14
2 Similarity of Model Experiments
Region is characterized by the crisscrossing gullies covered with the thick loess on
top of ancient landforms. Ridges and mounds are the most typical micro-relief form
on the Loess Plateau.
The prototype watershed, the Yangdaogou Catchment, is a typical small watershed
located at the Loess Hill Ravine Region, and covers an area of 0.206 km
2 (Xu et al.
2006). The ground surface in the watershed is covered with Lishi loess and Malan
loess (Cai et al. 1998). The major landform parameters, such as the drainage area,
watershed elevation, hillside gradient, and length of the main gully, are all scaled
down according to the dimensions of the Yangdaogou Catchment with the same
horizontal and vertical proportions. Precipitation in the Yangdaogou Catchment is
typical of that on the Loess Plateau. In the 1960s, when no dam existed, the average
annual runoff rate was 36,700 m
3 /(km
2 a), and the average annual soil erosion rate
was 20,811 t/(km
2 a) (Zhang et al. 1995). Therefore, the annual soil loss, namely,
the soil loss of the conceptual prototype in a rainfall event, S p , could be obtained by
multiplying the watershed area with the annual soil erosion rate, which is 4.29 × 10
6
kg/a. Generally, the annual erosion rate is adopted as an index of the structural practice
while the check dam system is designed, and a rainfall event in the conceptualized
prototype watershed accounts for the amount of soil loss in one year in the prototype
watershed.
2.4.2 Landscape Simulator
The landscape simulator for the Model B consisted of a rainfall simulator suspended
above a flume containing the large scale watershed model. The other two landscape
simulators for the small scale models, the Model Da and Model Db, of which relative
ratio R DB was not equal to 1 and was close to 1, respectively, were used to verify
the semi-scale model. The experimental apparatus and techniques in this study have
been used extensively to examine the behavior of the watershed at a model scale, and
could be used to generate the model watersheds with many features that are similar
to field watersheds (Xu et al. 2006). The determinant parameters of the three models
are listed in Table 2.1.
Table 2.1 shows the determinant parameters of the prototype watershed. The table
also lists the erosion data for the prototype watershed together with those for the large
scale model and the small scale model, and similarity conditions.
Figure 2.2 presents a schematic representation of the experimental setup for the
Model B. Two lines of five SX2004 Sprayer-styled Rainfall Simulators were utilized
to simulate the rainfall in the experimental plot, which measured 6.0 m × 10.8 m.
The simulator contained a nozzle with an inner rotor, which was screwed onto the top
of the pipes and sprayed downwards to the landform located 5.5 m below. Figure 2.3
presents a schematic representation of the experimental setup for the Model Db. A
line of three SX2004 Sprayer-styled Rainfall Simulators was used to simulate the
rainfall in the experimental plot, which measured 1.5 m × 2.7 m. Figure 2.4 presents
a schematic representation of the experimental setup for the Model Da. A SX2002
2 Similarity of Model Experiments
Region is characterized by the crisscrossing gullies covered with the thick loess on
top of ancient landforms. Ridges and mounds are the most typical micro-relief form
on the Loess Plateau.
The prototype watershed, the Yangdaogou Catchment, is a typical small watershed
located at the Loess Hill Ravine Region, and covers an area of 0.206 km
2 (Xu et al.
2006). The ground surface in the watershed is covered with Lishi loess and Malan
loess (Cai et al. 1998). The major landform parameters, such as the drainage area,
watershed elevation, hillside gradient, and length of the main gully, are all scaled
down according to the dimensions of the Yangdaogou Catchment with the same
horizontal and vertical proportions. Precipitation in the Yangdaogou Catchment is
typical of that on the Loess Plateau. In the 1960s, when no dam existed, the average
annual runoff rate was 36,700 m
3 /(km
2 a), and the average annual soil erosion rate
was 20,811 t/(km
2 a) (Zhang et al. 1995). Therefore, the annual soil loss, namely,
the soil loss of the conceptual prototype in a rainfall event, S p , could be obtained by
multiplying the watershed area with the annual soil erosion rate, which is 4.29 × 10
6
kg/a. Generally, the annual erosion rate is adopted as an index of the structural practice
while the check dam system is designed, and a rainfall event in the conceptualized
prototype watershed accounts for the amount of soil loss in one year in the prototype
watershed.
2.4.2 Landscape Simulator
The landscape simulator for the Model B consisted of a rainfall simulator suspended
above a flume containing the large scale watershed model. The other two landscape
simulators for the small scale models, the Model Da and Model Db, of which relative
ratio R DB was not equal to 1 and was close to 1, respectively, were used to verify
the semi-scale model. The experimental apparatus and techniques in this study have
been used extensively to examine the behavior of the watershed at a model scale, and
could be used to generate the model watersheds with many features that are similar
to field watersheds (Xu et al. 2006). The determinant parameters of the three models
are listed in Table 2.1.
Table 2.1 shows the determinant parameters of the prototype watershed. The table
also lists the erosion data for the prototype watershed together with those for the large
scale model and the small scale model, and similarity conditions.
Figure 2.2 presents a schematic representation of the experimental setup for the
Model B. Two lines of five SX2004 Sprayer-styled Rainfall Simulators were utilized
to simulate the rainfall in the experimental plot, which measured 6.0 m × 10.8 m.
The simulator contained a nozzle with an inner rotor, which was screwed onto the top
of the pipes and sprayed downwards to the landform located 5.5 m below. Figure 2.3
presents a schematic representation of the experimental setup for the Model Db. A
line of three SX2004 Sprayer-styled Rainfall Simulators was used to simulate the
rainfall in the experimental plot, which measured 1.5 m × 2.7 m. Figure 2.4 presents
a schematic representation of the experimental setup for the Model Da. A SX2002