This repository introduces an efficient surrogate model with the aim of accelerating joint estimation of subsurface geological properties and relative permeability parameters for high-dimensional inversion problems. A combination of model linearization and our recently proposed smooth local parameterization where the Gaussian geological parameters and non-Gaussian facies indicators are locally parameterized forms a set of efficient subdomain linear models which fully replace the high-fidelity model for the forward simulations. This Subdomain Linear Model with Smooth Local Parameterization, referred to as SLM-SLP, allows us to linearize model in each subdomain simultaneously with only a few fidelity model evaluations. Furthermore, we also introduce an adaptive scheme, called weighting smooth local parameterization (WSLP) where the inversion procedure adaptively determines the weighting coefficients and the optimal domain decomposition correspondingly, to mitigate the negative effects of inappropriate domain decomposition strategies.
  
  In this repository includes two folders to separetly construct surrogate model for Gaussian parameters and non-Gaussian binary facies system. Those codes are used for our recent submission titled "Surrogate-Based Joint Estimation of Subsurface Geological and Relative Permeability Parameters for High-Dimensional Inverse Problem by Use of Smooth Local Parameterization" to Water Resources Research.
  
  To run the code, you also need to download MRST (Lie et al, 2012). and put our provided files to the root directory. Please contact Cong Xiao (email: [email protected]) if you have any questions.