Nitrate (NO3-) is one of the major nonpoint source pollutants in the vadose
zone. In this study, NO3- distributions were estimated in an 800 by 800 by
1.8 m soil volume using kriging and cokriging with nonsymmetric pseudo-cro
ssvariograms. Cokriging with pseudo-crossvariograms maximized the use of av
ailable information at different soil depths. Cokriging allowed easily obta
ined information at shallow layers to be used to improve soil chemical esti
mations at deeper layers, Compared with kriging, cokriging significantly re
duced the mean squared errors (MSEs) and mean kriging variances (MKVs) of t
he NO3- estimations in the soil. For the same estimation accuracy of krigin
g, cokriging with pseudo-crossvariograms used less than half the data; thus
potentially it could reduce more than half the sampling cost than kriging
estimation, Cokriging with pseudo-crossvariograms was shown to be a precise
and an economic way for determining nonpoint source pollutant distribution
s in large fields.