Hierarchical Bayesian model averaging for hydrostratigraphic modeling: Uncertainty segregation and comparative evaluation

摘要: [1] Analysts are often faced with competing propositions for each uncertain model component. How can we judge that select a correct proposition(s) an component out of numerous possible propositions? We introduce the hierarchical Bayesian averaging (HBMA) method as multimodel framework uncertainty analysis. The HBMA allows segregating, prioritizing, and evaluating different sources their corresponding through hierarchy BMA models forms tree. apply to conduct analysis on reconstructed hydrostratigraphic architectures Baton Rouge aquifer-fault system, Louisiana. Due in data, structure, parameters, multiple produced calibrated base models. study considers four uncertainty. With respect data uncertainty, two calibration sets. three variogram models, geological stationarity assumptions fault conceptualizations. following combinatorial design allow segregation. Thus, these components result 24 results show systematic dissection along detecting robust major