The importance of incorporating nonuniform flow effects in determining
field-scale values of longitudinal dispersivity from tracer tests in
aquifers is illustrated. We show that under conditions of Fickian disp
ersive flux, where a tracer has traveled at least 10 times the longitu
dinal dispersivity value, nonuniform flow effects account for the typi
cal tailing exhibited by data from radial flow and doublet (two-well)
tracer tests. Approximate analytical solutions for solute transport in
nonuniform flow fields based on earlier general work are presented. S
pecific cases evaluated include pulse and step inputs in diverging rad
ial flow, a pulse input in converging radial flow, and doublet (two-we
ll) tests with pulse and step inputs. The additional effect of longitu
dinal dispersivity increasing linearly with distance for radial flow c
ases is examined. Also presented is an analysis of the effect of injec
tion borehole flushing represented by an exponentially decreasing trac
er input in the converging radial flow case. The solutions are applied
to four sets of field data where previous analyses did not adequately
address nonuniform flow effects. Some data reanalyses yield longitudi
nal dispersivity values which are up to a factor of three smaller than
those previously reported, indicating a previous overprediction of pl
ume dilution. We demonstrate that an understanding of the sensitivity
of the mathematical solutions to tracer input conditions, nonuniform f
low effects, and scale can lead to better tracer-test design.