance the development and photosynthesis rate; however the duration of reproductive period could be less affected than that of vegetative period, and wheat productivity could benefit from enhanced photosynthesis due to climate change and rising [CO2]. Furthermore, wheat could become mature earlier, which could prevent it from severe high temperature stress. Our study parameterized explicitly the effects of high temperature stress on productivity, accounted for a wide range of crop cultivars with contrasting phenological and thermal characteristics, and presented new findings on the probabilistic responses of wheat productivity and water use to climate change in the NCP.
