norm ratios also show relatively small, but consistent, increases with slope. Both measures increase with
flow at both locations of the FT bin and in the C bin at 0.3L. There are no trends in either FT/C ratio for
0.3L, but these ratios show a general increase with flow at 0.7L. These results are generally consistent
with the measures obtained for group-averaged sediment load, but the trends are not as simple and clean.
The importance of FT induced erosion and its dependence on soil moisture is a consistent theme.
Nine-point averages of maximum channel width and maximum channel depth were obtained and
are presented with flow condition in Figures 17 and 18, respectively, by slope and soil moisture series.
Average maximum channel width (Fig. 17) generally increases with flow and slope in both bins. The
average maximum widths in each bin were generally highest at mid soil moisture, and comparable at both
higher and lower soil moisture. Corresponding FT/C ratios generally increase with soil moisture and do
not display trends with flow or slope. Average maximum channel depth (Fig. 18) generally decreases
with increasing soil moisture, increases with slope, and has some dependence on flow. Corresponding
FT/C ratios again generally increase with soil moisture, but do not depend on flow or slope. Group-
averaged maximum rill widths and depths in Table 9 were obtained for soil moisture, slope, and flow, as
well as overall. Grouped maximum rill widths increased with both slope and flow. However, the FT/C
ratios indicate that only soil moisture caused a width change as a result of FT, and then only at high soil
moisture. Trends in grouped maximum rill depth are very similar to those of sediment load. Again, the
important effect of soil moisture change on FT effects is clearly indicated by the FT/C depth ratio, while
increasing slope and flow cause depth increases that are unrelated to FT. Overall, depth differences
between FT and C bins are greater than those of width, but trends in these parameters and their ratios are
similar. Trends in these parameters are also similar to those of the L2 and Linf cross-sectional measures.
Replication of Experiments
Part way through our experimental program it became clear that the outcome of a particular
experiment was greatly affected by the parameters that describe the condition of the soil bins. To obtain
results that could be quantitatively compared and interpreted required minimum differences between all
bins of each soil moisture series, including weight, bulk density, and soil moisture. As evidenced by
nonzero standard deviations in Table 1, these parameters could not be replicated exactly, especially the
soil moisture. In this section we compare three experiments that were approximately replicated to better
understand and quantify the relationship between variability in parameters and corresponding variability
in results. Parameter differences considered here are generally larger than the corresponding standard
deviations of our experimental series. Low moisture experiment 6 and high moisture experiment 4 listed
in Table 10 were included in the analysis above. The replicate of low moisture experiment 6 had
significantly lower soil weight, lower soil moisture, and higher applied flow. High moisture experiment 4
and its replicate were more similar, with a large difference only in soil moisture. Corroborating the
difference in FT soil moisture was an average groundwater level difference of 12 cm to 3 cm between
these replicates. Groundwater levels in both C bins were 12 cm. This same experiment was repeated
twice more at lower soil weight and volume, at about the same bulk density as the original, and was called
high moisture experiment 4light. These two experiments have the closest parameters of the group,
differing only in soil moisture by slightly more than the standard deviation of the high moisture series
10
Previous Page
