Methods of Groundwater Vulnerability and Protectability Assessment - Groundwater Vulnerability: Chernobyl Nuclear Disaster - page 22

map of groundwater contamination with the pesticide. The results obtained in this
work were further analyzed by Loague and Corwin [1998], they came to the
conclusion that 3D modeling using GIS technologies is, in many cases, most effec-
tive for groundwater vulnerability assessment. The GIS provides the direct data
support for modeling (preprocessing, postprocessing, reformatting, mapping,
etc.), especially in the analysis of non-point source vulnerability. It helps to char-
acterize the full information content of the spatially variable data required by
solute transport models.
The same conclusion is made by Zaporozec [1985] as the result of a
groundwater vulnerability assessment in Wisconsin using the SUPRA index-
rating method. He notes that the step after the preliminary assessment should
be development of a regional hydrogeological flow transport model for the
study area.
Statistical models of groundwater flow and transport used for groundwater
vulnerability assessments are in most cases equivalent to deterministic ones
because their general solutions also satisfy the groundwater flow and contami-
nant mass balance equations. However, the problem solution methods are based
on stochastic algorithms such as the Monte Carlo method. Another aspect of
statistical models is represented by the use of special probability density functions
for the solution of groundwater migration problems. For groundwater modeling
applications, this approach was developed by Jury and Roth [1990].
Statistical algorithms and data processing methods (regression analysis, inter-
polation and extrapolation methods, gridding methods, etc.) are directly employed
in groundwater vulnerability assessment by analogy , that is, by associating a given
research area with known areas in which groundwater contamination already
occurred. If the analogue area is the same as the studied area, then we have the
case of groundwater vulnerability assessment by real contamination . For example,
Evans and Maidment [1995] used such a method for a statistical assessment of the
groundwater vulnerability in Texas to nitrate contamination using linear regres-
sion analysis. They built a spatial distribution map for groundwater contamination
probability based on water sampling data from 29,485 wells in the study area. It is
clear that this method requires high volumes of initial information available only
using the monitoring network facilities. An extended review of statistical methods
for groundwater vulnerability assessment is presented in the National Research
Council reports [ NRC , 1993a,b].
Among all groundwater vulnerability assessment methods described above,
in a higher or lesser degree, only a few consider the pathways and zones of
preferential flow and transport. An attempt at the experimental assessment of a
“fast migration component” for Chernobyl-born 137 Cs was implemented by
Rogachevskaya [2002]. In other methodologies the preferential flow phenomena
were taken into account indirectly, particularly in the German, EPIC, PI, and
COP methods for karst areas [ Hoelting et al ., 1995; Doerfliger et al ., 1999;
Zwahlen , 2004].