172
11 Detecting Fingerprints of Gravity Erosion Drivers …
11.2 Methods and Materials
The gully bank collapse experiments were conducted under closely controlled conditions in the Joint Laboratory for Soil Erosion of Dalian University of Technology
and Tsinghua University located in Beijing, China. The landscape simulator consisted of a rainfall simulator and a conceptual landform covering an area of 3.0 m
by 3.0 m (Fig. 10.2). The conceptual landforms with a gentle upper slope of 3° were
made with loess by hand patting. More details of the gully bank collapse experiments
(slope height-gradient, rainfall intensity-duration) are provided in Table 11.1. Five
runs of rainfall were applied, in turns, on a landform, and the rainfall interval was
approximately 12 h. Furthermore, the physical properties of the model soil—the 50%
diameter of soil particles (d 50 ) and the specific gravity (γ s ), were 0.05 mm and 2.56,
respectively.
The form of a failure scar (i.e., the geometry of slip surfaces and the shape of
the vertical profiles on the slope surface) is created as a result of the soil block
separating from a sloped face. Herein, we used the type of intersection line to define
the morphology of failure scar in the experiments. The line which ran through the
intersection of the failure scar and the longitudinal section which passed through the
center of the scar was obtained.
During, and 20 min after, the rainfall, slope failure occurrence time, slip mode,
type of failure scar, location, and slope failure retrogression behavior were recorded
by direct observation and via the topography meter designed by us (Guo et al. 2016).
To ensure the accuracy of the scar pattern obtained from direct observation during
the experiments, the videos in the topography meter were used to check the pattern.
For each mass failure, we calculated the volume and classified the scar form.
Then, we obtained the total amount of soil loss (g n ) and the sum amount of mass
failures (g v ) corresponding to all types of scars. All failure masses with volume
greater than 500 cm
3 were considered in the study. To evaluate the independent
effects of slope gradient and height, and rainfall intensity and duration on the mass
volume of each scar pattern, for this study the experiments were divided into the
following seven experimental groups. Each experimental group had the same slope
height or gradient, and rainfall intensity or duration:
(1) G a (experiments L1, 2, 5 and 6) versus G b (experiments L3, 4, 7 and 8). Slope
height of the initial lower slope in the former experimental group was 1.0 m,
while the latter was 1.5 m.
(2) G c (experiments L1, 3, 5 and 7) versus G d (experiments L2, 4, 6 and 8). Slope
gradient of the initial lower slope in the former experimental group was 70°,
while the latter was 80°.
(3) G e (experiments L5–6) versus G f (experiments L9–10). Rainfall intensity in the
former experimental group was 0.8 mm/min, while the latter was 2.0 mm/min.
(4) G f (experiments L9–10) versus G h (experiments L1–2). Rainfall duration in the
former experimental group was 30 min, while the latter was 60 min.
Then we used the increase-rate-analysis method to assess variations in the gravity
erosion in regard to changes in other causal factors such as rainfall intensity and
11 Detecting Fingerprints of Gravity Erosion Drivers …
11.2 Methods and Materials
The gully bank collapse experiments were conducted under closely controlled conditions in the Joint Laboratory for Soil Erosion of Dalian University of Technology
and Tsinghua University located in Beijing, China. The landscape simulator consisted of a rainfall simulator and a conceptual landform covering an area of 3.0 m
by 3.0 m (Fig. 10.2). The conceptual landforms with a gentle upper slope of 3° were
made with loess by hand patting. More details of the gully bank collapse experiments
(slope height-gradient, rainfall intensity-duration) are provided in Table 11.1. Five
runs of rainfall were applied, in turns, on a landform, and the rainfall interval was
approximately 12 h. Furthermore, the physical properties of the model soil—the 50%
diameter of soil particles (d 50 ) and the specific gravity (γ s ), were 0.05 mm and 2.56,
respectively.
The form of a failure scar (i.e., the geometry of slip surfaces and the shape of
the vertical profiles on the slope surface) is created as a result of the soil block
separating from a sloped face. Herein, we used the type of intersection line to define
the morphology of failure scar in the experiments. The line which ran through the
intersection of the failure scar and the longitudinal section which passed through the
center of the scar was obtained.
During, and 20 min after, the rainfall, slope failure occurrence time, slip mode,
type of failure scar, location, and slope failure retrogression behavior were recorded
by direct observation and via the topography meter designed by us (Guo et al. 2016).
To ensure the accuracy of the scar pattern obtained from direct observation during
the experiments, the videos in the topography meter were used to check the pattern.
For each mass failure, we calculated the volume and classified the scar form.
Then, we obtained the total amount of soil loss (g n ) and the sum amount of mass
failures (g v ) corresponding to all types of scars. All failure masses with volume
greater than 500 cm
3 were considered in the study. To evaluate the independent
effects of slope gradient and height, and rainfall intensity and duration on the mass
volume of each scar pattern, for this study the experiments were divided into the
following seven experimental groups. Each experimental group had the same slope
height or gradient, and rainfall intensity or duration:
(1) G a (experiments L1, 2, 5 and 6) versus G b (experiments L3, 4, 7 and 8). Slope
height of the initial lower slope in the former experimental group was 1.0 m,
while the latter was 1.5 m.
(2) G c (experiments L1, 3, 5 and 7) versus G d (experiments L2, 4, 6 and 8). Slope
gradient of the initial lower slope in the former experimental group was 70°,
while the latter was 80°.
(3) G e (experiments L5–6) versus G f (experiments L9–10). Rainfall intensity in the
former experimental group was 0.8 mm/min, while the latter was 2.0 mm/min.
(4) G f (experiments L9–10) versus G h (experiments L1–2). Rainfall duration in the
former experimental group was 30 min, while the latter was 60 min.
Then we used the increase-rate-analysis method to assess variations in the gravity
erosion in regard to changes in other causal factors such as rainfall intensity and