runoff estimation

Determination of Potential Runoff Using GIS-based SCS-CN and Remote Sensing Abstract Runoff estimation is one of the necessary procedures required in a watershed to achieve soil control measures, reservoir operation, and hydraulic structures. For instance, in Saudi Arabia, as a case study, there is limited research on accurate runoff, a challenging aspect for hydrologists. The region forms a good part for case study analysis and development of further research on runoff information. Therefore, Wadi-Rahjan in Saudi Arabia is a crucial case study for the topic, as one of the several catchment areas in the area, especially in the western region of the site. The region is primarily selected in the contemporary study to act as a case study on runoff estimation. The principal methodologies that are fundamental to developing the research in the chosen field include and are not limited to geographical information system (GIS), remote sensing, and soli conservation services curve number (SCS-CN). There are multiple parameters that researchers embrace for the study: they comprise rainfall data, hydrologic soil groups (HSGs), land use/land cover (LULC), and digital elevation model (DEM). Therefore, LULC and HSG layers are critical elements that researchers have employed in estimating the catchment response to storm events through curve number (CN). The runoff potential map is a vital component for the research generated from the SCS-CN model and GIS-based model, which the primary inputs for the study are rainfall and derived CN map. The research also classified the study area into three main phases, HSGs B, C, and D, based on the result calculations. 90 is the average CN for everyday conditions, with dry and wet condition averages being 80 and 97 in that order. The SCS-CN model calculations and findings postulate that the yearly runoff varied from 194 to 295mm. The runoff ranges present the highest percentages of runoff water, with results as follows: between 289 and 295mm, there was 35 percent, 269 to 288mm resulted to 24

percent. Therefore, the overall approach is applicable in other Saudi Arabia watershed areas to facilitate development and planning purposes. Keywords: GIS, remote sensing, SCS-CN, Runoff. Introduction Among the worldwide regions that suffer from water scarcity is Saudi Arabia, with the leading causes being increased non-renewable groundwater abstraction and rainfall absence. Also, the parts face critical issues with its runoffs, especially heading to the deserts and the sea, without mobilization and usage. Besides, the region faces inadequate information on and research on watershed runoff records and hydrological studies, which are pivotal to triggering development and planning. Modern study and research postulate fundamental intervention to soil conservation services centered on the GIS-based model and SCS-CN methodology [1,2]. Although current research and historical records show insufficient data on useful information regarding how experts can save Saudi Arabia runoffs, the process is possible through embracing the SCS-CN method, which does not rely on records. The SCS-CN model has been verified successfully by numerous professionals. The tool is one of the vital developments by the United States hydrologists in the Agriculture department [3-5]. The mechanism, the SCS-CN model, utilizes curve number (CN), a prominent land condition factor. The curve number (CN) primarily involves parameters convertible into numerical values included in a developed formula; the result is used in the watershed direct runoff estimation [6]. Two results are present from the formulation using the CN; high curve numbers indicate low infiltration and increased runoff, while smaller curve numbers indicate high infiltration and minimal runoff. In the SCS-CN model, the central factors that affect it include soil data, land use/cover, and rainfall [7-9].

Current research and analysis methods incorporate modern technology to provide better results and advance traditional critical techniques, such as the SCS-CN model. Since the conventional SCS-CN model is time-consuming and tedious, modern experts incorporate it with the GIS method to achieve more accessible detailed analysis [4]. Multiple experts []10-12 embraced critical techniques such as GIS mechanisms and remote sensing in runoff and CN estimation. The researchers confirmed the instrument's quickness, reliability, and relativity in the easy assessment of composite CN and watershed runoff. Other critical researchers and scholars such as Geena et al. [2] are innovative in developing other region’s curves that suit their conditions, such as Indian conditions, embracing the original skills in GIS and SCS-CN primary the Red hills watershed. The authors argued that RS tools and GIS are applicable in estimating temporal variables and hydrological parameters. For instance, Indian Uri River study, which incorporates inadequate hydrological data, Nayak et al. [13] argue that SCS-CN and RS models make faster and better runoff estimations, hence establishing a good correlation between computed and observed runoff findings. In the western Saudi Arabian side, prominent experts such as Al-Ghobari et al. [14] develop the study by employing SCS-CN, GIS-based, and RS models in surface runoff simulation. The researchers reported that GIS, RS, and SCS-CN integration in the field study is critical in runoff estimation; they recommended that the technique combination is fundamental for planning and basins conservation and management purposes. Additional research by Shadeed and Al Masri [4] shows a comparison between GIS and SCS-CN runoff estimation and runoff measured data to draw a contrast between the techniques. The researchers focused primarily on the methods’ accuracy, postulating that SCS-Cn and GIS- based model yielded 85 percent accuracy, a reliable runoff estimation model. Additionally, Ningaraju et al. [15] employed the SCS-CN method in a new region, a watershed in Kharadya, Mandya, Karnataka district. The professionals found a positive

correlation between rainfall and runoff estimation spanning eleven-year period datasets. In Sudan Khartoum state, Mohamed and Aldona [16] conducted central research and runoff estimation. The researchers observed the runoff and advocated for a strong correlation and high runoff estimation accuracy, thanks to the SCS-CN method. Tirkey et al. [17] conducted estimation research in India, Jharkhand, which yielded accurate SCS-Cn runoff estimation capabilities. Lastly, Liu and Li [18] tested the SCS-CN mechanism under Loess Chinese plateau watershed and found legitimacy in the model through its effective runoff stimulation and generation. Therefore, the study's central focus is to map the standard Wadi-Rahjan catchment hydrological features and assess SCS-CN model applicability in surface runoff simulation in the study area. Materials and Methods Study Area The study's critical selected region was Wadi-Rahjan, a catchment area found in the Saudi Arabian western part [Figure 1]. The case study area is geographically found in 210 11’ 0’’ N to 210 20’ 0’’ latitude and 400 2’ 0’’ to 400 12’ 0’’ E longitude. the Wadi-Rahjan catchment areas is 157 square kilometers, 359 to 2021 meters elevation range above the mean sea level. Physiologically, Wadi-Rahjan catchment areas are divided into hills, barren areas, and urban masses. The regions have typically semi-arid climate conditions that encompass a long hot summer season with a short dry winter season, and then rainfall occurs. The area’s average yearly precipitation and temperature are indicated in Figures 2 and 3. Figure 1: Wadi-Rahjan catchment area in Saudi Arabia western region Wadi-Rahjan Figure 2: Yearly average temperature within the area of study; Wadi-Rahjan Figure 3: Wadi-Rahjan yearly average precipitation Data and Methodology

The research study's fundamental adopted methodology is illustrated in figure 4, which displays SCS-CN run-off estimation and collaboration with remote sensing (RS) and GIS. The study also involved derivation of critical elements such as digital elevation model (DEM), soil texture, and LULC. The variables were derived through freely assessable RS data and later analyzed using the ArcGIS software. United States Geological Data Survey (USGS) from the DEM data was used as illustrated in figure 5. It was crucial for delineating the Wadi-Rahjan catchment area with the aid of ArcGIS in figure 6. Also, figure 7 shows the created slope map. The study also prepared a LULC map using USGS website satellite imagery of 30 m resolution and the handsat ETM. The study also involved rainfall data collected climatic research unit (CRU) for a period extending from 1900 to 2019. The study’s soil data was collected from Saudi Arabia’s general soil map, environment, water, and agriculture ministry. Figure 4: Flowchart depiction of research study methodology adopted for runoff computation Figure 5: Wadi-Rahjan digital evaluation map Figure 6: Delineation of Wadi-Rahjan catchment map Figure 7: Slope map of Wadi-Rahjan Soil Map Soil texture is one of the study's major requirements. It involves the relative clay, silt, and sand particles proportions in a given soil area, which in our research involves Wadi- Rahjan [19]. The element is vital for analyzing and evaluating soil water aspects such as relation [19] and the soils’ hydraulic features [20]. Since the study fundamentally depends on SCS-S model soil data, the natural conversation service (NRCS) provides soil subcategories or subdivisions amounting to hydrologic soil groups; they include group A, B, C, and D. the soils are grouped depending on their water transmission rate, and infiltration capacity [21]

(Table 1). The HSG group A indicates low runoff potential and high permeability, while HSG group D indicates high runoff potential, high clay content, and or shallow clay content. The other HSG groups, B and C, form intermediate classes [21,22]. Table 1: The natural resource conservation service classification [21] Land Use/ Land Cover (LULC) Among the other vital elements that constitute the success of this study is LULC. The aspect is crucial thematic input in multiple studies since it is a land pattern and status utilization [23]. Satellite remote sensing is the primary LULC mapping tool since LULC is changed dynamic [23]. Watershed cover effects are other variables estimated using the land- use property to attain runoff or infiltration results, based on soil available elements such as rocks, bare terraced soil, and natural vegetation, among other vital factors [6]. The current study used a Landsat ETM+ satellite imagery encompassing 30 m resolution to generate spatial information in the study area’s land cover. Supervised clarification was also applied in the research study to implement ERDAS imagine software, which was then used to obtain LULC for the study area. Curve Number Method The critical establishment and development of the SCS-CN model by the United States Agriculture department was for its use in rural areas [3,24]. The model’s working mechanism is based on two essential variables; the curve number simplified runoff depth and the rainfall parameter [4]. The curve number is applicable in transforming the rainfall distribution frequency into runoff-based distribution frequency. The dimensionless CN number is defined as 1 ≤ CN ≤ 100 [3,4]. Dry antecedent soil moisture conditions (AMCI) are characterized by low CNs, while high CNs are indicated by wet conditions (AMCIII). Equations 1, 2, and 25 indicate conversions for CNs for normal conditions to the dry and wet conditions CNs associated with AMC-I and AMC-III.

4.2  CNII − 0.058  CNII ( )CNI= (1) 10 (2) 23 × CNII + 0.13  CNII ( )CNIII= 10 From the equations: CN1 indicates dry conditions curve numbers while CNII normal conditions are applied for curve numbers. CNIII means wet conditions used for curve numbers. Table 2 shows antecedent soil moisture range conditions. Composite CN (CNC) is applied for catchment areas that entail several lands uses and soil types. It is applied to estimate the direct runoff using the equation below [25]: CNc n CNi  Ai (3) = A i=1 From the equation: CNI is the sub-region CN value; AI represents the sub-region area, and A means the study catchment area. Table 2: Antecedent Soil Moisture Conditions (AMC) classification and corresponding CN Direct Runoff Depth The multiple factors that affect runoff generation are entailed in the SCS-CN USDA developed model, incorporating them into a singular CN parameter [26]. CN is a representation of potential runoff about rainfall, LULC, and HSG. Below is runoff equation four based on the SCS-CN model.

 (P − S )2 P  S  P  S Q =  P + (1−  ) S (4)   0 From equation 4: Q signifies direct runoff depth in mm, P characterizes rainfall depth in mm, S denotes the after runoff potential maximum retention in mm, while λ denotes dimensionless surface abstraction. Multiple country studies have indicated and documented λ value to range between 0.1 and 0.3 [27], with Shadeed and Al Masri [4] postulating the value λ value to be 0.2. Shrestha’s [27] sensitivity analysis shows found λ value to be 0.2. Therefore, λ substitution in equation 4 becomes: Q = ( P − 0.2S )2 (5) P + 0.8S Potential maximum retention S, which is calculated on runoff onset, is derived from the equation below: S = 25400 − 254 (6) CN Results and Discussion The LULC Map The research study involved four classifications of the LULC map, including rocks, bare soil, urban, and agricultures illustrated in figure 8. The figure shows how the catchment areas are occupied by rocky regions of most parts, followed by bare soils, urban masses, and the agricultural areas. Two critical elements dominate the study area; rock and bare soils. They cover 98 square kilometers, representing 62 percent, and 55 square kilometers, which represent 25 percent in that order. The urban masses have an area extent of 2.5 square kilometers, approximately 2 percent, as it is located at the study area’s central portion. The

agricultural area represents 1.5 square kilometers which are estimated to 1 percent. The overall results indicate that Wadi-Rahjan has a low vegetation cover. Zhao et al. [28] argue a strong correlation between runoff and vegetation cover whereby a high runoff was detected resulting from vegetation removal and urbanization. The researcher’s trend also similar when studying and evaluating low vegetation cover and its effect on rainfall and runoff. Figure 8: LULC map for Wadi-Rahjan catchment Soil Map Three soil texture types are present in the Wadi-Rahjan catchment area and comprise sandy clay loam, loam, and clay loam, as illustrated in figure 9. Wadi-Rahjan catchment study area is featured under HSGs B, C, and D based on infiltration and soil classification rates. Thirty- five percent of the region is represented by HSG-D and a large area covering 56 square kilometers. The analysis shows that the part is favorable for water retention and runoff generation. Twenty-five percent of the region represents HSG-C, a class characteristic of clay and loamy soils mixture, while 40 percent represents HSG-B, which is highly loamy textured. Therefore, the study area's soil composition and distribution are one critical factor contributing to high runoff amounts. Figure 9: Wadi-Rahjan soil textural class-map Figure 10: Wadi-Rahjan hydrologic soil group map CN Map CNII values related to AMCII for the study were obtained using cross-mapping between LULC ArcGIS software maps and HSGs, as illustrated in figure 11. Equations 1 and 2 were the vital requirements that the survey used to generate CNI and CNIII values associable with AMC-I and AMC-III, as illustrated in Figures 12 and 13. The highest CNII is represented by 90-94 class, which covers 93 square kilometers, 59 percent, while 74-80 type of the CN occupies one square kilometer of the total area having the lowest runoff potential. The 80-85

CN class, which represents the low, moderate runoff potential, covers 2 square kilometers while the reasonable runoff potential ranging between 85-90 occupies 61 square kilometers, 39 percent. The results are suggestive that Wadi-Rahjan catchment areas generate massive runoff. The high and increasing CN values indicate that the runoff would still increase in the study area. Equation 3 is a critical breakthrough for recalculating CN values by dividing the total CN product value and sub-catchment area summation. Accordingly, the value of weighted CNI, CNII, and CNIII were 80, 90, and 97, respectively. Figure 11: AMC-II map curve number (CNII) Figure 12: AMC-III map curve number (CNI) Figure 13: AMC-III map curve number (CN III) Potential Maximum Retention Map Both Equations 5 and 6 are critical in ArcGIS software for calculating potential maximum retention once the CN map is generated. Values of S ranged from 16 to 89 mm, as illustrated in figure 14. Built-up and rocky areas locate the lowest S values with low retention capacity. Inadequate retention capacity was also evident in the municipality’s bare areas. Figure 14: Wadi-Rahjan potential maximum retention (S) Runoff potential map Runoff computation involves basic parameters such as rainfall and CN. Figure 15 shows the runoff potential map generated from the GIS environment under the SCS-CN model’s implemented algorithms. 194 to 295 mm was the yearly range for the runoff. 289 to 295 mm, 35 percent, represented a higher percentage of runoff water. 194 to 268 mm is the runoff class that could have been harvested in Wadi-Rajan's western and southeastern parts. Choi and Ball's [29] findings resonate with the results from the study analysis. Their research is conclusive that soils with larger particles have a higher infiltration rate than soils with fine particles.

Figure 15: Runoff potential map of the study area Rainfall-runoff correlation analysis Figure 16 is a vital illustration of testing CN calculated values as it represents a graph of rainfall and runoff data pairs. Since coefficient R2 is 0.90, it relates to the rainfall runoffs. Research and findings by Peng and You [30] agree with the study area findings. Their conclusion shows that the SCS-CN model is a better simulation effective in Wadi-Rahjan as it has a co-efficient higher than 0.5 compared to those with coefficient runoffs less than 0.5. Figure 16: Rainfall-runoff correlation analysis for the study area Conclusion There is a missing conventional hydrological data research to support design and water systems operation purposes at the watershed levels. Therefore, a collaboration between past and current soil analysis models is the solution to estimating runoffs in study areas. The critical mechanisms and models that do not require pre-existing data include GIS SCS-CN based and RS models as they serve as essential runoff estimation techniques. Through embracing GIS and RS-based SCS-Cn model, the research study results offer recommendable Wadi-Rahjan runoff estimation in the Saudi Arabian western region. Further evaluation and classification of the Wadi-Rahjan catchment area classify the region into fundamental groups, HSGs B, C, and D groups. The final assessment shows that 9 is the standard normal condition CN. Therefore, the research found out that the wet and dry areas CN conditions for the study area were 97 and 80 in that order. The Saudi Arabian region runoff varied from 194 to 295 mm, based on SCS-CN obtained results and calculations. Therefore, the results show a higher percentage of runoff ranges between 289 and 295 mm as runoff water, representing 35 percent, followed by 269 to 288 mm, representing 24 percent. In summary, the unavailability of appropriate runoff estimation and measurement equipment in most of Saudi Arabia remote areas, which provides no past research or records on the region, offers experts in the region a

practical approach to embrace the techniques applied in Wadi-Rahjan in other Saudi Arabia watersheds to support watersheds, their development and planning purposes.