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.

Abstract: Climate change scenarios of seasonal minimum and maximum temperature over different Italian agricultural areas, during the periods 2021–2050 and 2071–2100 against 1961–1990, are assessed. The areas are those selected in the framework of the Agroscenari project and are represented by: Padano–Veneta plain, Marche, Beneventano, Destra Sele, Oristano, Puglia and Sicilia, all areas of prominent agricultural vocation with excellence productions. A statistical downscaling technique applied to ENSEMBLES global climate simulations, emission scenario A1B, is used to achieve this objective. The statistical scheme consists of a multivariate regression based on Canonical Correlation Analysis. The scheme is constructed using large-scale fields derived from ECMWF reanalysis and seasonal mean minimum, maximum temperature derived from national observed daily gridded data that cover 1959–2008 period. Once the most skillful model has been selected for each season and variable, this is then applied to GCMs of ENSEMBLES runs. The statistical downscaling method developed reveals good skill over the case studies of the present work, underlying the possibility to apply the scheme over whole Italian peninsula. In addition, the results emphasize that the temperature at 850 hPa is the best predictor for surface air temperature. The future projections show that an increase could be expected to occur under A1B scenario conditions in all seasons, both in minimum and maximum temperatures. The projected increases are about 2 °C during 2021–2050 and between 2.5 and 4.5 °C during 2071–2100, respect to 1961–1990. The spatial distribution of warming is projected to be quite uniform over the territory to the end of the century, while some spatial differences are noted over 2021–2050 period. For example, the increase in minimum temperature is projected to be slightly higher in areas from northern and central part than those situated in the southern part of Italian peninsula, during 2021–2050 period. The peak of changes is projected to appear during summer season, for both minimum and maximum temperature. The probability density function tends to shift to warmer values during both periods, with increases more intense during summer and to the end of the century, when the lower tail is projected to shift up to 3 °C and the upper tail up to 6 °C. All these projected changes have important impacts on viticulture, intensive fruit and tomatoes, some of the main agricultural systems analyzed in the Agroscenari project.

Abstract: The changing climate and environmental conditions play a key role on plant physiology. In this context, crop simulation models represent a useful tool for investigating the main plant processes and provide a reliable estimation of crop productivity and quality. However, the most common crop models showed many limitations, with particular concern on the effect of some meteorological variables on plant processes during sensitive stages of development. Improving models by implementing the effect of such variables on crop processes may help to improve the accuracy of models, thus their usefulness. Here we focus on the analysis of the effect of high and low temperatures during flowering in grapevine. To this, the fruit-set index, developed for taking into account for the effect of temperature on setting the number of berries per cluster and the fruit-set percentage, was applied in a preliminary explorative study to assess the impact of different conditions during flowering at European scales. The sensitivity of the index allowed to identify the differential impact of temperature around flowering in different environment and for different varieties. Once meteorological variables are available at field or sub-field scale, the index can be used to provide information about the spatial variability of crop growth, thus allowing to identify the most appropriate interventions to improve productivity.