A spatial stochastic approach was adopted to estimate the capture zones of
a wellfield in eastern Netherlands near the town of Wierden. The main sourc
es of uncertainty were found to be the hydraulic conductivity of the produc
tion aquifer and the spatial distribution and thickness of an impeding clay
layer situated above the production aquifer, Both were modeled as random s
pace functions and a Monte Carlo analysis was performed to assess the proba
bility distributions of the capture zones for t = 25 years and t --> infini
ty. The distributions were compared with two deterministic capture zones to
show how the stochastic approach preserves the discontinuous character of
the impeding clay layer while the deterministic approach smooths out the di
scontinuities. Furthermore, it is shown how the stochastic approach reveals
the uncertainty about the location of the capture zones while the determin
istic approach obscures it.