9.1 Impact Factors of the Gravity Erosion
143
The mode of the rain-induced mass failure strongly depends on the initial state of
the slope materials, together with the pore water pressure distribution and magnitude
of apparent cohesion due to variations in the soil water content (Zhang et al. 2014;
Lourenço et al. 2006). Failures may differ in respect of thickness of failed mass,
time of failure occurrence, rotational inclination (Au 1998), or be also classified in
part by distribution on the ground, in the duration of time that the process acts, by
the rate of movement (Wang et al. 2014; Shroder and Bishop 1998). The amount of
gravity erosion is pivotal but not readily observable (Xu and Zhao 2014). Site-specific
observation is almost impossible due to the uncertainty and non-continuity of gravity
erosion (Keefer and Larsen 2007; Benda and Dunne 1997). The volume of individual
failure was normally calculated by multiplying the slide area by the thickness of the
slide mass (Guzzetti et al. 2009; Haflidason et al. 2005). Nevertheless, the calculated
volume involves an amount of tinkering, since the scars caused by the shallow debris
flow rapidly heal and are difficult to be detected after a few years (Montgomery and
Dietrich 1994). Moreover, erosion volumes caused by water and gravity could not
be distinguished with the above approaches.
9.2 Method and Materials
The Loess Plateau is located in the upper and middle reaches of the Yellow River,
covering a total area of 624,000 km
2 (Fig. 9.1). The plateau is one of the severest area
in the world suffered from gravity erosion (Zhang et al. 2004), and 30–50% of the
land is subjected to gravity erosion (Wang et al. 1993). Gravity erosion frequently
occurs induced by the rainstorm, because the undulating terrain on the Loess Plateau
is characterized by crisscrossing gullies, the vegetation is so sparse, and especially
the loess is collapsible due to vertical joints. On the Loess Plateau, a steep bank with
the slope of more than 70° in the upper reach of the small watershed is the main
source of gravity erosion. Forms of gravity erosion include avalanche, landslide,
earth flow, creep and so on (Tang 2004).
Most of the gravity erosions occurred during or after the storms (Ali et al. 2014;
Fourie 1996; Montgomery and Dietrich 1994). The specific processes are most easily
studied and quantified in a flume using a rainfall simulator under controlled laboratory conditions (Acharya et al. 2011). Here we define and classify the observed
mechanisms of rainfall-induced mass failure that occurred on the steep loess slope in
the laboratory tests, and then present the latest method to assess the gravity erosion.
A topography meter designed by us was used to quantitatively measure the process
of gravity erosion.
To classify different failure mechanisms and observe the instability conditions,
we conducted a series of gully bank collapse experiments under closely controlled
conditions in the Joint Laboratory for Soil Erosion of Dalian University of Technology and Tsinghua University in 2010 and 2012. The landscape simulator was
consisted of a rainfall simulator and a slope model covering an area of 3.0 m by
4.0 m (Fig. 9.1). 5 runs of rainfalls each at an amount of 48 mm, were applied in
143
The mode of the rain-induced mass failure strongly depends on the initial state of
the slope materials, together with the pore water pressure distribution and magnitude
of apparent cohesion due to variations in the soil water content (Zhang et al. 2014;
Lourenço et al. 2006). Failures may differ in respect of thickness of failed mass,
time of failure occurrence, rotational inclination (Au 1998), or be also classified in
part by distribution on the ground, in the duration of time that the process acts, by
the rate of movement (Wang et al. 2014; Shroder and Bishop 1998). The amount of
gravity erosion is pivotal but not readily observable (Xu and Zhao 2014). Site-specific
observation is almost impossible due to the uncertainty and non-continuity of gravity
erosion (Keefer and Larsen 2007; Benda and Dunne 1997). The volume of individual
failure was normally calculated by multiplying the slide area by the thickness of the
slide mass (Guzzetti et al. 2009; Haflidason et al. 2005). Nevertheless, the calculated
volume involves an amount of tinkering, since the scars caused by the shallow debris
flow rapidly heal and are difficult to be detected after a few years (Montgomery and
Dietrich 1994). Moreover, erosion volumes caused by water and gravity could not
be distinguished with the above approaches.
9.2 Method and Materials
The Loess Plateau is located in the upper and middle reaches of the Yellow River,
covering a total area of 624,000 km
2 (Fig. 9.1). The plateau is one of the severest area
in the world suffered from gravity erosion (Zhang et al. 2004), and 30–50% of the
land is subjected to gravity erosion (Wang et al. 1993). Gravity erosion frequently
occurs induced by the rainstorm, because the undulating terrain on the Loess Plateau
is characterized by crisscrossing gullies, the vegetation is so sparse, and especially
the loess is collapsible due to vertical joints. On the Loess Plateau, a steep bank with
the slope of more than 70° in the upper reach of the small watershed is the main
source of gravity erosion. Forms of gravity erosion include avalanche, landslide,
earth flow, creep and so on (Tang 2004).
Most of the gravity erosions occurred during or after the storms (Ali et al. 2014;
Fourie 1996; Montgomery and Dietrich 1994). The specific processes are most easily
studied and quantified in a flume using a rainfall simulator under controlled laboratory conditions (Acharya et al. 2011). Here we define and classify the observed
mechanisms of rainfall-induced mass failure that occurred on the steep loess slope in
the laboratory tests, and then present the latest method to assess the gravity erosion.
A topography meter designed by us was used to quantitatively measure the process
of gravity erosion.
To classify different failure mechanisms and observe the instability conditions,
we conducted a series of gully bank collapse experiments under closely controlled
conditions in the Joint Laboratory for Soil Erosion of Dalian University of Technology and Tsinghua University in 2010 and 2012. The landscape simulator was
consisted of a rainfall simulator and a slope model covering an area of 3.0 m by
4.0 m (Fig. 9.1). 5 runs of rainfalls each at an amount of 48 mm, were applied in