Abstract: Statistical learning is a fairly new term referring to a set of supervised and unsupervised modelling and prediction techniques. It is based on traditional statistics but has been highly enhanced inspired by developments in machine learning and data mining. The main focus of statistical learning is to estimate the functions that quantify relations between several parameters and observed responses. These functions are further used for prediction, inference or a combination of both. For a particular case of quantitative responses, regularization techniques in regression are developed to overcome the weaknesses of ordinary least square (OLS) regression in prediction. These new shrinkage methods outperform OLS for prediction, especially in large datasets. In this study, a large dataset of field experiments on winter oilseed rape in Denmark for 22 years (1992 to 2013) was collected. Biweekly climatic data along with sowing date, harvest date, soil type and previous crop are considered as the explanatory variables. Yield of winter oilseed rape is considered as response variable. LASSO and Elastic Nets are the regularization techniques used to estimate the functions. Hold-one-out cross validation method for testing the prediction power reveals that these techniques are much useful in both prediction and inference. Since these techniques are included in recent versions of some software packages (e.g. R), they can be easily implemented by users at any level. The estimated function (model) is further used to predict the oilseed rape yield responses to climate change for several scenarios using representative weather data produced by a weather generator.

Abstract: Mitigation of greenhouse gas emissions from agriculture should be assessed across cropping systems and agroclimatic regions. In this study, we investigate the ability of the FASSET model to analyze differences in the magnitude of N2O emissions due to soil, climate and management factors in cereal-based cropping systems. Forage maize was grown in a conventional dairy system at Mabegondo (NW Spain) and wheat and barley in organic and conventional crop rotations at Foulum (NW Denmark). These two European sites represent agricultural areas with high and low to moderate emission levels, respectively. Field trials included plots with and without catch crops that were fertilized with either mineral N fertilizer, cattle slurry, pig slurry or digested manure. Non-fertilized treatments were also included. Measurements of N2O fluxes during the growing cycle of all the crops at both sites were performed with the static chamber method with more frequent measurements post-fertilization and biweekly measurements when high fluxes were not expected. All cropping systems were simulated with the FASSET version 2.5 simulation model. Cumulative soil seasonal N2O emissions were about ten-fold higher at Mabegondo than at Foulum when averaged across systems and treatments (8.99 and 0.71 kg N2O-N ha(-1), respectively). The average simulated cumulative soil N2O emissions were 9.03 and 1.71 kg N2O-N ha(-1) at Mabegondo and at Foulum, respectively. Fertilization, catch crops and cropping systems had lower influence on the seasonal soil N2O fluxes than the environmental factors. Overall, in its current version FASSET reproduced the effects of the different factors investigated on the cumulative seasonal soil N2O emissions but temporally it overestimated emissions from nitrification and denitrification on particular days when soil operations, ploughing or fertilization, took place. The errors associated with simulated daily soil N2O fluxes increased with the magnitude of the emissions. For resolving causes of differences in simulated and measured fluxes more intensive and temporally detailed measurements of N2O fluxes and soil C and N dynamics would be needed. (C) 2015 Elsevier B.V. All rights reserved.

Abstract: The complexity of risks posed by climate change and possible adaptations for crop production has called for integrated assessment and modelling (IAM) approaches linking biophysical and economic models. This paper attempts to provide an overview of the present state of crop modelling to assess climate change risks to food production and to which extent crop models comply with IAM demands. Considerable progress has been made in modelling effects of climate variables, where crop models best satisfy IAM demands. Demands are partly satisfied for simulating commonly required assessment variables. However, progress on the number of simulated crops, uncertainty propagation related to model parameters and structure, adaptations and scaling are less advanced and lagging behind IAM demands. The limitations are considered substantial and apply to a different extent to all crop models. Overcoming these limitations will require joint efforts, and consideration of novel modelling approaches. No Label