Lawrence Berkeley National Laboratory

Models simulating multiphase flow and transport in porous media require as input hydrogeologic parameters that are often difficult to obtain from laboratory experiments. Inverse modeling provides the means to estimate model-related, process-specific, and scale-dependent parameters. Multiphase inverse modeling faces various challenges, ranging from computational inefficiency to substantial estimation bias as a result of systematic modeling errors. The purpose of this presentation is to discuss some of these issues by presenting examples of multiphase inverse modeling studies performed at LBNL.