Reliability of Contaminant Transport Modeling on Vadose Zone Sampling Methods in Structured Soils

摘要

Development and validation of predictive models is highly dependent on the type (sampling method) and robustness (spatial and temporal frequency) of available data. Thus, emphasis was placed on contaminant sampling methods and the use of models to accurately predict contaminant fate in field soils. Suction cup lysime-ters, vapor samplers and soil cores, were used to extract water samples from the vadose zone of structured soils. Tritium composition of soil core samples was determined at random intervals near the sampling clusters. A mass balance of the applied tritium and precipitation was performed. Soil profile tritium analysis and breakthrough curves were used to indicate the predominant flowpaths sampled by each technique. The fractions of the tritium sampled by each sampling method were predicted with: 1) a simple local mass balance of water and solutes transport model (Burns method), 2) the steady-state con-vective-dispersive model (HYDRUS 1-D), and 3) the field-based Cornell preferential flow model. Comparison of results from these three models with actual data from two field plots demonstrate how different field sampling methods can affect modeling results.