Evaluation of the Rackwitz‐Fiessler Uncertainty Analysis Method for environmental fate and transport models

Abstract

The use of contaminant transport modeling has become an integral component of the regulatory and decision process for the disposal and cleanup of hazardous wastes. Because many of the input parameters to these models are uncertain, analysis of this uncertainty and its impact on the decision process has become increasingly important. Many contaminant transport models are computationally intensive and require run times that make traditional Monte Carlo analysis impractical. This paper therefore evaluates and tests an approximate technique, the Rackwitz‐Fiessler method, that can be used when computation time prohibits the use of Monte Carlo simulation. The accuracy and efficiency of the method is assessed and compared to Monté Carlo simulation for three contaminant transport models.

Illustration of Sampling‐Based Methods for Uncertainty and Sensitivity Analysis

“…Many approaches are available for uncertainty and sensitivity analysis, including differential analysis, ( 102–115 ) response surface methodology, ( 116–126 ) the Fourier amplitude sensitivity test (FAST), ( 127–131 ) variance decomposition, ( 132–141 ) and fast probability integration. ( 142–148 ) Differential analysis involves approximating a model with a Taylor series and then using variance propagation formulas to obtain uncertainty and sensitivity analysis results. Response surface methodology is based on using classical experimental designs to select points for use in developing a response surface replacement for a model; this replacement model is then used in subsequent uncertainty and sensitivity analyses based on variance propagation and Monte Carlo simulation.…”

Sensitivity and Uncertainty of Ground‐Water Discharge Estimates for Semiarid Shrublands1

“…, 2000). Fast probability integration is mainly used to estimate the tails of uncertainty distributions of model predictions (Rackwitz and Fiessler, 1978; Chen and Lind, 1983; Wu and Wirsching, 1987; Schanz and Salhotra, 1992). Differential analysis approximates a model with a Taylor series and then uses a variance propagation approach to analyze sensitivity (Tomovic and Vukobratovic, 1972; Frank, 1978; Dougherty et al.…”