A Laboratory Watershed Model to Study the Effect of Rainfall Intensity and Soil Surface Slope on Surface Runoff Rate of Karbala Desert Soil

Abstract


Introduction
Water is one of the most important natural resources for countries and a mainstay for their growth and development, which necessitates searching for ways to preserve water wealth and optimally exploit all the water that the hydrological cycle gives.Water harvesting is an economical and effective means of benefiting from rainwater in arid areas and mitigating the negative effects of drought in these areas (Al-Hamoud, 2021).Water harvesting is an old practice of making use of runoff water resulting from precipitation, especially in arid regions that are characterized by scarcity of traditional water resources (Hammadi and Turki, 2023).Recent years have witnessed a "great" interest in the design, development, and management of water harvesting systems for human, agricultural, and even industrial consumption.Runoff from the land surface is the flow of water that results from extra water from rain, melt-water, or other sources that flow over the Earth's surface.It has a significant role in the regional and worldwide hydrological cycle, since it is a crucial water source for agriculture, industry, urban water consumption, and other uses, and it directly affects human life (Mahdi and Sudkhan, 2022).For effective catchment design, planning, and management, it is essential to comprehend the intricate relationships between rainfall and runoff processes and to accurately estimate surface runoff.Hydrological modeling can help with this since it estimates continuous surface runoff, models the effects of climate change and land use changes on surface water balance, and provides insight into catchment behavior (Li et al., 2012;Zhao et al., 2012).Good simulations and predictions of surface runoff require model calibration, which is an essential step, in order to adjust the model parameters for a catchment's inputs and water fluxes, hydrological models are typically calibrated against observed stream flow (Duan et al., 1992;Zhang et al., 2009).It is not only time-consuming but also expensive to generate hydrologic data for forecasting watershed hydrologic reactions.The better alternative to this lies in the development of runoff prediction models for ungauged watersheds with acceptable accuracy (Garg et al., 2003).Direct runoff from a watershed is influenced by the timing and spatial distribution of rainfall as well as the characteristics of the soil (Lee, 1998).Surface runoff occurs when the water-holding capacity of the soil decreases.Numerous field studies have demonstrated that infiltration excess runoff (Hortonian runoff) and saturation excess runoff (Dunne runoff) are the two main mechanisms that generate surface runoff.The studies also demonstrated that the spatial variability of soil properties, antecedent soil moisture, topography, and rainfall will result in different surface runoff generation mechanisms that produce storm hydrographs.Urban areas are more vulnerable to flooding under conditions of high rainfall intensity.Rainfall generated runoff is very important in various activity of water resources development and management such as a: flood control and its management, irrigation scheduling, design of irrigation and drainage networks, hydropower generation etc. (Sarita et al., 2013).Previous studies have shown that the soil surface slope and rainfall intensity affect the soil surface runoff rate, where the surface runoff rate increases with increasing the rainfall intensity and the soil surface slope (Dawood and Saeed, 2009;Mohammad and Al-Saleem, 2012;AL-jubouri and Al-douri, 2020;Ali et al., 2023;Abdul-Wahhab and Al-Jaberi, 2023).
In the present study, the effect of rainfall intensity and soil surface slope on the surface runoff rate of Karbala desert soil was investigated by using a laboratory watershed model.Also, The Gene Expression Programming (GEP) model was utilized to establish a runoff equation for Karbala desert soil.This equation is considered useful for hydrological studies required to develop this desert.

The Laboratory Watershed Model
The following is the contents of this model:

Rain Simulator
The rainfall are simulated according to the following procedures: The distribution of drop sizes should closely resemble that of a typical rainstorm.Rain should be uniformly distributed at the desired intensities of application.Raindrops ought to reach their natural rain terminal velocities.Each rain intensity was obtained by controlling the valve located after the water pump, and then the volume of falling water from the rain simulator accumulated inside a pan for one minute was calculated.Then the rainfall intensity is calculated by dividing the accumulated volume of water for one minute by the surface area of the soil incubator.This process was repeated for each test.
The conditions have been met to varied degrees by a variety of simulator types that have been proposed (Mutchler and Hermsmeier, 1965;Hall, 1970).After reviewing the literature and making attempts on many nozzles, the best nozzle in performance was chosen.It is made of polyethylene and consists of two parts: the reed and the reed cover.The reed is closed on one side and contains one hole, and the reed cover contains four holes as shown in Fig. 1.The rain simulator is made up of seven branch pipes that are internally connected to the four main plastic pipes to form a rectangular shape measuring 1 m by 1.5 m in size.Each branch pipe is 1 m long and has a diameter of 16 mm.An array of 35 nozzles were used, that is, each branch pipe carries 5 nozzles spaced 10 cm apart, and the first and last nozzles, which each are 10 cm apart from the main pipe.The distance between the branch pipes is 10 cm, and the pipes on the sides are apart 15 cm from the main pipe, as shown in Fig. 2.These distances between the spray nozzles were chosen because they achieved the best overlap and covered the entire study area.An inlet was placed in one of the main pipes to supply the system with water.

Catchment Area (Soil Incubator)
The catchment area (soil incubator) was made of plates with a thickness of 3 mm, a length of 1.8 m, a width of 1 m, and a height of 1 m, as shown in Fig. 3.An internal barrier of the same material, 0.75 m high, was placed at a distance of 1.5 m, and the rest of the length of the basin was 0.3 meters, in which a channel was made to collect surface runoff water.The amount of surface runoff is measured through the opening in the side of the channel, and the basin is provided with two glass windows on both sides to monitor the movement of the wetted side.
The soil incubator was designed with the above dimensions to build a laboratory watershed model of the Karbala desert soil that can be applicable to examine the surface runoff of the soil in a laboratory to represent reality.

The Experiments
After the completion of manufacturing the laboratory watershed model the rainwater representation system was connected to a water pump connected to four water tanks, the capacity of each tank was 250 liters.The line exiting the water pump branched into two branches, the first branch supplied the rainwater system with water and the other returned the excess water to the four tanks by manipulating the control valves that were attached to the pipe leaving the pump, in order to obtain different rainfall intensities as shown in Fig. 4. Four rain intensities were used in the current study (1.83 cm/min, 1.67 cm/min, 0.9 cm/min, and 0.64 cm/min).After that, the watershed (soil pod) is filled with soil transported from the Karbala desert in the form of layers and stacked.The height of the soil on one side of the soil slope was 75 cm, and the other side of the basin had a height of about (75 cm, 78 cm, 80 cm, and 85 cm), in order to achieve the slope of the soil surface.A metal wire clamp is placed at the beginning of the surface runoff collection channel to prevent soil erosion into the water collection channel.After completing the preparation of the soil sample for testing, a pan (1m x 1.5m) is placed over the catchment area (soil catchment) and the system is operated for one minute to measure the intensity of rain, after that the pan is removed from the soil catchment and the volume of the collected water inside the pan is measured.After that, the system is run over the soil for 128 minutes to begin measuring the amount of surface runoff occurred on the soil surface.Through this bottle, the surface runoff is measured.The operation time for the experiments is selected by conducting runs under different rainfall intensities until reaching the constant runoff rate and the longest time was selected.After the completion of the first experiment, the soil incubator is emptied and refilled each time for subsequent experiments.

Results and Discussion
The results of laboratory experiments are analyzed as follows:

Effect of the Surface Soil Slope on the Surface Runoff
The natural rainfall intensities did not give any surface runoff during laboratory experiments, so high values of rainfall intensities were used in order to get a surface runoff.The soil surface slope is also selected to be within the range of the ground surface slope of the Karbala region, which is between 0% and 11% according to the study of (Al Waeli and Mohsin, 2021).
Table 1, shows the values of surface runoff rates that were directly calculated using the laboratory watershed model and for different rainfall intensities (1.83 cm/min, 1.67 cm/min, 0.9 cm/min, and 0.64 cm/min) obtained from laboratory experiments, and for different slopes of the soil surface (6.7%, 3.3 %, 2%, and 0%).Also, Figs.5. to 8. show the variation of surface runoff rates with different slopes of the soil surface.
The greater slope of the soil surface should always produce greater runoff with the same intensity, but in these experiments, the opposite of this condition was obtained in some experiments.This is due to the fact that the studied soil contains 18% gypsum, in addition to the variation in the initial water content of the soil.These are two factors that affect the infiltration rate, as the higher the soil moisture content in the presence of gypsum, the infiltration rate increases, and thus a decrease in surface runoff occurs, and this is consistent with what was mentioned by (Kudayr and Salim, 2019).
In general, for all rainfall intensities, the surface runoff rate increases with the increase in soil surface slope, where the highest surface runoff rate was obtained at the highest soil surface slope of (6.7%), while the lowest surface runoff rate was obtained at the lowest surface soil slope of (0%).
It can be concluded that there is a direct relationship between the surface runoff rate and the slope of the soil surface for all rainfall intensities used in this study.The surface runoff rate of the rainfall intensity is 1.83 cm/min, under to the effect of the different slopes of the soil surface (0.0%, 2.0 %, 3.3 %, and 6.7%) Fig. 6.The surface runoff rate of the rainfall intensity is 1.67 cm/min, under to the effect of the different slopes of the soil surface (0.0%, 2.0 %, 3.3 %, and 6.7%) Fig. 7.The surface runoff rate of the rainfall intensity is 0.9 cm/min, under to the effect of the different slopes of the soil surface (0.0%, 2.0 %, 3.3 %, and 6.7%)The surface runoff rate of the rainfall intensity is 0.64 cm/min, under to the effect of the different slopes of the soil surface (0.0%, 2.0 %, 3.3 %, and 6.7%)

Effect of Rainfall Intensity on the Surface Runoff
Figs.9. to 12. show the variation of surface runoff rates with different rain intensities.
The results generally showed that the surface runoff rate increases with the increase in the rain intensity.Thus, the rain intensity of 1.83 cm/hr had the highest rate of surface runoff, whereas intensity of rain 0.64 cm/hr had the lowest rate of surface runoff.
The results show that the surface runoff rate and rainfall intensity have a direct correlation, as it is noticed that the depth of the wetness front increases with the increase in the intensity of the rain, that is, there is a direct relationship between the depth of the wetness front and the intensity of the rain.

Gene Expression Programming (GEP) Model
In this model, the data are randomly divided into two groups, training and testing (validation).The training set consists of 96 values representing about 75% of the total data, were used to develop the of GEP models, while the test set consists of 32 values about 25 % of the total data, were used to validate the of GEP models.The equation of the surface runoff rate is a function of the expression tree (ET) that is indicated in Fig. 13. and state in Eq. 1.
(1) In Table 3, the results show that the GEP model achieved a high value of R 2 with low values of RMSE and MAE for the training and testing data.Therefore, this model gives an excellent estimation of the surface runoff rate.

Fig. 1 .Fig. 2 .
Fig.1.Photo of the spray nozzle used in the rain simulator

Fig
Fig.5.The surface runoff rate of the rainfall intensity is 1.83 cm/min, under to the effect of the different slopes of the soil surface (0.0%, 2.0 %, 3.3 %, and 6.7%) Fig.8.The surface runoff rate of the rainfall intensity is 0.64 cm/min, under to the effect of the different slopes of the soil surface (0.0%, 2.0 %, 3.3 %, and 6.7%)

Fig. 14 .Fig. 15 .
Fig.14.Curve fitting between measured and predicted surface runoff for the training data

Table 2 .
Definition of parameters in (ET)

Table 3 .
The statistical parameters of the GEP model