The experimental site, situated in the middle of Erlaohugou on the Ordos Plateau, Inner Mongolia Autonomous Region, China (39° 47′ 38.79″N, 110° 36′ 2.74″E; Fig. 1), is within the bare Pisha sandstone area. The area has an arid and semi-arid temperate continental monsoon climate with cold winter and hot summer. The mean annual precipitation is about 400 mm with 70%-80% falling in July to September. The area is mainly dominated by water and gravity erosion. The overall vegetation coverage is less than 20%.

Fig. 1

	Figure Option ViewDownloadNew Window
	Fig. 1 Location of the study site and types of Pisha sandstone sub-areas

W-OH is a type of hydrophilic polyurethane produced by Toho Chemical Industry Co. Ltd., Japan. It can rapidly react with water to form an elastic gel with strong adhesion capabilities. It does not contain any heavy metal ions and thus has no harmful effect on plants and animals (Wu et al., 2009).

To evaluate the erosion resistance of Pisha sandstone slopes treated with W-OH solution, we conducted simulated scouring tests on two slopes: an original hard Pisha sandstone slope (slope angle, 43° ; S1) without vegetation coverage and a quicksand slope (slope angle, 34° ; S2) with some vegetation. The experimental area was isolated by iron sheets from the surrounding environment. The iron sheets were 15 cm high with 5 cm above the ground. Each experimental slope was divided into the treated area and the control (not treated) area and the control area was separated by iron sheets as well. The experimental design is shown in Figure 2. The 6% and 3% of W-OH solutions were sprayed on the surface of treated areas of S1 and S2, respectively. The spraying amount was 3 L/m². The scouring test started after the surface was completely dry. The scouring flow was controlled at approximately 3 L/min and the scouring process lasted for 1 h. We repeated the scouring process for three times and selected a typical one for analyses. The sediment yield was collected by barrel at the bottom of the slopes. We collected the sediment yield every other minute over the entire scouring process and obtained 30 collections. Then we mixed all collections to form a representative sediment sample and dried it in an electric oven at 105° C to a constant weight. The morphological changes on the slopes were documented by taking pictures.

Fig. 2

	Figure Option ViewDownloadNew Window
	Fig. 2 Illustration of the field scouring test. The length of the scouring slope is 4.0 m and the width is 1.0 m.

A 5-month field demonstration was conducted on a Pisha sandstone slope from May to September 2014 to monitor the effects of W-OH solution. The slope was 43.5 m long with a vertical height of 29.9 m and an average gradient of 44.5° (Fig. 3a). We surrounded the experimental area with iron sheets and divided the area into A and B parts. Part A was set as the control and part B was treated with W-OH solutions. According to the slope gradient and potential for vegetation growth, we further divided B into consolidation-waterproof (B1), consolidation-growth promotion (B2), and consolidation-green (B3) sub-sections (Fig. 3b). The three sub-sections were treated with W-OH solutions in different concentrations. The slope angles of sub-sections B1, B2, and B3 were greater than 60° , 30° -60° , and less than 30° , and the concentrations of the solutions sprayed were 8%-10%, 4%-5%, and 3%, respectively. W-OH solutions were artificially sprayed (Fig. 3a) and the sprayed amount was controlled at 2.0-3.0 L/m² to ensure the penetration depths reaching 5-8 mm. There were four barrels installed at the slope bottom to collect the eroded sediments and runoff. The seeds of local bushes and grasses, such as Buchloe dactyloides and Pseudostellaria maximowicziana (Zeng et al., 2013), were sowed in B2, B3 and the similar areas in A. The B was treated with W-OH solutions. Micro-irrigation systems were constructed to guarantee the water supply for the vegetation growth. A meteorological station, and a water and sediment monitoring station were set up to collect the rainfall intensitiy, temperature, humidity, wind, sediment, etc.

Fig. 3

	Figure Option ViewDownloadNew Window
	Fig. 3 Experimental area in Erlaohu Gou on the Ordos Plateau, Inner Mongolia Autonomous Region, China. (a) spraying W-OH solutions, (b) the experimental design and control partition.

The sediment yield and vegetation growth (Zhang et al., 2009) were recorded after each rainfall event. The vegetation coverage digital pictures, taken with a focus on a 50 cm× 50 cm area, were measured using Adobe Photoshop.

As shown in Figure 4a, the sediment yield from the control area of slope S1, an original hard Pisha sandstone slope with an angle of 43° without vegetation coverage, increased dramatically first and then gradually decreased with scouring time. The peak value of sediment yield, approximately 3200 g/min, appeared at 10 min after the process started. The formation of erosion gullies at the early stage of scouring led to the increase of sediment yield. For the area treated with the 6% W-OH solution, the sediment yield had some fluctuations in the scouring process but the value of the collection was less than 7.0 g/min.

Fig. 4

	Figure Option ViewDownloadNew Window
	Fig. 4 Sediment yield on two different Pisha sandstone slopes. (a), an original hard Pisha sandstone slope (S1; slope angle of 43° ) without vegetation coverage; (b), a quicksand slope (S2; slope angle of 34° ) with some vegetation.

The sediment yield process of the control area of S2, a quicksand slope with an angle of 34° with some vegetation, presented a different trend (Fig. 4b) in comparison with S1. Specifically, the sediment yield increased relatively gradually at the beginning of the scouring process and then became stable at about 20 min after the process started. For the treated area of S2 (i.e., treated by 3% W-OH in Fig. 4b), the sediment yield presented a similar trend with that in treated area of S1. The sediment yield was also less than 7.0 g/min in the entire process.

After each rainfall event, we collected the sediment in the barrels. We found that the sediment yield tended to increase with increasing rainfall intensity in both treated and control cases (Table 1). But, the difference between the treated case and the control case was rather large. For example, at the rainfall intensity of 41.2 mm/h, the sediment yield of the control case reached approximately 1895.7 kg with the average unit loss of 11.8 kg/m², while the sediment yield of the treated case was only 11.9 kg with the average unit loss of only 0.074 kg/m². It means that the W-OH treatment has reduced the erosion by 95%. These results indicate that the consolidation layer formed by W-OH has greatly enhanced the erosion resistance of Pisha sandstone slopes.

Table 1

Table 1

Table 1 Results of the experimental and control areas in rainy days from July to September 1 Jul 22 Jul 28 Jul 2 Aug 12 Aug 22 Aug Rainfall (mm) 16.4 8.6 23.2 35.6 20.6 34.4 30-min rainfall intensity (I30, mm/h) 26.0 4.4 15.6 41.2 26.8 20.0 Rainfall erosion force (MJ• mm/(hm2• h)) 94.8 9.4 77.2 373.3 131.0 133.3 Sediment yield in experimental area (kg) 8.6 0.0 1.8 11.9 0.4 1.6 Sediment yield in control area (kg) 216.1 42.0 27.3 1895.7 52.1 18.7 Reduced sediment yield (%) 96.0 100.0 93.4 99.4 99.2 91.4

Table 1 Results of the experimental and control areas in rainy days from July to September

Again as shown in Figure 3b, Part A was the control and Part B was divided into three sub-sections. The performance of the consolidation layer in protecting the slopes was dependent on the thickness of layer and the concentration of W-OH solution as well. For the steeper portion of the slope under B1 treatment, a few cracks appeared (Fig. 5a) even under the consolidation layer reaching the maximum penetration depth of 8 mm. Cracks occurred because the residence time of the W-OH solution was short. In addition, the steeper slope was not very smooth and was difficult to guarantee a uniform W-OH spraying. As a result, water-eroding and gravity collapsing forces led to minor damage to the consolidation layer in some places. Consequently, the number and the magnitude (i.e., depth and width) of the grooves increased gradually. Continuous expansion of the grooves with time and with rainfall intensity eventually resulted in the collapse of the entire slope.

Fig. 5

	Figure Option ViewDownloadNew Window
	Fig. 5 Comparison between the consolidation-waterproof (B1) and consolidation-green (B3). (a) cracks in the consolidation-waterproof section; (b) comparison between A and B (also among B1, B2, and B3) in September, 2014. (1)-(5) are cracks appeared in sections under different treatments.

In the consolidation-green area (i.e., B3) where the concentration of the sprayed solutions was the smallest (approximately 3%), vegetation grew quickly from May to August (Fig. 5b). When heavy rainfall occurred from July to September, the vegetation had already grown up and the mature vegetation played an important role in reducing the erosion force of raindrops and water flow. Therefore, just a few small erosion gullies appeared in the treated area.

In the case of B2 (i.e., consolidation-growth promotion), as the vegetation coverage increased, the plants became more effective in protecting the weathered Pisha sandstone from further erosion. In the treated area, the vegetation coverage reached approximately 95% and the average height of vegetation reached 50-60 cm with the average root length reaching 50 cm at the end of September (Fig. 6). However, in the control area, the vegetation coverage was only 26% and the average height of the vegetation was only 10 cm. The vegetation not only increased the roughness of the slope surface and effectively reduced the kinetic energy of raindrops and flow velocity of water, but also played a role in improving soil permeability (Zhao et al., 2014). In brief, W-OH treatment can protect the Pisha sandstone from erosion and provide the vegetation seeds a chance to grow. Once the vegetation matured, consolidation-growth mutual promotion effect took place.

Fig. 6

	Figure Option ViewDownloadNew Window
	Fig. 6 Vegetation coverage and height in the treated area (TA) and control area (CA) from May to September 2014