Abstract. A methodology and hypothetical case study are presented for incorporation of uncertainty and variability into calculations of human health risk appropriate for regional, or basin-scale, groundwater management problems. Uncertainty in well water concentration is introduced through complex contaminant migration patterns in the subsurface. Variability is considered in parameters related to individual behavior patterns and biological effects and to groundwater extraction and distribution networks. A joint uncertainty and variability (JUV) analysis is used to generate a two-dimensional distribution or risk surface that spans both transport uncertainty as well as individual variability. Cuts in this distributional surface (fractiles of variability and percentiles of uncertainty) are presented and discussed. Comparisons with alternative approaches based upon deterministic transport models are also made. In addition, important distinctions are made between the case where household water is derived from the nearest well and the case where household water is mixed from many wells in a distribution system. IntroductionGroundwater is an important natural resource in the United States, comprising 40% of all public supply water and over This paper develops a new integrated methodology of linkedgroundwater contaminant transport and human exposure that includes uncertainty and variability. This methodology explicitly addresses the variability in multipathway exposure and treats groundwater transport in regional scale, heterogeneous aquifers. Important physical features (such as wells and resultant well distribution systems) are included, as is the uncertainty in geophysical parameters. The methodology is also used to study parameter effects over a wide range of heterogeneity. An idealized groundwater contamination problem is then analyzed using the methodology, and insights regarding different water management strategies are described. BackgroundUnder the Safe Drinking Water Act (SDWA) of 1974, 1986, and 1995, the U.S. Environmental Protection Agency (USEPA) is required to regulate levels of contaminants in water supplies in order to protect public health. As part of this protection, the USEPA establishes a nonenforceable maximum contaminant goal level (MCGL) and an enforceable maximum contaminant level (MCL) for contaminants considered to have adverse health effects [Pontius, 1990a, b]. In addition, states may have their own standards. In planning the restoration or remedial action for contaminated groundwater, target cleanup levels may be established either in terms of 833

A framework of simulation models for including human exposure to contaminants in regional scale aquifer management problems is presented. The framework includes horizontal flow and transport of contaminant plumes in the aquifer and multiple‐pathway human exposure. Well water from the aquifer simulation model is used as the source of contaminant in the human exposure model. The exposure pathways considered for regional analysis are ingestion of foods grown using well water as part of the irrigation supply; ingestion and dermal absorption of contaminants in tap water; and inhalation of vaporized contaminants while bathing. An environmental compartment model is used to track the contaminant in irrigation water into the soil layers in contact with food products. The simulation framework is demonstrated in a study of the sensitivity of exposure to various aquifer and water supply parameters. The region used is hypothetical; however, the parameters are typical of California.

System Engineering tools have been used extensively to model the technical characteristics of large systems or processes. However, little use of these tools has been made in the social sciences, and the modeling of the social factors that contribute to design choices-in the case of systems that involve political processes-remains relatively crude. This paper discusses the efforts of the Center for Nuclear and Toxic Waste Management, University of California, Berkeley to employ system design tools as part of social science research into US surplus plutonium disposition options.