Sustainable Water Management Using Rainfall-Runoff Modelling in Rift Valley Basin, East Africa

Agegnehu Kitanbo Yoshe ^*

1. Department of Water Resources and Irrigation Engineering, Arba Minch University, 21 Post Office Box, Arba Minch, Ethiopia

* Correspondence: Agegnehu Kitanbo Yoshe

Academic Editor: Wen-Cheng Liu

Special Issue: Advances in Hydrology, Water Quality and Sediment Simulation Modelling

Received: September 03, 2024 | Accepted: December 19, 2024 | Published: January 06, 2025

Adv Environ Eng Res 2025, Volume 6, Issue 1, doi:10.21926/aeer.2501003

Recommended citation: Yoshe AK. Sustainable Water Management Using Rainfall-Runoff Modelling in Rift Valley Basin, East Africa. Adv Environ Eng Res 2025; 6(1): 003; doi:10.21926/aeer.2501003.

© 2025 by the authors. This is an open access article distributed under the conditions of the Creative Commons by Attribution License, which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original work is correctly cited.

1. Introduction

Examining water resources and river basin problems stands at the forefront of environmental research, presenting a critical nexus between human societies and the intricate ecosystems they inhabit [1,2]. Sustainable water resource management has become increasingly urgent against the global challenges posed by climate change and population growth. Rivers, the lifeblood of countless civilizations, play an irreplaceable role in facilitating agriculture, industry, energy production, and freshwater supply [3]. Consequently, comprehension and mitigation of river basin issues are paramount for preserving biodiversity, protecting human livelihoods, and equitable access to these invaluable resources. The consequences of the unmitigated river basin issue can be far-reaching, with potential effects that extend both regionally and globally. Unpredictable surface runoff in river basins, stemming from changing precipitation patterns, deforestation, and transformations in land use, can lead to catastrophic floods or devastating droughts [4]. Communities residing along river areas face an increased risk of displacement, infrastructure damage, and economic losses, thereby exacerbating issues of social inequality [5,6]. In light of these considerations, administrative and research-oriented steps are essential to implement sustainable water resource management strategies. Such steps should delve into the complexity of river surface runoff, promote resilience, and foster harmony between human activities and the natural environment [7,8,9].

Hydrologists have faced a fundamental challenge in flood forecasting in ungauged and/or poorly gauged basins for decades. Simulating reliable continuous streamflow has been an essential and common objective for the scientific hydrological community since the middle of the 20th century. Such an issue is paramount for addressing engineering and environmental problems, which span from designing hydraulic structures to flood-risk assessment and stormwater management, from forecasting the climatic, groundwater Replenishment and anthropogenic effects on water resources to assessing economic benefits from hydropower production. Additionally, rainfall-runoff models are the most common tools for predicting discharge and estimating water balance. The study of rainfall-runoff estimation is a critical problem in hydrological modeling. Because of this context, it is recommended that accurate hydrological prediction is essential for water resource planning and management, mentoring lakes and dams to avoid soil erosion and to replace groundwater [10,11]. The problem that arises most frequently in hydrological research is the importance of estimating runoff from a river basin where there are precipitation records but no runoff records [12]. Hence, runoff is one of the most important hydrological variables used in most applications involving water resources [13]. Collecting this data is costly, time-consuming, and challenging [14], hence the advantage of hydrological models for estimating runoff from a river basin. So, the hydrological models are grouped into physical process-based models, empirical-based models, data-driven models, and conceptual models used for runoff estimation [15,16,17]. Conceptual hydrological models, sometimes called gray-box models, are precipitation-runoff models built based on observed or assumed empirical relationships among different hydrological variables. They differ from black-box models, which only statistically consider the precipitation-runoff relationship. They also differ from the physically distributed hydrological models based on solving differential equations describing the physical laws of mass, energy, and momentum conservation. Data-driven hydrological models are widely used for many practical purposes. However, the reliability of such models depends heavily on the strategy used to partition available observations into model calibration and evaluation subsets. An empirical model is built upon the observation of input and output without seeking to represent explicitly the process of conversion. In physically based hydrologic modeling, the hydrologic process of water movement is modeled either by the finite difference approximation of the partial differential equation representing the mass, momentum, and energy balance or by empirical equations. For example, the SCS-CN model, the Horton model, and the Green Ampt infiltration model, etc., are empirical methods [18]; the Sacramento model, the Stanford model, etc., are conceptual models [19]; SHE models, SWAT models, etc., are physical process-based models [20]; and data-driven models are machine learning and deep learning [21]. The Soil Conservation Service Curve Number (SCS-CN) method (SCS, 1993) is commonly used to estimate the depth of surface runoff in a watershed for a given rainfall event. The Natural Resource Conservation Service Curve Number (NRCS-CN) method has recently been renamed. The process is simple and valuable for ungauged watersheds and accounts for four major runoff-producing watershed characteristics: soil type, land use/treatment, surface condition, and antecedent moisture conditions. The method has been a topic of much discussion in hydrologic literature for the last three decades.

In the late 1940s, Sherman [22] proposed building a relationship between rainfall and runoff via unit hydrograph theory. Later, Mockus [23] suggested the inclusion of variables like land use, land cover type, soil, antecedent moisture condition, spatial distribution, etc. To evaluate the surface runoff based on infiltration rate, Musgrave [24] classified soil into four classes, namely A, B, C, and D. This classification was based on SCS-CN methods [25] that provide an empirical relationship to estimate runoff that considers initial abstraction based on land use, land cover, and soil type, called the SCS-CN method. The SCS-CN method is the most popular model due to its simplicity [26,27]. SWAT and HEC-HMS also use the SCS-CN model to calculate surface runoff [28,29]. Initially, the SCS-CN model was developed for small watersheds in the USA to estimate direct surface runoff from small rainfall events; then, it was used for several areas, climate conditions, and land uses [30,31]. Due to increasing urbanization, land use and land cover patterns drastically changed, which influenced SCS-CN and hydrological parameters affecting runoff from the catchments [32].

Advancements in geospatial technologies help include these variables for evaluating runoff from watersheds spatially and temporally. Different remote sensing products from satellites like IRS, NOAA, Landsat, TRMM, etc. can obtain reliable input parameters like precipitation, soil type, land use type, DEM, etc., from a regional to global scale [33]. In the present study, the SCS-CN model, which belongs to empirical models, was used due to its reliability and popularity among the hydrological models [34,35,36] with GIS; hence, the majority of researchers used GIS in conjunction with the curve numbers approach, which has been proven to be quick and accurate for estimating runoff [16]. This modeling of runoff and identifying the best locations for collecting water or recharging underground water using GIS and remote sensing have shown that this is an essential tool [37]. RS and GIS data are more dependable, modern, and quick than traditional techniques, and they are vital for gathering data on many elements of land use and soil type to create curve numbers, which are particularly important for runoff estimation [38]. Hence, hydrological research can be conducted accurately using GIS and RS technology, and GIS and RS data are becoming more prevalent in planning the development and management of natural resources. In particular, GIS supports decision-making by integrating multiple datasets and performing spatial analysis [39]. The curve number runoff model is helpful in predicting the direct runoff volume for any specific rainfall events using antecedent moisture conditions [40] and also for different tools and software developed to capture and process remote sensing and GIS data for runoff modeling. But either they are expensive or require large computer hardware [41]. Few free and open-source platforms are available for water resource studies [42] but require advanced skills. Presently, no model can include the temporal changes or dynamics of LULC changes in the process, and changes can only be accommodated by repeating the entire simulation process. This problem becomes more severe if the data to be analyzed is spatially and temporally larger or has multiple parameters. Spatial data have made it possible to accurately predict the runoff, which has led to significant increases in its use in hydrological applications. The curve number method [43] is adaptable and widely used for runoff estimation. This method is an essential property of the watershed, specifically soil permeability, land use, and antecedent soil water conditions, which are considered [44]. The SCS-CN approach is also commonly employed in studies measuring the influence of wildfires on the hydrological response of vegetated regions and consequent hydrological risks [45,46].

A chain of lakes and wetlands with unique hydrological and ecological characteristics characterizes the Rift Valley Basin. Largely foreign and national horticulture and floriculture enterprises have been settling in the area recently. Still, the crisis of water for drinking and irrigation purposes has adversely affected human society and has also been affected by drought events. Drought is a natural disaster in every climatic condition, adversely affecting socio-economic and environmental degradation [47]. In the study area, agriculture is utilized chiefly. It depends on rainfall systems, but rainfall deficiency is one of the significant fundamental causes of drought occurrences, and it is considered the meteorological drought, which is also responsible for the occurrences of hydrological, agricultural, and socio-economic drought. First, factors like topographical, climatic, and geological conditions are responsible for frequent drought occurrences and associated water scarcity in the study area. To overcome the water shortage problem in this water-scarce area, it is important to analyze rainfall-runoff, as it estimates the discharge rate of the river basin, to conserve water resources and ensure sustainable watershed management in the study area. No such investigation has been noted in the existing literature on the area studied. As a result, the techniques used, the methodology for estimating rainfall runoff, and the final result will be regarded as a new and honest contribution to the present study area. In general, this study aims to estimate rainfall runoff using the SCS-CN model in the Rift Valley basin through GIS and remote sensing. Therefore, the methodology applied in this study is quite helpful in characterizing rainfall runoff and associated water shortage problems, which is the novelty of this study. The present study would fill the research gap in rainfall-runoff dynamics in the data-scarred Rift Valley basin. Moreover, the present study will also fulfill the needs of different stakeholders for sustainable watershed management and water resource planning.

2. Method and Materials

2.1 Description Study Area

The study was located in the southern Ethiopian Rift Valley basin, located between 36°–40° E and 4°–9° N. A complex and rugged topography characterizes the basin. Based on the digital elevation model data, the elevation of the area ranges from 450 to 4190 m above sea level, and western and eastern escarpments bound it with altitudes of up to 4190 m above sea level. The basin contains lakes, forests, grasslands, wetlands, and savannahs, which provide a range of ecosystem services, e.g., as a natural habitat for diverse birds, fish, and wildlife species. It contains over 500 bird species of African, European, Arabian, and Asian origin and 33 fish species [48]. In particular, many lakes in the region are a focal point for internationally significant numbers of wildfowl and water and rated or endangered species. Additionally, BirdLife International has identified it as one of the most important bird habitats in the world [49]. Figure 1 shows the location map of the study area.

Figure 1 Location map of Rift Valley Basin in Ethiopia.

2.2 Methods and Data Collection

The methodology used to analyze the characteristics of rainfall-runoff for the study area using the GIS-based SCS-CN method is presented in Figure 2. To estimate rainfall runoff, different types of secondary data were used to generate various thematic layers, such as land use, slope, DEM, and soil textural map. The secondary data was collected from multiple sources. The main inputs required for the SCS-CN are as follows: watershed boundary and slope map from the digital elevation model from USGS datasets; the land use land cover map obtained from MCD12Q1.006 MODIS Land cover [50]; the hydrological soil group generated from the World Food and Agriculture Organization (HYSOGS250m) [51]; and rainfall data collected from Climate Hazards Group InfraRed Precipitation (CHIRPS) with station data [52]. In this model, land use is categorized by antecedent moisture conditions, which are integrated with GIS to estimate the curve number value of the hydrological soil group [53]. Then the rainfall-runoff for the watershed can be calculated using the computed weighted curve number incorporated into a GIS system for quick and precise assessment of runoff curve numbers [54,55].

Figure 2 Methodology for estimation of rainfall runoff in Rift Valley basin.

2.2.1 Precipitation

The most important source of fresh water on Earth is precipitation, a crucial part of the water cycle. Approximately 505,000 cubic kilometers of water falls as precipitation yearly, 398,000 cubic kilometers over the oceans and 107,000 cubic kilometers over land. So, the SCS-CN method uses the precipitation data as input to estimate river basin runoff and discharge [55]. But surface runoff creation is affected by precipitation. Therefore, when precipitation is excessive, there is a corresponding increase in surface runoff. However, the characteristics of the land surface, such as land use and soil textural conditions, can affect surface runoff either directly or indirectly [56]. Fuka et al. [57] highlighted that the weather value generated by CFSR for river basin simulations is as good as or better than the weather value used in simulations generated by conventional meteorological stations. For this study, area precipitation data was extracted from climate forecast system reanalysis (CFSR) [58]. The spatial distribution of the precipitation for the study area ranges from 33.812 to 141.79 and is presented in Figure 3.

Figure 3 Characteristics of precipitation in Rift Valley bain.

2.2.2 Digital Elevation Model (DEM)

The SCS-CN model requires DEM as a primary and necessary input variable to calculate flow accumulation, stream networking, catchment boundary, slope, and others [59]. Moreover, watershed delineation is the area that includes all the land and the contributors of runoff water to the central location. But every watershed has unique features, such as gradient, form, area size, drainage, vegetation cover, soil type, landforms, land use, and climatic conditions. Therefore, watershed management indicates the wise utilization of all the land and water resources in the river basin for optimal output with the least risk to the natural resources. Due to that, the DEM is an elevation model that restores the dataset in pixel or grid format. The spatial resolution used for this study is an arc-second 30m global digital elevation model from USGS (https://earthexplorer.usgs.gov), used in GIS and presented in Figure 4.

Figure 4 Characteristics of Digital elevation model in Rift Valley basin.

2.2.3 Slope

The slope is an essential topographical feature, explained by the contour area and horizontal spacing. It is crucial to determine surface runoff, infiltration, and groundwater recharge [60,61,62,63]. The slopes in the study area are extracted from the DEM using spatial analysis tools in the GIS environment. The hill has a vital role in calculating rainfall runoff and estimating groundwater recharges [64]. Reclassification of the slope depends on its steepness and flatness. When the slope is flat, the rainfall-runoff over the land surface is slow, which increases infiltration, whereas when the slope is steep, the rainfall-runoff is high with a low infiltration rate. The area's slope was classified into five classes and presented in Figure 5.

Figure 5 Slope characteristics of the Rift Valley basin.

2.2.4 Land Use Land Cover

Change in Land Cover and Land Use (LCLU) influences the runoff characteristics of a drainage basin to a large extent, affecting the area's surface and groundwater availability and hence leading to further change in LCLU. Therefore, the land use map is produced from the processing of the satellite Landsat 8OLI/TIRS USGS data (https://earthexplorer.usgs.gov, 2023) with a spatial resolution of 30 m, and the supervised classification is carried out to reclassify the land use land cover in the study area [65,66]. After supervised classification, the accuracy of the image classification was analyzed using the confusion matrix of the spatial analysis tool ArcGIS 10.4. Classification accuracy was calculated using the Kappa value (K), which is a statistical coefficient that is used for the estimation of classification accuracy. According to Demir and Keshin [67], cited by Ayhan et al. [68], when the kappa value is 75% or more, the classification accuracy is considered to be very excellent; when it is between 40% and 75%, it is deemed to be medium-good; and when it is below 40%, it is considered to be weak. The estimated kappa value for this study area was 96%, which is very good and acceptable for further hydrological analysis. The land use and cover for the study area are presented in Figure 6. Also, land use changes occur constantly and at many scales. They can have specific and cumulative effects on air and water quality, watershed function, waste generation, wildlife habitat extent and quality, climate, and human health.

Figure 6 Land uses land cover characteristics in the Rift Valley basin.

2.2.5 Hydrological Soil Group

Any water that cannot immediately seep into the ground flows downslope as runoff. Ground permeability affects runoff significantly. Hard-packed clay soils might absorb very little water, while loose sand might absorb almost all the precipitation that falls onto them, resulting in virtually no runoff. So, the soil group affects rainfall runoff and is an important variable for the estimation of rainfall runoff [69]; therefore, the generation of runoff is controlled by soil characteristics, which are determined by the hydrological soil group with the parameters of soil texture, average clay content, soil depth, and infiltration rate. The hydrological soil group is obtained using soil data [70]. The characteristics of the soil texture with rainfall runoff, water transmission, and infiltration rate are presented in Table 1. Figure 7 shows the study area's hydrological soil group extracted from the world's FAO soil map, which has a 1:500,000 scale digital soil map and was converted to a raster dataset, projected, and reclassified based on the hydrological soil group. For the study area, we observed two hydrological soil groups.

Table 1 Soil conservation service classification (Source: [71]).

Figure 7 FAO soil type, soil texture, and hydrological soil group in the study area.

2.2.6 Antecedent Moisture Condition

Soil moisture is a key control for runoff generation and biogeochemical processes on hillslopes. Precipitation events can evoke different soil moisture responses with depth through the soil profile, and responses can differ among landscape positions along a hillslope. The antecedent moisture condition shows the water storage in the soil at a specified time when it is affected by total precipitation over the five days before the precipitation event [72]. Therefore, it is key in determining the maximum curve number value. However, the antecedent moisture condition value aims to show how the infiltration curves influence both the volume and the speed of surface runoff [73]. Generally, SCS-CN shows three antecedent soil moisture situations and is represented as (I, II, and III) based on the change in soil condition and dormant and growth season rainfall amounts [74]. The antecedent soil moisture condition (I) shows the lowest surface runoff possibility due to sufficiently dry soils, whereas the antecedent soil moisture condition (II) represents the normal soil moisture condition, and the antecedent soil moisture condition (III) shows the highest surface runoff possibility. The curve number is determined by integrating the soil map and land use. So, this stage used antecedent moisture condition II, an average condition, to estimate the curve number value in the Rift Valley basin. The following equations (1) and (2) were used to calculate the values of curve numbers for antecedent soil moisture conditions I and III [75]:

\[ CNI=\frac{4.2\times CNII}{10-0.058\times CNII} \tag{1} \]

\[ CNIII=\frac{23\times CNII}{10+0.13\times CNII} \tag{2} \]

Where CNII is the curve number for everyday conditions, CNI is the curve number for dry conditions, and CNIII is the curve number for heavy rain and saturated soils. However, the SCS method was initially intended for 15-square-kilometer river basins. However, it has been altered to be used for bigger river basins by weighing curve numbers in relation to the river basins' land use areas [76]. The following equation (3) is used for the weighted curve number (CNw) [74]:

\[ CN_w=\frac{\Sigma(CN_i\times A_i)}{A} \tag{3} \]

Where CN_w is the weighted cure number, CNi is the CN from 1 to any number, Ai is the area with the curve number CNi, and A is the total area of the river basin. The land use and HSG were combined using ArcGIS 10.4 software to produce a new polygon with a new attribute table. However, using the SCS-CN Tables, the CN was provided for each polygon depending on the properties of the land use and soil type, and equation 3 was applied to estimate the CNw for each land used in the Rift Valley basin. This method was also used by Matomela et al. [76] for the Tayiba Valley basin in Egypt. Table 2 shows the categorization of antecedent moisture condition

Table 2 Categorization of antecedent moisture condition (Sources: [71]).

2.2.7 SCS-CN Model

There are a number of models for calculating runoff, but the Soil Conservation Service Curve Number (SCS-CN), developed in 1954 by the USDA SCS [77], is the most popular empirical model for estimating rainfall runoff and is widely used [70], also called the Hydrological Soil Cover Complexes Model [43]. This model is based on water balance estimation, which depends on two input variables: precipitation and curve number [78]. The soil conservation service curve number approach is frequently used empirically to estimate the direct runoff from a watershed [43]. Moreover, it also takes into account two primary presumptions. First, the ratio of surface runoff quantity to maximum rainfall is equal to the ratio of infiltration capacity to the volume of maximum retention. Secondly, the initial abstraction is one factor that could possibly result in maximum retention [40,79]. The losses of infiltration are combined with external storage based on equation (4) [80].

\[ Q=\frac{(P-I_a)^2}{P-I_a+S} \tag{4} \]

Where Q is surface runoff quantity, Ia is initial abstraction in mm before the onset of surface runoff and includes surface for storage, P is the depth of precipitation in mm, and S is the maximum possible retention after surface runoff in mm. The relationship between Ia and S is shown by equation (5) [81].

\[ I_a=0.2S \tag{5} \]

S can be obtained after knowing the CN; it is shown below by equation (6) [81].

\[ S=\frac{25400}{CN}-254 \tag{6} \]

The idea of CN is the curve number, which ranges from 0 to 100 [40] and is obtained from the table in accordance with the SCS manual of hydrology, relying on the land use and hydrological soil group consisting of four groups. Such as A, B, C, and D, antecedent soil moisture condition (AMC). So, the SCS curve number calculates how the soils allow water infiltration in relation to land use and antecedent soil moisture conditions [82]. Now, the equation can be rewritten as equation 7 below.

\[ Q=\frac{(P-0.3S)^2}{P+0.7S} \tag{7} \]

Significantly higher the value of CN, the runoff from the watershed was calculated from the two equations above (6 and 7). Then, the rainfall runoff for the lake basin was calculated after estimating the necessary parameters.

2.2.8 water Management Using Rainfall-Runoff Modelling

The following Table 3 shows the related works for water management using rainfall-runoff modeling.

Table 3 Related works for water resource management using SCS-CN methods.

3. Result

This study used the SCS-CN model for rainfall-runoff evaluation in the Rift Valley basin. The rainfall-runoff generation model is highly complex, nonlinear, dynamic in character, and affected by different interconnected physical factors. Hence, precise runoff estimation is carried out for valuable water resource management and development. In the Rift Valley River basin, drought-like situations prevail yearly due to low average annual precipitation and high runoff. The basin constitutes different land use land cover classes, and present studies show that rangeland is the dominant land use class, followed by cropland. In contrast, flood vegetation covers the lowest area in the study region. The built area and bare land play a crucial role in the rainfall runoff among the other land-use land cover types. The hydrological soil group also has important functionality in estimating the rainfall runoff potential and showing the soil character, category, and capability of infiltration rate. In this study, the hydrological soil groups A, B, C, and D were prepared based on the FAO soil map, geospatial data, and other secondary data. The obtained result shows that the C and D types of hydrological soil groups are mainly covered throughout the Rift Valley River basin. By overlay analysis of land use, land cover, and hydrological soil groups in the ArcGIS environment, the curve number values were assigned according to the USDA SCS, and the AMC values are AMC I, AMC II, and AMC III. The characteristics of precipitation, evapotranspiration, temperature, and soil moisture with rainfall runoff were evaluated.

3.1 Land Use Land Cover

The land use cover of our study is grouped into six land use types such as water bodies, trees, flooded vegetation, crops, built-up areas, bare land, and rangeland. As presented in Table 4 and Figure 6, the land use land cover map demonstrates that nearly 22763.82 km² (39.53%) of the Rift Valley basin area has been covered by rangeland, 19012.21 km² (30.01%) of the area is covered by cropland, 9003.65 km² (15.63%) is covered by trees, 3122.07 km² (5.42%) is covered by built-up areas, 2556.65 km² (4.44%) is covered by water bodies, 984.29 km² (1.71%) is covered by bare land, and 144.3 km² (0.25%) of the area is covered by flooded vegetation. One of the most crucial factors considered when estimating rainfall runoff is the type of land use, which significantly affects the pattern of rainfall runoff in any geographical region over time. This is because different types of land use affect a river basin differently, which also impacts its infiltration capacity and rate of rainfall runoff, which agrees with similar findings [87,88]. Areas with forest cover generally increase the infiltration rate and reduce rainfall runoff, whereas bare land surfaces or built areas increase rainfall runoff and lower the infiltration rate [87]. Recent studies conducted in Narok town revealed that land use and land cover change directly affect rainfall runoff intensity [89].

Table 4 Characteristics of land use land cover and its area coverage for the study area.

3.2 Hydrologic Soil Group (HSG)

Soil has been grouped into four different hydrological soil classes, such as A, B, C, and D groups, according to infiltration rate, texture, water transmission capacity, drainage order, and transmissivity of water based on the USDA-SCS [90]. In this study, five textural soil classes were observed, such as clay, clay loam, loam, silty loam, and water, presented in Table 5 and Figure 7. As demonstrated in Table 4, the corresponding hydrological soil group has been assigned to each textural polygon, and two hydrological soil groups are identified using MWSWAT2012, i.e., C and D (Figure 7 and Table 4). The characteristics of the hydrological soil groups are specified regarding each soil group's reasonable infiltration rate, runoff potential, and water retention capacity in Table 4 based on USDA-SCS [87]. Surface runoff is generated when the movement of water into the soil is slow enough to result in water flow along the land surface and into water bodies. Areas with water bodies generally decrease the infiltration rate and increase rainfall-runoff, whereas regions with loam soil increase infiltration and reduce rainfall-runoff, consistent with previous studies [91].

Table 5 The characteristics of hydrological soil texture, hydrological soil group and area coverage.

3.3 Curve Number (CN)

For this study, rainfall--runoff curve numbers have been taken from a hydrological modeling system [92]. The value of the curve number ranges from 0 to 100; a lower value shows low surface runoff with a high infiltration rate, whereas a higher value presents adequately higher rainfall runoff with a lower infiltration rate [93,94]. For this study, the area curve number ranges from 74 to 80. The CN with low values is generally given to areas covered by trees, crops, and range land, whereas the CN with high values in the area is given to bare land, built areas, flooded areas, and water bodies. The different types of land use and land cover related to their respective hydrological soil groups and curve number values are presented in Table 6. Additionally, using equation 3, it has been found that the weighted curve number for the study area is 74.71, which is represented as AMC-II for normal soil moisture conditions. The curve numbers for AMC-I and AMC-III were calculated by equations 1 and 2, which give a result of 55.37 and 87.20, respectively, presented in Table 5. The rainfall-runoff curve number is an empirical variable used to calculate rainfall runoff in particular river basins, which agrees with similar studies [82,95,96].

Table 6 Distribution of land use land cover, hydrological soil group, and respective curve number values.

After the CN value for each antecedent moisture content was evaluated, the probable maximum retention capacity (S) and preliminary abstraction loss (I_a) were estimated using equations 6 and 5, respectively. Their results are presented in Table 7.

Table 7 Estimation hydrological calculation for the Rift Valley basin (probable maximum retention, (S), preliminary abstraction loss (I_a) and CN_w).

3.4 Characteristics of Slope

For rainfall runoff estimation, the slope of an area is the foremost regulatory factor that determines the surface runoff and associated rate of infiltration capacity. Areas with a very steep slope permit a high rate of surface-runoff with a low infiltration rate, whereas areas with a gentle slope permit slow runoff with a high infiltration rate [97]. This study's topography is characterized by undulating, rugged, hilly terrain. The area's slope varies from 0 to 75.08 percent, as shown in Table 8. Figure 6 shows the slope map of the study area. Based on the characteristics of slope, a very low to high rate of runoff was estimated; i.e., slow runoff was estimated in areas with a low slope, whereas high runoff was estimated on the steepest slope of the hilly areas. 22.64 percent of the area was covered by nearly level terrain experiencing low rainfall-runoff, whereas 14.65 percent of the study area was covered by very steep sloping terrain experiencing very high rainfall runoff. The full details of rainfall runoff related to the characteristics of the slope in the area are presented in Table 7.

Table 8 Characteristics of the slope with landform and runoff.

3.5 Precipitation

Runoff is the proportion of rainfall that does not infiltrate and is not taken up by evapotranspiration. Precipitation serves as an important source of water and is one of the fundamental components of hydrology. However, the curve number method uses precipitation data to estimate river basin runoff and discharge [17]. Moreover, rainfall runoff creation is affected by the precipitation in the area. So, when precipitation is high, there is a corresponding increase in surface runoff. However, surface characteristics, like land use type and the textural class of the soil, can affect rainfall runoff directly or indirectly [56]. Hence, these values from 1991 to 2022 were collected from the climate forecast system reanalysis (CFSR): https://climatedataguide.ucar.edu/climate-data/climate-forecast-system-reanalysis-cfsr. The spatial variation of the precipitation for the study area is presented in Figure 3. Figure 3 shows a low spatial distribution of precipitation in the southern parts of the river basin, whereas high precipitation was observed in the same parts of the midstream section. Additionally, Fuka et al. [57] evaluated that the weather value generated by CFSR for river basin simulations is as good as or better than that used in simulations generated by conventional meteorological stations. The cumulative precipitation for this study is presented in Figure 8. Using multiple regressions, the statistical parameters of precipitation were evaluated. From the estimated result, the R-multiple is 0.11, the R-square value is 0.012, the adjusted R-square is -0.021, and the standard error is 119.13. The trend of precipitation shows rising and falling trends during the study period. The maximum yearly precipitation was observed in 2020 with a value of 1165.46 mm, followed by 1088.84 mm in 1997. In contrast, the minimum precipitation was observed in 2021 with a value of 673.22 mm, followed by 692.08 mm in 1999. The region's uneven precipitation results from changing weather patterns, which may bring more extreme weather events and lengthen drought and flood seasons, consistent with similar previous studies [1,98].

Figure 8 Annual precipitation for Rift Valley River basin from 1991 to 2021.

3.6 Estimation of Rainfall-Runoff for the Study Area

Using the available daily precipitation data from 1991 to 2021, the rainfall-runoff for the Rift Valley basin was calculated using equations 6 and 7 of the SCS-CN method. From the precipitation, we observed that the cumulative annual precipitation ranges from 673.22 to 1165.46 mm from 1991 to 2021, and the estimated average rainfall runoff varies between 562.70 and 1047.1 mm, as presented in Table 8. For the estimated result, the average volume of rainfall-runoff for the Rift Valley basin ranges from 32403589400 to 60303064900 Cubic Meters. The result in Table 8 shows cumulative precipitation, rainfall-runoff for AMC-I, II, and III, average rainfall runoff for the three-moisture condition, and the volume of rainfall-runoff for the study area. The correlation coefficient between precipitation and estimated rainfall runoff was 1, which shows precipitation was strongly correlated with rainfall runoff. This strong correlation shows that when precipitation increases, rainfall-runoff will also increase, and vice versa. The yearly rainfall-induced runoff rate is presented in Table 9. The result shows that maximum runoff occurred in 2020 with a value of 1047.11 mm, whereas the minimum rainfall runoff was observed in 2021 with a value of 562.66 mm.

Table 9 Estimated rainfall runoff for (AMC-I, II, III, average) in mm, Volume in cubic kilometers and cumulative precipitation in mm.

The long-term temporal variation of rainfall runoff for the study area shows rising and falling trends for the study durations. The variation in rainfall-runoff for the study area was observed due to the variation in precipitation caused by climatic change and greed, with similar previous findings [6,35,98]. Figure 9 demonstrates the temporal variation of rainfall runoff in the Rift Valley basin from 1991 to 2022.

Figure 9 Estimated average rainfall-runoff characteristics in the Rift Valley basin.

3.7 Spatial Distribution of Rainfall-Runoff in Rift Valley Basin

The estimated spatial distribution of rainfall runoff in the Rift Valley basin from 2011 to 2022 is presented in Figure 10. As presented in Figure 10, the low spatial distribution of rainfall-runoff has been observed in the downstream sections of the river basin for the entire year. As shown in Figure 10, most of the river basin experiences low rainfall-runoff, whereas a few parts of the river basin experience high rainfall runoff. As indicated in Figure 10, the rainfall-runoff of the river basin ranges from 1.25 to 71.75 mm in 2011, 2.08 to 66.67 mm in 2012, 2.08 to 74.25 mm in 2013, 1.5 to 78.5 mm in 2014, 1.5 to 64.33 mm in 2015, 1.33 to 62.08 mm in 2016, 1.58 to 56.92 mm in 2017, 2.08 to 72.83 mm in 2018, 1.33 to 57.5 mm in 2019, 1.92 to 100.7 mm in 2020, 0.83 to 38.08 mm in 2021, and 0.67 to 29.83 mm in 2022. As presented in Figure 10, the lowest rainfall runoff was observed in 2022, with a value range of 0.67 to 29.83 mm, whereas the highest rainfall runoff was observed in 2020, with a value range of 1.92 to 100.7 mm. As observed in Figure 10, the variation in rainfall-runoff for this study is due to land use, land cover, soil type, anthropogenic activities, and climatic change, consistent with similar previous studies. A similar result was achieved by Umukiza et al. [89], who demonstrated how the configuration and composition of land use can affect rainfall runoff and flow volume in selected catchments of Narok Town, Kenya.

Figure 10 Spatial distribution of rainfall runoff in the Rift Valley basin from 2011 to 2022.

4. Discussions

4.1 Relationship between Precipitation, Evapotranspiration, Temperature, and Rainfall Runoff

Figure 11 presents the relationship between rainfall-runoff, precipitation, evapotranspiration, and temperature. As indicated in Figure 11, precipitation and rainfall runoff follow a similar trend. As precipitation is the source of the hydrological cycle, increasing precipitation leads to increasing rainfall runoff, and decreasing precipitation leads to decreasing rainfall runoff, which shows a direct relationship that agrees with previous similar studies [1,99,100]. As demonstrated in Figure 11, the relationship between evapotranspiration and rainfall runoff shows significant and varying trends. As observed in Figure 11, at certain times, evapotranspiration and rainfall runoff show a similar trend. Still, at other times, they show variation in trend, which depends on the seasonal variation of the environment. During the rainy season, the effect of evapotranspiration on rainfall-runoff is insignificant. Still, during the dry season, the effect of evapotranspiration is high. We agreed that a similar trend was observed during the wet season due to the low impact of evapotranspiration on rainfall-runoff. In contrast, a varying trend is observed during dry and drought conditions, confirmed by similar previous studies [101]. As presented in Figure 11, increasing temperature leads to a decrease in rainfall-runoff, and a reduction in temperature leads to an increase in rainfall runoff. This is due to climatic change and is confirmed by similar findings [1,102]. As presented in Figure 11, the relationship between soil moisture and rainfall runoff shows a similar trend. As soil moisture increases, the rainfall-runoff in an area also increases, which agrees with previous studies [103].

Figure 11 Relationship between precipitation, soil moisture, evapotranspiration and temperature with rainfall runoff.

Understanding the hydrological behaviors of river basins is an important part of effective river basin management and planning. In this study, rainfall-runoff modeling is carried out using CN-SCN methods in the Rift Valley basin for water resource management and optimal utilization. Different datasets are collected and utilized to evaluate rainfall runoff for the study area, like precipitation, land use, land cover, slope data, and soil datasets. As presented in Figure 6, the land use cover of the study area is classified into seven classes: rangeland, bare ground, water bodies, trees, flooded vegetation, crops, and built area. A comprehensive understanding of the impact of land use and land cover change, especially on hydrology, is important for developing sustainable water resource management strategies that optimize economic and social welfare without jeopardizing ecosystem sustainability. The impact of land use and land cover change on hydrological responses is critical and also presented by other findings [104,105,106]. As presented in Table 4, the study area covered 4.44 percent by water, 15.63 percent by tree, 0.25 percent by flood vegetation, 33.01 percent by crop, 5.42 percent by built area, 1.71 percent by bare land, and 39.53 percent by rangeland. Recent studies in Narok town revealed that land use and land cover change directly affect rainfall runoff intensity [6]. As presented in Figure 7, the hydrological soil groups of the study area are grouped into C and D, with five textural classes (clay, clay loam, loam, silty loam, and water). Soil physical characteristics affect water infiltration rate into the soil and influence rainfall runoff in any area. Surface runoff is generated when the soil is saturated. In clay soil, rainfall-runoff is higher than in sandy soil due to the permeability of the soil porosity, which allows the downward movement of water to recharge groundwater and reduce rainfall runoff. A similar result was reported by Chen[107] in hillslope cropland of purple soil in China. They reported that sand, loam sand, and sandy loam have a minimum infiltration rate of more than 7.26 mm per hour, whereas clay and silty clay have a minimum infiltration rate range of 0 to 1.24, which affects rainfall runoff. Table 5 and Figure 7 show that the study area accounts for 34.26 percent of the D hydrological soil group and 65.74 percent of the C hydrological soil group. On land, surface precipitation is one of the most natural resources for obtaining fresh water and creating rainfall runoff. It is an essential part of the hydrological cycles, and for the study area, the spatial distribution of precipitation ranges from 31.69 to 142.5 mm and the year precipitation ranges from 673.22 to 1165.46 mm, as presented in Figure 3 and Figure 8, respectively, which are important parameters for the estimation of the spatial and yearly variation of rainfall runoff in the study area. As presented in Figure 3, some parts of the river basin in the midstream section are experiencing very high precipitation, the southern parts are experiencing very low precipitation, and the rest are experiencing medium to low precipitation. As precipitation increases, the rainfall-runoff also increases, whereas as precipitation decreases, the rainfall runoff also decreases, which is also reported by similar previous studies [1,108,109,110]. Soil moisture before rainfall will significantly affect the rainfall-runoff capacity and CN values of the area because it can check and balance the infiltration rate and quantity of moisture. Considering the effects, using a 5-day antecedent precipitation index, we evaluated the antecedent soil moisture content for dry, normal, and wet conditions, and the results are presented in Table 5. As presented in Table 5, the CN for AMCI is 55.37, AMCII is 74.71, and the CNIII for AMCIII is 87.20. The CN with low values is generally given for areas having crops, trees, and rangeland; however, the CN with high values is for built areas, bare land, water bodies, and flooded areas due to its contribution of rainfall runoff, which agrees with previous studies [1,111]. As presented in Table 7, the probable maximum retention capacity and preliminary abstraction loss were evaluated to estimate rainfall runoff for the study area. Maximum potential retention represents the maximum amount of water in mm retained on the watershed surface after rainfall runoff begins. This is an important variable in the CN-SCS method of runoff estimation. It is not a directly measurable watershed variable. Preliminary abstraction loss is a parameter that accounts for all losses before runoff and consists mainly of interception, infiltration, evaporation, and surface depression storage. In theory, all rainfall minus preliminary abstraction loss will generate the runoff from a specified river basin. As presented in Table 7, the maximum probable retention for AMC-I is 213.73, AMC-II is 94.98, and AMC-III is 46.38, whereas the preliminary abstraction loss is 42.75 for AMC-I, 19.0 for AMC-II, and 9.28 for AMC-III. As shown in Table 7, the maximum value of maximum probable retention and abstraction losses were observed for AMC-I. This is because before rainfall runoff begins, it balances the soil moisture condition in the soil and maximum infiltration and decreases as precipitation increases, which agrees with similar findings [99,110]. The slope is an important parameter that affects rainfall runoff. As the slope of the area increases, infiltration decreases with increasing rainfall-runoff, whereas decreasing the land surface slope increases the soil infiltration rate and reduces rainfall runoff. As presented in Table 7 and Figure 6, the slope of the study area ranges from 0 to 75.08 percent. As demonstrated in Table 7, 22.64 percent of the study area slope ranges from 0 to 4.42 with nearly level rainfall-runoff, 26.69 percent of the area slope ranges from 4.43 to 10.31 with a very gentle slope having moderate rainfall-runoff, 19.7 percent of the area slope ranges from 10.32 to 18.26 percent with gently sloping having moderate rainfall-runoff, 16.33 percent of the study area slope ranges from 18.27 to 28.27 with moderate sloping with high rainfall runoff, and 14.65 percent of the area slope ranges from 28.28 to 75.08 with strong sloping creating very high rainfall runoff. We understand that under the condition of equal horizontal projective lengths so that all slopes receive the same amount of rainfall, there are increases in infiltration with increasing slope gradients and decreasing rainfall runoff, and the steeper the slopes, the faster the rainfall-runoff, which is consistent with similar studies [1,99,110,112].

For this study, soil conservation service curve number rainfall runoff was generated based on precipitation, abstraction loss, and probable maximum soil retention potential, and these datasets are used as input for the ArcGIS-based SCS-CN rainfall-runoff simulation model. As presented in Table 8, the estimated rainfall runoff ranges from 562.70 to 1047.10 mm, and the temporal variation of rainfall-runoff during this study period is observed due to changes in precipitation and climatic change, which agrees with previous similar studies [1,98]. As presented in Figure 10, the spatial distribution of rainfall-runoff ranges from 0.67 to 100.7 mm during the study period, and the variability of this spatial distribution of rainfall runoff is observed due to the variability of precipitation across the study area. The curve number resulted from land use, land cover, hydrological soil group, anthropologic activity, and climatic change, which agrees with similar previous findings [98,113,114]. Rainfall-runoff estimation from a river basin is used to develop hydraulic structures, regulate soil erosion, analyze the water supply potential of a river basin, and develop groundwater recharge by collecting rainfall runoff in artificial recharge areas. The SCS-CN method is probably a very well-defined and straightforward conceptual model for calculating rainfall runoff from precipitation. Therefore, in the current study, we utilized the SCS-CN method to estimate rainfall runoff in a drought-prone area of the Rift Valley basin, relying on the daily precipitation from the basin for 31 years, i.e., 1991–2022, where the SCS-CN method is the most popular for estimating the volume of rainfall-runoff for a certain precipitation event, and this method offers more dependability. It is a fact that the SCS-CN model has been widely used across the globe under different land use, land cover, and climatic conditions to evaluate rainfall runoff since its development [75,115,116]. Although this model has been used for a range of applications outside of its initial focus, such as rainfall-runoff analysis in vast river basins and integration into daily and long-term hydrological models [117,118]. Different researchers have done their work using the SCS-CN method to calculate rainfall runoff, and their output is optimal for sustainable watershed management [31,38,75,119]. Also, currently, machine learning and various approaches have been used to calculate rainfall runoff in the natural river basin [19,107,120,121]. As the SCS-CN is primarily affected by precipitation, land use, land cover, hydrological soil group, and antecedent soil moisture condition, all of these variables are utilized as input variables for the estimation of rainfall-runoff by the SCS-CN method to create and evaluate the weighted curve number in the Rift Valley basin. In this undulating drought-prone watershed, the rainfall-runoff is robustly correlated with an R-value of 1. Therefore, there is high seasonal precipitation during the wet season (April to September), and the lowest rainfall runoff is estimated in the dry season for the study area from November to February. Figure 12 shows the rainfall-runoff characteristics for the three antecedent moisture conditions with precipitation in the study area. The estimated rainfall runoff for the study area is directly related to the precipitation in the area. As presented in Figure 10, the spatiotemporal variation of the rainfall-runoff in the Rift Valley basin shows significant variation during the study period. As shown in Figure 10, the majority of the study area experiences very low rainfall-runoff, whereas very few parts of the study area demonstrate very high rainfall runoff. As indicated in Figure 10, the rainfall-runoff of the river basin ranges from 1.25 to 71.75 mm in 2011, 2.08 to 66.67 mm in 2012, 2.08 to 74.25 mm in 2013, 1.5 to 78.5 mm in 2014, 1.5 to 64.33 in 2015, 1.33 to 62.08 mm in 2016, 1.58 to 56.92 mm in 2017, 2.08 to 72.83 in 2018, 1.33 to 57.5 mm in 2019, 1.92 to 100.7 mm in 2020, 0.83 to 38.08 mm in 2021, and 0.67 to 29.83 mm in 2022. As presented in Figure 10, the lowest rainfall runoff is observed in 2022, with a value range of 0.67 to 29.83 mm, whereas the highest rainfall runoff is observed in 2020, with a value range of 1.92 to 100.7 mm. As observed in Figure 10, the variation in rainfall-runoff for this study is due to land use, land cover, soil type, anthropogenic activities, and climatic change, consistent with similar previous studies. A similar result was achieved by Umukiza et al. [89], who demonstrated how the configuration and composition of land use can affect rainfall runoff and flow volume in selected catchments of Narok Town, Kenya. The relationship between precipitation, evapotranspiration, soil moisture, and temperature with rainfall runoff was evaluated. As presented in Figure 11, precipitation and soil moisture show a similar trend with rainfall-runoff, whereas evapotranspiration and temperature show a significant and varying trend with rainfall-runoff. The variation of rainfall-runoff is mostly observed due to the variability of precipitation, temperature, and other climatic parameters, consistent with similar previous findings [1,99,100,101,103,122].

Figure 12 Characteristics of rainfall runoff with precipitation for AMC-I, II, and III in Rift Valley basin.

5. Conclusion

This paper investigates the characteristics of rainfall runoff in the Rift Valley basin using the SCS-CN method with integrated ArcGIS, which can effectively be used in water resource management. The land use, land cover, soil data, slope, and precipitation data were used to evaluate rainfall runoff in the Rift Valley basin. For the generated land use, land cover, and hydrological soil group, antecedent moisture conditions for wet, normal, and dry conditions were evaluated, which were equal to 54.96 for AMC-I, 74.39 for AMC-II, and 86.98 for AMC-III. In addition to the curve number, probable maximum retention and preliminary abstraction loss for the estimated curve number were evaluated. The estimated probable maximum retention is 213.73, 94.98, and 46.38 for AMC-I, II, and III, respectively, whereas the preliminary abstraction loss is 42.75, 19.00, and 9.28 for AMC-I, II, and III, respectively. The estimated average rainfall runoff in the Rift Valley basin varies between 562.70 and 1047.1 mm, with an average volume of 32403589400 to 60303064900 Cubic Meter from 1991 to 2021. The spatiotemporal characteristics of rainfall runoff in the Rift Valley basin are also evaluated, and the results range from 0.67 to 100.7 mm. The characteristics of precipitation, temperature, soil moisture, and evapotranspiration with rainfall runoff are evaluated in the study area. In addition to that, this study highlights the importance of using SCS-CN GIS-based methods and GIS and remote sensing data. So, the GIS SCS-CN method merged with remote sensing, was utilized to calculate surface runoff in the Rift Valley basin. Hence, the rainfall runoff depth and volume in the Rift Valley basin will provide surface water availability. In addition, the potential is important for various uses, such as domestic water supply, irrigation, and power production in the area, as well as the construction and design of drains, bridges, and others. In conclusion, the Soil Conversations Service-Curve Number approach is efficiently proven to be a better method, which consumes less time and has the facility to handle an extensive data set and a larger environmental area to identify the site selection of artificial recharge structures.

Sedimentation affected the Lake Chamo water body, and our work's future prospect was to evaluate sedimentation for effective water resource management and ecological restoration.

Abbreviations

Author Contributions

Conceptualization, methodology, original draft writing, preparation, and development of model was curried by the A. K. Yoshe.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or non-profit sectors.

Competing Interests

The author declares no competing interests.

Data Availability Statement

The dataset utilized for the estimation of rainfall runoff during this study are available from the corresponding author on reasonable request.