A hybrid regularized inversion methodology for highly parameterized environmental models

By M. Tonkin, John Doherty. 2005

234 citations recorded by the source corpus.

Subfields: Environmental Engineering, Geochemistry and Petrology, Geophysics

View Paper: https://www.semanticscholar.org/paper/c2edb349233f5d66a8a2d666f35747774b82ceb2

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

In Geophysics

Citations in the corpus

On the use of spatial regularization strategies to improve calibration of distributed watershed models (2010, methodology): “Other strategies for constraining the parameter space could include the use of interior point data [Khu et al., 2008], soft information [Seibert and McDonnell, 2002; Dunn, 1999], and manual calibration within a Tikhonov regularization framework [Tikhonov and Arsenin, 1977; Tonkin and Doherty, 2005 ”
Efficient nonlinear predictive error variance for highly parameterized models (2006, methodology): “This enables nonlinear predictive error variance analysis of highly parameterized models to be completed efficiently following a hybrid truncated singular value decomposition (TSVD)-Tikhonov calibration [Tonkin and Doherty, 2005].”
Sensitivity analysis and parameter estimation for distributed hydrological modeling: potential of variational methods (2009, unknown): “As emphasised by Tonkin and Doherty (2005), the subspace determined from the truncated singular value decomposition of the Jacobian (TSVD) is determined from the observations information content whereas the subspace constructed from a prior parsimony strategy is not.”
High‐dimensional posterior exploration of hydrologic models using multiple‐try DREAM(ZS) and high‐performance computing (2012, methodology): “It required a joint use of manual and computer-based parameter estimation methods, including the covariance matrix adaptation evolutionary strategy (CMAES) [ Hansen et al. 2003], truncated singular value decomposition (SVD) [ Tonkin and Doherty , 2005; Marquardt , 1963], and automatic user intervent”
A comparison of seven methods for the inverse modelling of groundwater flow. Application to the characterisation of well catchments (2009, methodology): “The hybrid regularised inversion methodology [127] obtains a ‘single best’ solution (including uncertainty estimates) that also displays details of small-scale variability.”
Real‐time groundwater flow modeling with the Ensemble Kalman Filter: Joint estimation of states and parameters and the filter inbreeding problem (2008, methodology): “…captures some of the details at smaller scales, namely those which are important for groundwater flow and transport predictions, and also estimates an uncertainty associated with the prediction [e.g., Hernandez et al., 2003; Vermeulen et al., 2004; Tonkin and Doherty, 2005; Tonkin et al., 2007].”
Sensitivity-guided reduction of parametric dimensionality for multi-objective calibration of watershed models (2009, background): “gineering, Pennsylvania State given the amount of information contained in hydrological data [3,35,43].”
A spatial regularization approach to parameter estimation for a distributed watershed model (2008, methodology): “Such strategies have been used to address high dimensional inverse modeling problems in hydrology [see, e.g., Skaggs and Kabala, 1998; van Loon and Troch, 2002; Tonkin and Doherty, 2005; Doherty and Skahill, 2005, etc.].”
Efficient posterior exploration of a high‐dimensional groundwater model from two‐stage Markov chain Monte Carlo simulation and polynomial chaos expansion (2013, methodology): “This is particularly true for CPU-intensive hydrogeologic models, and hence MCMC simulation becomes a daunting task [e.g., Tonkin and Doherty, 2005, 2009; Keating et al., 2010].”
Calibration‐constrained Monte Carlo analysis of highly parameterized models using subspace techniques (2009, methodology): “…(5) can be reformulated as follows: D ¼ TV1 ð10aÞ XtQX ffi V1 ZtQZþ b2DtSD ð10bÞ where V1 is obtained through TSVD of X tQX as described above, and the m x ns matrix Z contains the calculated sensitivities of model outputs with respect to superparameters [see Tonkin and Doherty, 2005, equation 12].”
The Future of Sensitivity Analysis: An essential discipline for systems modeling and policy support (2020, unknown): “These methods work by restricting resources to identifying and exploiting subspaces that are informed by data and can reduce the computational cost of the inverse problem by orders of magnitude (Tonkin and Doherty, 2005; Spantini et al., 2015).”
Optimization and uncertainty assessment of strongly nonlinear groundwater models with high parameter dimensionality (2010, methodology): “As an additional efficiency measure for calibration, we used the “SVD‐ Assist” technique described by Tonkin and Doherty [2005] to minimize the objective function; this gains efficiency through directly estimating linear combinations of parameters spanning the calibration solution space, thereby e”