Environmental data often have features that are distinct from data in
other branches of science. These features include spatial and/or tempo
ral auto-correlation, natural heterogeneity, measurement errors, small
sample sizes, and simultaneous existence of different types and quali
ties of data. Realistic environmental modeling requires simulation pro
cedures that account for all of these features. In this study, a model
of uncertainty analysis, BUDA, is used to account for the noted featu
res and provide a unified framework for quantification, propagation, a
nd reduction of uncertainty, The BUDA model is used to analyze the dev
elopment of a chloride plume around an old landfill to the year 2020,
This article describes the different components of BUDA as they relate
to the landfill application.