Search

There is an urgent need to better understand and predict crop yield responses to weather disturbances, in particular, of extreme nature, such as heavy precipitation events, droughts, and heat waves, to improve future crop production projections under weather variability, extreme events, and climate change. In this paper, we develop quantile regression models for estimating crop yield probability distributions depending on monthly temperature and precipitation values and soil quality characteristics, which can be made available for different climate change projections. Crop yields, historical and those simulated by the EPIC model, are analyzed and distinguished according to their levels, i.e., mean and critical quantiles. Then, the crop yield quantiles are approximated by fitting separate quantile-based regression models. The developed statistical crop yield meta-model enables the analysis of crop yields and respective probabilities of their occurrence as a function of the exogenous parameters such as temperature and precipitation and endogenous, in general, decision-dependent parameters (such as soil characteristics), which can be altered by land use practices. Statistical and machine learning models can be used as reduced form scenario generators (meta-models) of stochastic events (scenarios), as a submodel of more complex models, e.g., Integrated Assessment model (IAM) GLOBIOM. [ABSTRACT FROM AUTHOR]

The authors discuss connections between the problems of two-stage stochastic programming, robust decision-making, robust statistical estimation, and machine learning. In the conditions of uncertainty, possible extreme events and outliers, these problems require quantile-based criteria, constraints, and "goodness-of-fit" indicators. The two-stage stochastic optimization (STO) problems with quantile-based criteria can be effectively solved with the iterative stochastic quasigradient (SQG) solution algorithms. The SQG methods provide a new type of machine learning algorithms that can be effectively used for general-type nonsmooth, possibly discontinuous, and nonconvex problems, including quantile regression and neural network training. In general problems of decision-making, feasible solutions and concepts of optimality and robustness are characterized from the context of decision-making situations. Robust machine learning (ML) approaches can be integrated with disciplinary or interdisciplinary decision-making models, e.g., land use, agricultural, energy, etc., for robust decision-making in the conditions of uncertainty, increasing systemic interdependencies, and "unknown risks." [ABSTRACT FROM AUTHOR]

The paper presents a consistent algorithm for regional and sectoral distributed models' linkage and optimization under asymmetric information based on iterative stochastic quasigradient (SQG) solution procedure of, in general, nonsmooth nondifferentiable optimization. The procedure is used for linking individual sectoral and regional models for integrated and interdependent Food–Energy–Water–Environmental security analysis and management. [ABSTRACT FROM AUTHOR]