of groundwater for local wells originates relatively near the well locations. Each traceline on
figure 2 represents the three-dimensional path of a single hypothetical "particle" of water from the
water table to a particular well. The collection of particle pathlines for each well outlines the
steady-state capture zone for that well. In figure 2, these three-dimensional traces are projected
onto a two-dimensional map, resulting in some apparent angular refractions of the particle tracks
where the particles are moving vertically from one model layer to the next. All traces on figure 2
are steady-state particle paths, meaning that these are the paths that groundwater would take
assuming that the groundwater system is steady (ie recharge is constant, pumping rates and water
levels are unchanging) for an infinitely long period of time. Therefore, the length of a particular
pathline is not necessarily related to the velocity of the particle, although in general the longer
paths are associated with longer travel times. Steady-state travel times for many wells from
recharge to entry into the well are on the order of several thousand years.

Maps of the ZOCs and ZOIs of each individual well in Dane County (Appendix F) provide more
detail and also constrain the steady-state capture zones to 5-, 10-, and 100-year flow boundaries.
Such boundaries should be more useful for WHPA delineation in real human terms. For each of
these wells, the ZOC is best viewed in a three-dimensional sense as an ellipse with a cross-section
represented by the shaded area of the ZOC and extending from the water table to the bottom of
the deep sandstone aquifer. It is interesting to note that many of the time-constrained ZOCs are
rather small, and that even the travel paths for the 100-year flow boundaries are only on the order
of one or two miles long.

Accuracy of the capture zone delineations

The accuracy of the locations of the capture zones delineated in figure 2 and in Appendix F
depends on the accuracy of the groundwater flow-model and of the field data and data
interpretations used to construct it. The MODFLOW and PATH3D codes themselves are
mathmatically very precise, and numerical errors associated with these codes are probably
insignificant. However, the calibration of the groundwater flow model (the "fit" of the model to
observed field data) is not perfect, although it is considered good from a groundwater modeling
standpoint (Krohelski and Bradbury, in preparation). In general, the model results are probably
most precise in areas where hydrogeologic data are abundant, such as in the Madison  0
metropolitan area. The model is less accurate in areas where hydrogeologic data are sparse, such
as in western Dane County, where very few deep water wells exist. Also, the precision of the
capture zone delineations decreases with increasing travel time. Therefore, the time-constrained
capture zones shown in Appendix F are much more precise than are the long tails of the steady-
state particle paths show in figure 2.
All capture zones show in this report assume steady-state conditions, meaning that groundwater
levels and recharge rates do not change with time. In areas where this assumption is not met the
capture zones may differ slightly from those shown in the report.

In two villages, Brooklyn and Verona, the potential for error in the capture zone delineations is