Tr. Ellsworth et Cw. Boast, SPATIAL STRUCTURE OF SOLUTE TRANSPORT VARIABILITY IN AN UNSATURATED FIELD SOIL, Soil Science Society of America journal, 60(5), 1996, pp. 1355-1367
Three solutes were applied in series under steady, unsaturated flow to
determine the spatial scale dependence of transport. Twenty-seven day
s after solute application, a 2 by 2 by 2 m(3) field plot was excavate
d using two measurement scales, 22.2 by 22.2 by 10 cm(3) (scale L) and
7.4 by 7.4 by 10 cm(3) (scale I), providing two sets of 81 samples fo
r each 0.1-m layer to a depth of 2.0 m. For each chemical and for each
scale, vertical moment analysis was performed on the data for each of
the 81 vertical columns. Evidence of enhanced local horizontal mixing
with increasing mean travel time was observed in two ways: (i) a decr
ease in mass recovery variance at scale L for NO3- (center of mass at
80-cm depth) vs. Br- (46-cm depth of center of mass), and (ii) a horiz
ontal correlation range in solute concentrations in individual 0.1-m-t
hick layers, which was lowest between 70 and 130 cm deep. Point variog
rams of the vertical moments for each solute were estimated by simulta
neous deconvolution of the two observed (at scales l and L) area-avera
ged variograms, Dispersion variance analysis (based on these point var
iograms) characterized the variability within the plot of mass recover
y and center of mass depth as a function of sample size. Several appro
aches for characterizing plot- or field-scale transport features from
soil-core scale measurements were evaluated. The horizontal dependence
of the vertical variance increased with mean travel time. For each tr
acer, this dependence was reasonably characterized as the sum of the d
ispersion variance in the center of mass and the average L-scale verti
cal variance.