Data from ten monitoring wells in a region of suspected underground pollution are used to assess the source, time, and amount of pollutant released into the ground. The chemicals are sorted based on changes recorded in their concentrations over time to determine which were active pollutants during the data collection period and to account for discrepancy caused by an incomplete data set. Those chemicals found to be active during this time period were observed to change concentration simultaneously, indicating that each chemical was a component of a single leaking liquid involved in two major spills. The concentrations of selected active chemicals were then combined to form a composite indicator whose concentration value was found at each well on each date. A graph showing this concentration at each well (see appendix 4) revealed that two spills occurred, the first between July 1991 and March 1993, and the second between January 1995 and April 1997, possibly continuing until the end of the data collection period. The primary chemical constituents of the leaking liquid are included in appendix 2.

A Delaunay triangulation is used to interpolate a gradient of concentration for the composite indicator at each date between the monitoring wells. Given that the general flow of groundwater in this region is directed towards well # 9, the time and location of the pollution source can be approximated based on changes in the concentration gradient over time. This spill is estimated to have originated in the region surrounding the point (8000, 4500). Following the initial triangulation, Voronoi Polygons are used to construct a convex hull representing the total volume and position of the spill (the volume of the contaminated area). This polygon is comprised of smaller segments, each of a specific uniform concentration. The program Geomview is used to generate graphics of these polygons and convex hulls. A volume can be calculated at each concentration and ultimately the total volume of polluting liquid can be found if the concentration of the composite indicator in the original polluting liquid is known.

Finally, various testing and interpretation methods are explored and incorporated into a procedure for evaluating underground pollution. Each method is discussed in terms of its application to the scenario in problem two, and uses information given in the data set to test the validity of the method.