ABSTRACTThe theoretical shape of a zone of contaminated ground water (henceforth called an enclave) can be predicted from a knowledge of the three‐dimensional, ground‐water flow pattern. The reasoning in LeGrand (1965) and Sendlein and Palmquist (1973) suggests that the horizontal shape of the enclave is a flame‐like plume extending from the source, parallel to the ground‐water flow lines, in a downflow direction. According to LeGrand, the ultimate size of the enclave will depend upon the relative rates of decay, diffusion, dilution and absorption of the contaminants. Similar reasoning suggests that, in vertical section, the boundaries of the enclave will follow the ground‐water flow lines between contamination source and ground‐water discharge site. This reasoning suggests that the three‐dimensional shape of an enclave is that of a tongue‐like lobe, the length, width and depth of which increases with increasing distance from and increasing height above the discharge site.In Iowa, the hydrology of refuse sites emplaced in alluvium has been analyzed to determine the size and shape of the associated enclaves. The refuse sites range in size from 13 to 117 acres (5.25 to 47.3 hectares), in age from 9 to 40 years, and in topographic position from floodplain adjacent to river to terrace. The ground‐water quality and water‐table elevations were determined from multiple nests of piezometers (from depths of 15 to 45 feet) (4.7 to 13.6 meters) around the refuse sites. The surficial geology of the sites was established from both borehole and geophysical data.The Iowa data suggests that the enclave in alluvium is plume‐shaped with the long axis parallel to the ground‐water flow lines and extending from the refuse site to the nearest stream. The SO4enclave at the Ames site, for example, is over 7000 feet (2121 meters) long, 4500 feet (1362 meters) wide and extends to a maximum depth of 60 feet (18.2 meters). The highest SO4concentrations are along the axis of the enclave at a depth of 30 feet (9.09 meters). The concentrations decrease with distance along the axis, laterally away from the axis and vertically away from the axis, such that the enclave is entirely surrounded by uncontaminated ground water except at the source. Analysis of the variation in water quality data with time indicates that most of the enclaves are not increasing in size but have achieved their maximum size and are in a steady‐state equilibrium condition.The Iowa studies indicate that the size and shape of the contamination enclave resulting from refuse disposal sites can be predicted from the initial geohydrologic conditions and that it may become possible in the near future to estimate the concentrations within the enclave at any point in time and space. These possibilities open the way toward a strategy of minimizing potential contamination of aquifers through selective refuse site placement on the floodplain. The Iowa data indicates that floodplain sites may be desirable for disposal sites because of the predictability of enclave shape within alluvium, the tendency for floodplains to be ground‐water discharge sites and the low concentrations of the leachate produced in a high ground‐water flow environment.The results of this study strongly indicate that any monitoring system around a refuse disposal site be judiciously placed. Not only must the location of the wells be considered but also the depth of wells. To monitor maximum leachate concentrations, wells must be located in a downflow direction and along the axis of the enclave. Likewise, wells which are too shallow or too deep will miss the core of an enclave and yield leachate samples with low concentrations. It thus becomes necessary to estimate the size and shape of the enclave prior to establi