Prediction of Peak Discharge Using the SCS Curve Number Method in the Manikin Watershed

Surface runoff is a crucial hydrological variable in the analysis of water infrastructure planning. A reliable method for predicting surface runoff resulting from rainfall in an ungauged watershed is the SCS CN method. This research aims to represent the effectiveness of the Curve Number (CN) method in calculating peak discharge in the Manikin Watershed. The data used for this analysis includes rainfall data from three rain stations, each with a 25-year dataset, water level data of 11 years, digital elevation model (DEM) data, land use maps, and hydrogeological maps. The SCS curve number method is the most commonly used method for the estimation of peak discharge in a watershed. The calculated flood discharge values for the Manikin River Basin, with return periods of 5, 10, 20, 25, 50, 100, 500, and 1000 years, are as follows: 60.32 m3/s, 84.74 m3/s, 111.98 m3/s, 121.41 m3/s, 153.25 m3/s, 188.79 m3/s, 287.24 m3/s, and 337.30 m3/s, respectively. The reliability testing of the method in the Manikin Watershed was determined by a Nash–Sutcliffe efficiencies (NSE) value of 0.93 and root mean square error (RMSE) value of 19.70. As a result, the Curve Number Method proves to be highly reliable in representing the peak discharge in the Manikin Watershed.


Introduction
Kupang Regency is one of the administrative regions in East Nusa Tenggara Province (NTT), Indonesia.Astronomically, Kupang Regency is located at coordinates 9°15'11.78" to 10°22'14.25"S and 123°16'10.66"to 124°13'42.15"E, with a total area of approximately 5,298.13 km 2 .Kupang Regency experiences only two seasons: the dry season and the rainy season.Generally, the dry season occurs from April to November, while the rainy season extends from December to March [1].
In Kupang Regency, several major rivers are found, with the Manikin River being one of them.The Manikin River comprises a network of river tributaries situated within a Watershed, which functions to channel water into the main river.This river has become a priority for the government's attention in addressing water resource issues, including the construction of the Manikin Dam to enhance water resources.
An accurate prediction of peak discharge in watersheds is critical not only for water resource management but also for understanding the complex relationships of hydrological processes.Surface runoff is one of the crucial hydrological variables that play a significant role in enhancing water resources.In traditional models, the calculation of runoff into the river necessitates the consideration of IOP Publishing doi:10.1088/1755-1315/1343/1/012007 2 both hydrological and meteorological data.Acquiring this data is time-consuming, expensive, and a complex process.Remote sensing technology can significantly contribute to the development of models that traditionally test the relationship between rainfall and runoff.The primary role of remote sensing technology in runoff calculations is typically to generate input data and assist in calculating model coefficients and parameters.Experience has shown that various thematic information sources can be obtained through the interpretation of satellite imagery, including data related to land cover, soil, vegetation, and drainage patterns [2].By combining the results of satellite image interpretation with traditional measurements, such as rainfall, temperature, and topography, new data for rainfall-runoff modeling is generated.These overlapping data sets can be processed using Geographic Information Systems (GIS) [3].
Hydrological models that describe the relationship between rainfall and runoff are generally divided into two approaches: lump parameter and physical spatial distribution.One of the most commonly used and widely applicable methods is the SCS curve number method used in the estimation of watershed flood hydrograph ordinates.Easy usage and the availability of the model's inputs and numerous outputs such as peak discharge of flood, time to peak, lag time, and flood time make the SCS method more applicable [4].Abundant studies are devoted to the SCS-CN method due to its popularity and simplicity in rainfall-runoff modeling.The relationship between SCS-CN value and the rainfall depth (P) using rainfall-runoff data from experimental watersheds in the United States [5,6,7].
Research in the Manikin Watershed has also been carried out using different methods, including flood discharge analysis using the Nakayasu Unit Hydrograph (UH), ITB-1 UH, and Limantara UH methods in the Manikin Watershed [8].Other research has carried out the analysis of maximum discharge in the Manikin Watershed using the rational method and Nakayasu Synthetic Unit Hydrograph [9].
This paper aims to analyze flood discharge using the Soil Conservation Services -Curve Number (SCS-CN) UH method in the Manikin watershed.The current study focused on the hydrograph as a whole through the application of flood routing and considering the hydrograph's significant parameters (peak flow and runoff volume).

Materials and Methods
The Manikin Watershed is located in Kupang Regency, East Nusa Tenggara Province.The outlet point of the Manikin Watershed for this study is located in Tarus Village at coordinates 10° 08' 29.3" S and 123° 41' 22.9" E as shown in Figure 1.Based on these data, the analysis is carried out by first selecting the rainfall station.Selection of rainfall stations.The selected rainfall stations are those with 25 years of rainfall data, namely Baun Rain Station, Tilong Rain Station, and Penfui Rain Station.To fill in missing rainfall data through the normal ratio method [11] by using the following formula: IOP Publishing doi:10.1088/1755-1315/1343/1/012007 Where Px = predicted rainfall at station X; PA, PB = rainfall at stations A, B, ..; NA, NB = normal annual rainfall at stations A, B, …; Nx = normal annual rainfall at station X; n = number of reference stations.
Testing the consistency of rainfall data using the RAPS method.Consistency tests are conducted on annual rainfall data to identify any deviations, enabling a determination of whether the data is suitable for hydrological analysis.The consistency of rainfall data is represented by calculating the cumulative deviation value from the average.Calculation of area-averaged rainfall using the Thiessen polygon method [12]: Where R ̅ = average rainfall; R1, R2 = rainfall at stations 1, 2,..; A1, A2= Area of each sub-region (area at stations 1, 2); n = Number of reference stations.
To analysis of the rainfall return period was conducted using Hydrognomon 4.0 software.The rainfall return periods used in this test were 5, 10, 20, 25, 50, 100, 500, and 1000 years with the rainfall distribution methods used being the Log Pearson Type III and Gumbel methods.
Frequency distribution goodness-of-fit tests were conducted using the Chi-Square test and the Kolmogorov-Smirnov test.
1. Chi-Square Test Where χ = the test statistic; Ef = expected frequencies; Of = observed frequencies 2. Kolmogorov-Smirnov Test ∆maks = I Sn -Px I (4) Where Sn = empirical distribution function for n; Px = cumulative distribution function.The determination of watershed characteristics and the creation of land cover and hydrogeological maps were performed using ArcGIS 10.5 software.Determining the curve number values based on soil moisture conditions and the combination of land cover type data with hydrogeological data [13].
Where CNcomp = Composite CN; CN1, CN2 = Curve Number for soil type and land cover 1; ΣA= The amount of area combined soil type and land cover Flood discharge analysis using the HSS SCS-CN method.The CN method is based on the relationship between the infiltration of each soil type and the amount of rainfall that falls every time it rains.The CN values range between 1 and 100 which is a function of runoff resulting from soil types, land use, hydrological conditions, and antecedent moisture condition [14].The form of the equation is: Where Q = direct runoff (mm), P = rainfall depth/precipitation (mm), Ia = initial abstraction (Initial loss), and S = the water maximum retention potential by the soil, which is a large part is due to infiltration (mm).For a clearer understanding, the testing steps are illustrated in the flowchart depicted in Figure 4.
Where St = Simulation discharge; Ot = Observed discharge;  ̅  = Average of observed discharge 2. Root Mean Square Error (RMSE) Where X = observation value; Y = prediction value; n= amount of data

Results and Discussion
The rainfall data obtained from BWS Nusa Tenggara II consists of daily rainfall data.Daily rainfall data from Baun Rain Station, Tilong Rain Station, and Penfui Rain Station for the period of 1994-2018 (25 years) were summarized into annual maximum daily rainfall data.Missing rainfall data at the Tilong Rain station were filled in using the Normal Ratio Method.To ensure consistent data use and eliminate the influence of measurement location changes or other disturbances on data consistency, the data needed to be adjusted [16].Consistency tests were conducted to confirm that the obtained rainfall data were consistent and ready for further calculations [17].The consistency test for rainfall data used the RAPS Method.The results of the consistency test revealed that the Baun Rain Station, Tilong Rain Station, and Penfui Rain Station had consistent data and could be used for subsequent calculations.
The calculation of average rainfall used the Thiessen Polygon Method.The Thiessen Polygon Map is shown in Figure 5. Based on the Manikin Watershed's area of 115.49km², weights for the three rain stations representing the affected areas in the polygon were determined.The weights for the Baun Station, Tilong Station, and Penfui Station were 0.39, 0.30, and 0.31, respectively.The graph of average rainfall in the area using the Thiessen Polygon Method is shown in Figure 6.Then a frequency analysis was conducted to determine design rainfall for specific return periods.The selected return periods for this analysis were 5, 10, 20, 25, 50, 100, 500, and 1000 years.In the rainfall design analysis of the Manikin Watershed, both the Gumbel Type I and Log Pearson Type III methods were utilized.Rainfall design values obtained from both methods were then subjected to goodness-offit tests, which included the Chi-Square Test and the Smirnov-Kolmogorov Test.The results indicated that both Log Pearson Type III and Gumbel methods were accepted in both goodness-of-fit tests.However, the Log Pearson Type III method showed the smallest deviation.Therefore, for further analysis, the Log Pearson Type III design rainfall was selected.The design rainfall values using the Log Pearson Type III method are presented in Table 1.Based on the determination of watershed characteristics and the creation of land cover and hydrogeology maps using ArcGIS 10.5 software, the Curve Number (CN) value can be calculated using Equation 5.The CN value obtained for the Manikin Watershed is 71.91.Subsequently, flood discharge simulation calculations were conducted using the SCS-CN Unit Hydrograph method.The results of these calculations can be found in Table 2.The water level data in meters (H) obtained at the Manikin River gauge station is converted into discharge (Q) data using the flow rating curve equation [18,19]: The results of discharge calculations from water level data for 2010-2020 can be seen in Table 3. Table 3 shows the results of the consistency test conducted using the Rescaled Adjusted Partial Sums (RAPS) method, followed by a frequency analysis using the Log Pearson Type III method.The calculation results obtained from the Log Pearson Type III method are shown in Table 4.The simulation and observed discharge data can be compared to evaluate the correlation between the two datasets.The comparison of simulation discharge and observed discharge is depicted in Figure 7.   7 shows a strong correlation between the simulated discharge and observed discharge, as evidenced by the small data differences at each return period.Subsequently, the reliability of the method within the Manikin DAS was assessed through the Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE) tests.The Nash-Sutcliffe Efficiency (NSE) value obtained was 0.93, and the Root Mean Square Error (RMSE) value was 19.70.These results indicate that the reliability of the Curve Number method, as measured by Nash-Sutcliffe Efficiency (NSE), is close to 1 (one), and for Root Mean Square Error (RMSE), it is not close to 0 (zero).Therefore, the Curve Number method is sufficiently reliable in representing the relationship between rainfall and runoff in the Manikin Watershed.

Conclusion
The design rainfall in the Manikin Watershed obtained using the Log Pearson Type III Method for the return periods of

Figure 1 .
Figure 1.The map of Manikin watershed location.

Figure 2 .
Figure 2. The land use map of the Manikin watershed.

Figure 3 .
Figure 3.The hydrogeological map of the Manikin watershed.

Figure 4 .
Figure 4. Flowchart of the research Calibrating the simulated discharge values and observed discharge values using the Nash-Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE) tests [15].1. Nash-Sutcliffe Efficiency (NSE)

Figure 6 .
Figure 6.Average rainfall data in Manikin watershed

Figure 7 .
Figure 7.Comparison of simulation and observed discharge in Manikin watershed

Figure
Figure7shows a strong correlation between the simulated discharge and observed discharge, as evidenced by the small data differences at each return period.Subsequently, the reliability of the method within the Manikin DAS was assessed through the Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE) tests.The Nash-Sutcliffe Efficiency (NSE) value obtained was 0.93, and the Root Mean Square Error (RMSE) value was 19.70.These results indicate that the reliability of the Curve Number method, as measured by Nash-Sutcliffe Efficiency (NSE), is close to 1 (one), and for Root Mean Square

Table 1 .
The design of rainfall using the Log Pearson Type III method.

Table 2 .
Results of Flood Discharge Simulation Calculation.

Table 3 .
The results of discharge calculations from water level data.

Table 4 .
The Log Pearson Type III method for observed discharge.
5, 10, 20, 25, 50, 100, 500, and 1000 years are 99.98 mm, 120.28 mm, 141.53 mm, 148.66 mm, 171.89 mm, 196.97 mm, 263.85 mm, and 297.00 mm, respectively.The Curve Number (CN) value for the Manikin Watershed is 71.91.The flood discharge values obtained using the SCS-CN Unit Hydrograph Method for the return periods of 5, 10, 20, 25, 50, 100, 500, and 1000 years are 60.32 m 3 /s, 84.74 m 3 /s, 111.98 m 3 /s, 121.41 m 3 /s, 153.25 m 3 /s, 188.79 m 3 /s, 287.24 m 3 /s, and 337.30m 3 /s, respectively.The Nash-Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE) values for the comparison of simulated and observed discharge in the Manikin Watershed are 0.93 and 19.70.In the reliability test for the Manikin Watershed, it can be observed that the method's reliability values for Nash-Sutcliffe Efficiency (NSE) are close to 1, and for Root Mean Square Error (RMSE), none are close to 0. The Curve Number Method is sufficiently reliable in representing the relationship between rainfall and runoff in the Manikin Watershed due to the same pattern and small deviation between the simulation and observation graphs.