Role of the calibration process in reducing model predictive error

By Catherine R. Moore, John Doherty. 2005

261 citations recorded by the source corpus.

Subfields: Environmental Engineering, Geochemistry and Petrology

View Paper: https://www.semanticscholar.org/paper/33ade20e75c15e1eaab5edeaace34b43bce68b6a

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Environmental Engineering

In Geochemistry and Petrology

Citations in the corpus

Calibration‐constrained Monte Carlo analysis of highly parameterized models using subspace techniques (2009, methodology): “where V1 is obtained through TSVD of X t QX as described above, and the mxn smatrix Z contains the calculated sensitivities of model outputs with respect to superparameters [see Tonkin and Doherty, 2005, equation 12].”
A hybrid regularized inversion methodology for highly parameterized environmental models (2005, methodology): “For example, where a model is used to predict contaminant movement, predictions may depend upon aquifer heterogeneity [Moore and Doherty, 2005].”
Predictive uncertainty analysis of a saltwater intrusion model using null‐space Monte Carlo (2011, methodology): “Consequently, model predictions are always in error, especially those that depend on hydraulic property detail that cannot be inferred through the calibration process [ Moore and Doherty, 2005, 2006 ].”
Modeling Soil Processes: Review, Key Challenges, and New Perspectives (2016, methodology): “If MCMC is not feasible, uncertainty quantification is still possible when assuming that all model parameters and measurement errors follow multiGaussian distributions (at least after transformation) and that the model equations can be linearized, and then using linear error propagation (Moore and D”
Sensitivity analysis of environmental models: A systematic review with practical workflow (2014, unknown): “…case when the output y is the mean squared error and gradient-based local sensitivity (see Section 3.1) is an approximation of the curvature (Hessian matrix) of y evaluated at x (for practical examples see for instance Sorooshian and Gupta (1985) or the PEST approach by Moore and Doherty (2005)).”
Inverse methods in hydrogeology: Evolution and recent trends (2014, background): “…method,361 like the PiPM, only optimizes a few parameters (simple model) in the groundwater flow model, which is362 commonly insufficient for transport predictions such as travel times that require a higher level of system363 detail, as demonstrated in the case study by Moore and Doherty (2005).”
Efficient nonlinear predictive error variance for highly parameterized models (2006, methodology): “Parameter error, p p, incurred through calibration is expressed by [Moore and Doherty, 2005] p p ¼ I Rð Þp Ge; ð3Þ where R is the resolution matrix.”