Abstract: Crop production in the Northeast Farming Region of China (NFR) is affected considerably by variation in climatic conditions. Data on crop yield and weather conditions from a number of agro-meteorological stations in NFR were used in a mixed linear model to evaluate the impacts of climatic variables on the yield of maize (Zea mays L.), rice (Oryza sativa L.), soybean (Glycine max L. Merr.) and spring wheat (Triticum aestivum L.) in different crop growth phases. The crop growing season was divided into three growth phases based on the average crop phenological dates from records covering 1981 and 2010 at each station, comprising pre-flowering (from sowing to just prior to flowering), flowering (20 days around flowering) and post-flowering (10 days after flowering to maturity). The climatic variables were mean minimum temperature, thermal time (which is used to indicate changes in the length of growth cycles), average daily solar radiation, accumulated precipitation, aridity index (which is used to assess drought stress) and heat degree-days index (HDD) (which is used to indicate heat stress) were calculated for each growth phase and year. Over the 1961–2010 period, the minimum temperature increased significantly in each crop growth phase, the thermal time increased significantly in the pre-flowering phase of each crop and in the post-flowering phases of maize, rice and soybean, and HDD increased significantly in the pre-flowering phase of soybean and wheat. Average solar radiation decreased significantly in the pre-flowering phase of all four crops and in the flowering phase of soybean and wheat. Precipitation increased during the pre-flowering phase leading to less aridity, whereas reduced precipitation in the flowering and post-flowering phases enhanced aridity. Statistical analyses indicated that higher minimum temperature was beneficial for maize, rice and soybean yields, whereas increased temperature reduced wheat yield. Higher solar radiation in the pre-flowering phase was beneficial for maize yield, in the post-flowering phase for wheat yield, whereas higher solar radiation in the flowering phase reduced rice yield. Increased aridity in the pre-flowering and flowering phases severely reduced maize yield, higher aridity in the flowering and post-flowering phases reduced rice yield, and aridity in all growth phases reduced soybean and wheat yields. Higher HDD in all growth phases reduced maize and soybean yield and HDD in the pre-flowering phase reduced rice yield. Such effects suggest that projected future climate change may have marked effects on crop yield through effects of several climatic variables, calling for adaptation measures such as breeding and changes in crop, soil and agricultural water management.

Abstract: Highlights • Regularization techniques for regression outperformed the classical regression techniques in predicting crop yields. • Different regression techniques with similar prediction accuracy showed different responses of major climatic variables to crop yield. • The regression models showed some responses of crop yield to climatic conditions that is mostly absent in process based crop models. Abstract Statistical regression models represent alternatives to process-based dynamic models for predicting the response of crop yields to variation in climatic conditions. Regression models can be used to quantify the effect of change in temperature and precipitation on yields. However, it is difficult to identify the most relevant input variables that should be included in regression models due to the high number of candidate variables and to their correlations. This paper compares several regression techniques for modeling response of winter oilseed rape yield to a high number of correlated input variables. Several statistical regression methods were fitted to a dataset including 689 observations of winter oilseed rape yield from replicated field experiments conducted in 239 sites in Denmark, covering nearly all regions of the country from 1992 to 2013. Regression methods were compared by cross-validation. The regression methods leading to the most accurate yield predictions were Lasso and Elastic Net, and the least accurate methods were ordinary least squares and stepwise regression. Partial least squares and ridge regression methods gave intermediate results. The estimated relative yield change for a +1°C temperature increase during flowering was estimated to range between 0 and +6 %, depending on choice of regression method. Precipitation was found to have an adverse effect on yield during autumn and winter. It was estimated that an increase in precipitation of +1 mm/day would result in a relative yield change ranging from 0 to −4 %. Soil type was also important for crop yields with lower yields on sandy soils compared to loamy soils. Later sowing was found to result in increased crop yield. The estimated effect of climate on yield was highly sensitive to the chosen regression method. Regression models showing similar performance led in some cases to different conclusions with respect to effect of temperature and precipitation. Hence, it is recommended to apply an ensemble of regression models, in order to account for the sensitivity of the data driven models for projecting crop yield under climate change.