Hydrology

Abstract

Whereas methods for porous-media models are developed or undergoing rapid development, fractured-media models are not as well defined. For porous media, deterministic procedures may be useful if data are extensive, and may be especially useful if inverse methods are used to incorporate site-specific data. Monte Carlo procedures may be used to determine the release probabilities and bounds on estimated probability values. Subjective probability estimates should be clearly and explicitly stated. Stochastic models are an attractive alternative to deterministic porous-media models, because they include the inherent variability of the media. These may be combined with parameter information of the type discussed above and with geostatistical inverse procedures to perform site-specific evaluations. Monte Carlo analyses may be used both to estimate probabilities and to incorporate data via conditional simulations. These models include the variability inherent in nature more explicitly. The best fractured-media models are stochastic models. -from Author