Wallach, D., & Rivington, M. (2015). Identification and quantification of differences between models (Vol. 6).
Abstract: A major goal of crop model inter-comparison is model improvement, and an important intermediate step toward that goal is understanding in some detail how models differ, and the consequences of those differences. This report is intended as a first attempt at describing possible techniques for relating differences between model outputs to specific aspects of the models. No Label
|
Wallach, D., & Rivington, M. (2014). A framework for assessing the uncertainty in crop model predictions (Vol. 3).
Abstract: It is of major importance in modeling to understand and quantify the uncertainty in model predictions, both in order to know how much confidence to have in those predictions, and as a first step toward model improvement. Here we show that there are basically three different approaches to evaluating uncertainty, and we explain the advantages and drawbacks of each. This is a necessary first step toward developing protocols for evaluation of uncertainty and so obtaining a clearer picture of the reliability of crop models. No Label
|
|
Wallach, D., & Rivington, M. (2014). A framework structure to integrate improved methods for uncertainty evaluation, and protocols for methods application (Vol. 3).
|
Wallach, D., & Rivington, M. (2013). Development of a common set of methods and protocols for assessing and communicating uncertainties (Vol. 2).
Abstract: This reports sets out an outline approach to create definitions of uncertainty and how it might be classified. This is not a prescriptive approach rather it should be seen as a starting point from which further development can be made by consensus with CropM partners and across MACSUR Themes. We propose both a numerical quantification of uncertainty and text based classification scheme. The rational is to be able to both establish the terms and definitions in quantifying the impact of uncertainty on model estimates and have a scheme to enable identification of connectivity between types and sources of uncertainty. The aim is to establish a common set of terms and structure within which they operate that can be used to guide work within CropM. No Label
|
Wallach, D., Thorburn, P., Asseng, S., Challinor, A. J., Ewert, F., Jones, J. W., et al. (2016). Estimating model prediction error: Should you treat predictions as fixed or random. Env. Model. Softw., 84, 529–539.
Abstract: Crop models are important tools for impact assessment of climate change, as well as for exploring management options under current climate. It is essential to evaluate the uncertainty associated with predictions of these models. We compare two criteria of prediction error; MSEPfixed, which evaluates mean squared error of prediction for a model with fixed structure, parameters and inputs, and MSEPuncertain(X), which evaluates mean squared error averaged over the distributions of model structure, inputs and parameters. Comparison of model outputs with data can be used to estimate the former. The latter has a squared bias term, which can be estimated using hindcasts, and a model variance term, which can be estimated from a simulation experiment. The separate contributions to MSEPuncertain(X) can be estimated using a random effects ANOVA. It is argued that MSEPuncertain(X) is the more informative uncertainty criterion, because it is specific to each prediction situation.
|
