2.4 Experimental Methods and Materials
19
2.4.3 Performance of the Landscape Simulator and Data
Collection
Rainfall intensity was calibrated in advance of each rainfall event. The plot surface
was covered with a tarpaulin. Rainfall was collected in a calibrated tank for the
Model B or a small calibrated bucket for the Model Da and Model Db, and was
volumetrically measured at 10-s intervals.
Soil water storage was very sensitive to rainfall (Peugeot et al. 1997). Thus,
the significant effort was made to ensure equal initial soil moisture content before
each experiment. To measure the initial soil moisture content, six soil samples were
collected from the upper, middle and lower reaches of the model watershed surface, respectively, before each rainfall event. For Model B, an equal period of 24 h
was maintained after each rainfall event. Before each rainfall-erosion event, a lowintensity rainfall was applied to the landform surface for a few minutes. For the
Model Da and Model Db, before each rainfall-erosion event, atomized water drops
from a portable nebulizer were applied to the land surface until the model landform
was drenched but no runoff emerged.
The amount of soil loss for a downscaled watershed is typically estimated based
on the measured amount of water discharged at outlets (Muttiah et al. 2005). During
each rainfall event, the water-solid mixture was collected in a calibrated tank at the
outlet of each plot. At 2 min or 1 min intervals, runoff samples were collected in
100 ml sampling bottles to determine the sediment concentration. Flow rates were
synchronously measured using the label on the wall of the calibrated tank/bucket.
After a rainfall event, the bed load deposited at the bottom of the tank was dried
and weighed. The weight of the suspended load was calculated by the concentration
determined gravimetrically, and the volume was then measured using the bucket
gauge.
2.4.4 Executive Process
To validate the scale number of soil loss. Several groups of experiments were conducted to determine the model rain erosivity. Except that the initial landscape for the
first rainfall event in a group of experiments was made by hand patting, the other
subsequent landscapes were formed by the previous rainfall event. (1) Two groups of
experiments were conducted for the Model B. First, 7 rainfall events, generated by
the SX2004 Sprayer-styled Rainfall Simulator with intensities of about 1.60 mm/min
and a duration of 20 min, were applied to the initial ground cover of the Model B
without check dams. Then recovered the landform to the initial state, and gave 10
rainfalls same to the experiment above mentioned while 12 check dams were orderly
constructed as shown in Table 2.3. (2) Two groups of experiments were conducted
for the Model Da. First, 6 rainfall events, generated by the SX2002 Spout-type Rainfall Simulator, had intensities of about 1.65 mm/min and a duration of 10 min, were
19
2.4.3 Performance of the Landscape Simulator and Data
Collection
Rainfall intensity was calibrated in advance of each rainfall event. The plot surface
was covered with a tarpaulin. Rainfall was collected in a calibrated tank for the
Model B or a small calibrated bucket for the Model Da and Model Db, and was
volumetrically measured at 10-s intervals.
Soil water storage was very sensitive to rainfall (Peugeot et al. 1997). Thus,
the significant effort was made to ensure equal initial soil moisture content before
each experiment. To measure the initial soil moisture content, six soil samples were
collected from the upper, middle and lower reaches of the model watershed surface, respectively, before each rainfall event. For Model B, an equal period of 24 h
was maintained after each rainfall event. Before each rainfall-erosion event, a lowintensity rainfall was applied to the landform surface for a few minutes. For the
Model Da and Model Db, before each rainfall-erosion event, atomized water drops
from a portable nebulizer were applied to the land surface until the model landform
was drenched but no runoff emerged.
The amount of soil loss for a downscaled watershed is typically estimated based
on the measured amount of water discharged at outlets (Muttiah et al. 2005). During
each rainfall event, the water-solid mixture was collected in a calibrated tank at the
outlet of each plot. At 2 min or 1 min intervals, runoff samples were collected in
100 ml sampling bottles to determine the sediment concentration. Flow rates were
synchronously measured using the label on the wall of the calibrated tank/bucket.
After a rainfall event, the bed load deposited at the bottom of the tank was dried
and weighed. The weight of the suspended load was calculated by the concentration
determined gravimetrically, and the volume was then measured using the bucket
gauge.
2.4.4 Executive Process
To validate the scale number of soil loss. Several groups of experiments were conducted to determine the model rain erosivity. Except that the initial landscape for the
first rainfall event in a group of experiments was made by hand patting, the other
subsequent landscapes were formed by the previous rainfall event. (1) Two groups of
experiments were conducted for the Model B. First, 7 rainfall events, generated by
the SX2004 Sprayer-styled Rainfall Simulator with intensities of about 1.60 mm/min
and a duration of 20 min, were applied to the initial ground cover of the Model B
without check dams. Then recovered the landform to the initial state, and gave 10
rainfalls same to the experiment above mentioned while 12 check dams were orderly
constructed as shown in Table 2.3. (2) Two groups of experiments were conducted
for the Model Da. First, 6 rainfall events, generated by the SX2002 Spout-type Rainfall Simulator, had intensities of about 1.65 mm/min and a duration of 10 min, were