Soil Water Variation Due to Grass Water Uptake

Shallow landslide is the phenomenon of slope collapse frequently occurring in tropical rainforest regions such as Malaysia based on geotechnical problem. This paper aims to determine the effects of grass on the shallow slope’s stability in terms of the variability of soil moisture, and to establish 1D suction of soil moisture due to grass water uptake. Moisture variations were used to compare the moisture content with time using GID and Fotran code tools. Based on the literature review findings, numerical simulation modelling was applied to achieve the most suitable condition for replicating grass water uptake within the soil slope. Six types of grasses were used in this research. The correlation between numerical simulation results is appropriate for these six types of grasses, but only Indropogon Gayanus grass with field monitoring results was achieved for validation. This paper assessed slope stability due to the influence of six types of grasses induced suction. The long grassroots are special between both kinds of grass, ranging from 0.3048 m to 4.000 m. This work gives a clear belief that Axonopus Compressus grass extracts water much faster than other grasses and is useful in preserving unsaturated soil stability.


Introduction
Slope can be categorized as natural slope and man-made slope. Natural slope consists of the slope of residual soil and surface. The man-made slope, however, is usually cut-back slope and filled with well-compacted filling material. Surface protections are carried out on man-made slopes, including vegetation and soil concreting, with adequate surface drainage such as water channels, catch pits, and sand traps. This is to protect the slope surface from erosion caused by rainfall and to ensure that anticipated soil property capacities speak to soil conduct sensibly well [1].
Depending on the design, subsurface stabilization including soil nailing and ground anchors with underground drainage such as weep hole is required to keep the slope stable in position. Commonly, vegetation has two key roles in stabilizing slope stability; one is through mechanical reinforcement, and the other is through hydrological mechanisms that cause increased evapotranspiration, thus increasing suction or extraction of moisture from the soil [2][3]. The stabilizing effectiveness depends on the vegetation type with specific root distribution, including dense sod cover, grasses, herbs, shrubs, and trees [2,4]. This research focuses on the effects of variations in soil moisture due to water uptake.

Root systems
Root systems for produces dense mats are the most crucial component of facilities considered by engineers as shown in Table 1.

Grass
Appreciable grassroots concentrations hit a depth of 0.3 m to 0.5 m. The grass growth habit's most significant aspect is the ground level which is the critical growth stage. This implies that no permanent damage to the plant is done by mild mowing, grazing, burning, or abrasion [2].

Data collection
The curve of soil water retention and hydraulic conductivity also act as two other soil hydraulic properties, which pose a significant concern in this research. Information on hydraulic conductivity properties was collected from different water retention curves and was merged into a single connection to reflect the entire soil profile. Genuchten obtained the water retention curve [11]. The result of the relationships is given as: Where r  and s  are residual and saturated water content, respectively,  is the capillary potential and  ,n and m are the empirical shape parameters. Table 2 presents the exact values of those parameters.
Where Ks is the saturated hydraulic conductivity, and where l is a parameter specific to soil.
The appropriate values for the parameters used in this case are also set out in Table 2. The water retention and hydraulic conductivity relationships are shown in Figure 1 respectively.

Numerical simulation
The simplest type of model for water uptake assumes a linear variation with depth in water extraction. It is assumed that the possible conditions for transpiration, Smax is given by, Where Smax is the extraction rate, aj and -bj are the intercept and slope on the jth day, respectively and z is the rooting depth, whereby zrj is the maximum depth of the root zone. The boundary condition at the bottom of the root zone (z = zrj) is Smax = 0, therefore: The total transpiration, Tj, across the root zone is then obtained through active depth integration which is: Combining equation (3) and (5) gives: Combining Equation (6) yields: The root zone equation at the bottom (4) gives: The provision of replacing equation (8) into equation (7) produces: This can be re-arranged as: Equation (12) is only valid under optimal levels of soil moisture. If the moisture content is low, then the real transpiration is less than the potential value. A model describes the sink term for accurate transpiration as follows [9]: is a defined function of the capillary potential referred to as a waterstress function. For more conditions drier than wilting point and wetter than a certain anaerobiosis point, it may be assumed that the water uptake by the roots is zero.
Feddes [15] makes two assumptions: the first is that no water uptake by the roots occurs at the Anaerobiosis spot with a head pressure of -50 cm, and second is that linear water uptake variability is assumed when the pressure head is less than-50 cm, respectively. The analysis requires moisture and impact pressure head to yield quality of various vegetable crops. Feddes [15] concludes that the water pressure head in the ground for these plants usually begins to restrict plant growth by around -400 cm. Consequently, the intake of water by roots is constant and does not exceed -400 cm (h2) < < -50 cm (h1). It is expected that the water uptake for the plant will decrease linearly between h3 = -400 cm and h4 = -15000 cm. Thus, when soil moisture is limited, equation (14) becomes: This simulation uses the size of the time step at 86400 seconds, 345600 seconds, 1296000 seconds, 2592000 seconds, 5184000 seconds, and 7776000 seconds, all of which were constant for the entire period considered. Preliminary numerical tests were conducted to ensure that the solutions for the domain size and the size of the time step at 86400 seconds (day 1) were concentrated. The simulation was analysed and covered for a period of 90 days. Figure 2 shows the number of variations in soil moisture content with time for grass carpets formally named Axonopus Compressus. There was a comparison of the moisture contents for numerical simulations in four measures over a 15, 30, 60, and 90-day simulation period.     Figure 6 shows the moisture content of the grass with the scientific name of Melinis Minutiflora indicating the moisture content outcomes depending on the time. Results for the soil shows that the moisture content increased gradually from the depths of 0.2 m to 5 m, which is by 10.5 percent. For day 30, simulations increased steadily to 32.6 percent at depths of 0.2 m to 0.5 m. As shown in Figure  6, on day 60 the moisture content indicated a difference between 0.   Figure 7 indicates the moisture content of the Digit grass, scientifically named Digitaria Eriantha. The moisture content for 15 days appeared to increase slowly to 4.80 percent at a depth of 0.2 m to 5 m. In the meantime, day 30 indicated that the moisture content increased to 7.60% for 30 days and to 11% for up to 60 days, while it also increased gradually as the duration multiplied by 12.47% for day 90 at a depth of 0.2 m to 5 m. For a period of 90 days, the moisture content reduced to 17.75 percent at the same depth of 0.2 m that is proportional to the soil. From day 1 to day 90 at a depth of 0.2 m to 5 m, the rate was 12.03 percent as the moisture content decreased. At almost the same depth from day 1 to day 90 of 0.2 m to 5 m, the moisture content rate showed a slight or steady decrease with a rate of 1.06%.  Figure 8 indicates the results of computer simulation and field measurement [16]. The result at depth 0.4 m is differences of 0.00620 cm 3 / cm 3 or 20.06 percent. The difference increases to 21.21% at depth 0.6 m, and that is 0.00681. At a depth of 0.6 m, the gap rose to 21.21 percent, and that is 0.006811. The difference between the two approaches is 22.01% at the next depth, which is 0.8 m. The difference is 0.00743 or 22.86 percent at a depth of 1 m and gap at a depth of 1.2 m is 23.70 percent. The difference is 25.56 percent or 0.00808 at the end of validation at a depth of 1.4 m. Therefore, the comparison of field measurement with the outcome of numerical simulation is appropriate. Figure 9 presents the graph for six types of grasses with a current period range of their moisture content at a depth of 5 m. The grasses used for this research are Axonopus Compressus, Pennisetum Purpureum, Andropogon Gayanus, Brachiaria Humidicola, Melinis Minutiflora, and Digitaria Eriantha. It is apparent from the graphs that the grassroots have short-term rate of soil moisture content and faster drying rate, and the higher rates are seen in comparison to the grass-root length. For Axonopus Compressus of 0.3048 m, the drying rate is 45.63% for the six grass types for 90 days and 40.54% for Andropogon Gayanus with a root length of 1 m in the same period for the six grass types. Minutiflora Melinis grass also shows the highest rate among the three-water absorption rate, 28.29 percent for the soil with root length of 1.2 m. By contrast, with the three other types of grasses with the root lengths in the range of 2 m to 4 m, the water absorption rates and moisture content on the soil occurred very slowly within 90 days. As for Pennisetum Purpureum and Brachiaria Humidicola, the drying rate is 11.03 percent and 15.89 percent, respectively, and this made them be known to be low rates and the length of their grass-roots is 4 m, each. Here, it can be inferred that shorter roots have a faster absorption rate compared to longer roots, as well as water intake. The difference can be seen at a depth of 5 m where the rate of water absorption by the grass with the shortest roots was faster than that with longer roots like Axonopus Compressus. Also, long-rooted grass consumes water at prolonged rate, for example, Digitaria Eriantha.

Conclusions
The result of this research as shown in Table 1 from literature review indicated that the suction rate for the short duration of the grassroots is strong at the top. Therefore, it can be inferred from the table that the moisture content depends on the length of the grassroots. The observation above proves that the root length does not guarantee the roots' durability in absorbing water from the soil, especially for shallow slopes. If the diffusion rate is high in water, the soil will become too moist, and the grass will not be suitable for shallow slopes as it may collapse. Axonopus Compressus grass is also used to prevent soil erosion in shallow slopes. While planting this grass is not essential, however, it is the main factor that has led to landslides' problems as planting grasses with long grassroots allow the roots to drain the water from the soil gradually. Axonopus Compressus grass is seen from the graph showing the depth of the grass by day as contributing to a faster rate of dry soil. It is compatible for use in the occurrence of prolonged rainfall, as the grass roots will help to absorb the water quickly and reduce the pressure of pore water while increasing soil shear strength [17]. Although the roots are essentially long, it is suggested from the computational simulations that the soil moisture content takes time to dry up within 90 days and is therefore not ideal for shallow slopes, when compared to other grasses, especially Digitaria Erianta grass. Generally, stress on the slope may be altered by increasing shear stress,or decreasing the significant normal stress on movement by reducing the shear strength of the material within the landslide [17]. Also, moisture content will impact the strength of the soils and weathered materials. The humid tropics deep residual soils can often have relatively high hydraulic conductivities that allow rainfall to infiltrate. Hydraulic conductivity will vary for the research situation for unsaturated soils as a function of the moisture content compared to saturated soil as it depends on the different hydraulic conductivity of their structure and composition [18]. Lastly, an environmental change may have a potential impact on slope stability [19].