Assessing Hydrological Response in the Timah-Tasoh Reservoir Sub-Catchments: Calibration and Validation using the HEC-HMS Model

Hydrological modelling is a tool that is frequently used for assessing the hydrological response of a basin as a result of precipitation. It is also a vital component as water resources and environmental planning management. The study deals with calibrating and validating the hydrological response in the sub-catchments of the Timah-Tasoh reservoir using the hydrological model named Hydrologic Engineering Center – Hydrologic Modelling System (HEC-HMS). This study uses the SCS Curve Number, the SCS Unit Hydrograph, the constant monthly baseflow, and lag routing for the model development. The model was simulated for ten (10) years for calibration and nine (9) years for validation. The model calibration and validation efficiency were assessed using the coefficient of correlation (R). The findings show that the HEC-HMS model performs satisfactorily in simulating the observed daily inflow series, with the R-value of 0.4902-0.5139 during calibration and 0.5047-0.5559 during validation process. Thus, the result obtained from this study can be used as a preliminary development of hydrological modelling of the catchment of the Timah-Tasoh reservoir and can be used for extend application such as water inflow forecasting, impact of land use to the reservoir and others.


Introduction
Hydrology is the scientific study of water, including its flow on and beneath the Earth's surface and the effects of human activities on water supply and quality.Water assumes a vital role in human life, and the absence of this precious resource would lead to the incapacity of humans to endure, presenting an insurmountable challenge to human survival [1].Agriculture, generation of energy, manufacturing and recreational activities all require water.However, it is also dangerous to be without water while also having too much water, water resources management involves a systems approach that covers all hydrological components, including relations, effects and interactions.Changes in land cover, irrigation IOP Publishing doi:10.1088/1755-1315/1303/1/012029 2 and flow regulation are all examples of human-made environmental adjustments currently occurring on scales that have a major impact on the seasonal and yearly fluctuations in hydrologic patterns [2].
Generally, hydrology provides help for water resource planning and management, and control by applying engineering and geographical principles [3].Small-scale physical models can establish hydrological modelling, mathematical equivalents and computer simulation to represent hydrologic characteristics and system [4].Hydrological modelling focuses on individual flows within a system, with parameters that affect flow, soil land, climate and river characteristics.By applying the model, runoff data can be used to understand, control and measure the quality and quantity of water resources.In addition, due to advances in computing technology, hydrology models have recently become more detailed.There were three main reasons for developing hydrological models which are can better understand the impact of urbanisation on the natural water cycle to make up for the lack of reliable data that comes from the heterogeneous urban environment and to make future predictions such as flood forecasts and scenarios for land use and climate change [5].Furthermore, mathematical hydrological modelling is interesting because it enables a better understanding of reality and forecasting future water supply in different catchments [6].
With a variety of models available, it has recently been the most popular method for managing water resource and hydrological work.Besides, the model selection depends on the watershed and the purpose of the hydrological prediction in the catchment [7].Hydrologic Engineering Centers -Hydrologic Modelling Systems (HEC-HMS) is an open access hydrologic model developed by the United States Army Corps of Hydrologic Engineering Center.It is used globally for rainfall simulation and flood forecasting since it is an advanced modelling method.This model also processes rainfall loss, direct runoffs and routing [8].Therefore, due to calibration being the systematic process of modifying parameter values to its ease of use and common technique, HEC-HMS is commonly utilised in hydrological investigation [7].
Model calibration and validation play an important role.Calibration involves the process of modifying model parameter values for a particular situation in comparison with observable data for similar conditions gained through experimentation [9].The model parameters were tweaked to ensure the measurements yielded a satisfactory result.In contrast, validation is the process of evaluating the calibrated model's stability.The model was examined for various time steps, including daily, monthly and daily intervals [10].The model parameter obtained will be used to validate before it is proposed for use.To determine whether or not the model is suitable for hydrological simulations, it is calibrated and validated using statistical measurement techniques, the results of which reveal a correlation of between 0.4 and 0.7 are acceptable.
This paper aims to evaluate the performance of the HEC-HMS model for simulating the inflow of the Timah-Tasoh reservoir during the calibration and validation periods.In following sections, the data and materials are presented, followed by results, discussions and conclusion.

Description of study area
The study area focus of this study is the upper part of the sub-catchments in the Timah-Tasoh reservoir as shown in Figure 1.Timah-Tasoh reservoir is the largest reservoir in Perlis, Malaysia which lies approximately 13 km 2 north of Kangar town near the Thailand border.With an average surface of 13.33 km 2 and a storage capacity of 40 million m 3 , the reservoir has a storage capacity.Timah-Tasoh reservoir is used for flood mitigation in addition to its other functions, which include water supply, irrigation, and recreational activities.The Tasoh River and Pelarit River link and provide water to the reservoir.Approximately, 97 million m 3 of water per year flow into the reservoir from these two major rivers [11].Furthermore, the water supply comes from rainfall, flow in the catchment area and the Timah-Tasoh reservoir [12].The sub-catchments cover an area of 183.34 km 2 .

Hydrological data
At least nineteen years (2001-2019) of daily records of rainfall and discharge data from five rainfall stations and two discharge stations have been collected from the Department of Irrigation and Drainage (DID) Malaysia.The details and location of data are shown in Table 1 and Figure 1.

Development of HEC-HMS
HEC-HMS includes an integrated tool for simulating hydrologic processes of dendritic watershed systems [7].By using this modelling, it can learn more about the variability of hydrological processes.Each of the processes is represented by the HEC-HMS model, which computes runoff volume, direct runoff, baseflow and channel flow, and accounts for runoff loss from a specific watershed.The SCS-Curve Number had been applied for loss method, while SCS Unit Hydrograph, Constant Monthly and Lag had been used for transform, baseflow and routing method, respectively.Figure 2 presents the HEC-HMS model of a basin that contains 19 sub-catchments, 18 junctions and 14 reaches.

Calibration and Validation
Calibration is the systematic process of modifying model parameter values for a particular situation compared to observable data for similar conditions gained through experimentation [9].For this study, calibration was carried out to achieve reliable estimations of the parameters utilized to simulate rainfallrunoff.The values of the parameters were adjusted via trial and error.In contrast, validation is the process of evaluating the calibrated model's stability.Every calibrated model should be tested before usage.The model parameter obtained will be used to validate it before it is proposed for use.This study calibrated and validated the HEC-HMS model for 2001 until 2010 and 2011 until 2019, respectively.In this study, the performance was evaluated using correlation coefficient (R).The model was calculated using Equation (1): where  and  are the observed and mean observed discharge, while  and  are the mean simulated discharge.When R-values are closer to 1, it indicates that the simulated model is more accurate.

Calibration and Validation of HEC-HMS
To determine the best value of parameters of J8 and J18, the calibration and validation processes were done for the continuous-process model.The HEC-HMS model was calibrated from 2001 to 2010 and validated from 2011 to 2019.Modifications were made to the parameters to bring the observed and simulated hydrographs almost balanced adequately with one another.The estimation of losses utilised the SCS-CN method, while transformation employed the SCS-UH method, and for baseflow and routing, constant monthly and lag routing methods were applied, respectively, within the continuous process.
Calibrations findings improved model performance.Validation was done to test the accuracy of the calibrated model parameters and to confirm the parameters chosen during the calibration procedure.Each sub-catchments of calibrated parameters were adjusted, which affected curve number (CN), initial abstraction (  ), impervious percentage (imp), lag time (  ), baseflow and lag for routing (  ).The parameters are presented in Tables 2 and 3.As shown in Figures 3 and 4, the output of the model simulation was compared to the observed discharge of J8 and J18.

Daily simulation
Rainfall and streamflow data from each day are used to create the model's daily flow.Figures 3-4 display the agreement between the initial, calibrated, validated and observed hydrographs.For the calibration J8 and J18, the difference between the simulated initial and calibrated was slightly high and the calibrated fit with observed discharge.In addition, most of the parameters were adjusted for the calibration.[13] stated   , CN and   were the suitable parameters for model calibration.The lag time (  ) was adjusted so that it would correspond with the SCS unit hydrograph.Several studies also performed similar parameters for the transform process were calibrated to find the optimum fit such as [14] [10], [12] and [13].The average baseflow for J8 was changed from 0.83 m 3 /s to 0.00 m 3 /s, while the baseflow for J18 was changed from 1.15 m 3 /s to 0.28 m 3 /s.Therefore, peak discharge for J18 is higher than for J8.Following the completion of the calibration process for J8 and J18 for the years 2001 to 2010, a technique similar to the one used for the calibration was utilised to validate J8 and J18 between 2011 to 2019.All of the variables were kept the same during the validation process.In addition, the parameter values listed in Tables 2 and 3 were used for model validation.As shown in Figure 3, the simulated hydrograph for J8 fits well with the observed hydrograph.However, for the hydrograph shown for validation, J18, the simulation hydrograph cannot capture the observed start from 2018 until 2019.It might be J18 is located at the Kg.Jarum between sub-catchment B10a and B10b.Furthermore, the subcatchments B5 and B9, which are located in industrial and agriculture sectors, have high CN values, hence, the discharge is correspondingly high.In addition, an increase in imperviousness has contributed to this situation.Thus, the simulation data cannot fit with the observed data.
Compared to the actual data, the calibrated parameters for J8 and J18 produce a peak flow consistent with the observed values.Table 4 summarises the results and shows the peak flow for J18 (111.0 m 3 /s) is higher than for J8 (73.0 m 3 /s).The peak flow increases as the impervious percentage increases [10].However, the peak flow from calibrated parameters did not generate a similar time of peak and volume because this model is run for the continuous-process model.Similarly, for validation of J18 (51.80 m 3 /s) is lower than J8 (62.30 m 3 /s) in Table 4.The parameters of J18 were validated with the different environmental conditions because the location of the sub-catchments at J18 was started to build up with residence, industrial and agriculture [12].
Table 4 shows the model performance of calibration and validation.Calibration results for daily and monthly observations show that the model performs satisfactorily, with the R-values of 0.4902-0.5139.The finding shows the R-value for each junction increased after calibrated parameters were adjusted.During validation, the specified R-value of 0.5047-0.5559indicates that the model is satisfactorily on a daily time series.Both calibration and validation show an increase which considers satisfactory where the R-value 0.4 -1.0 is acceptable [16].According to [16], categorising the model as acceptable occurs when the R-value lies between 0.4 to 0.69.Moreover, labelling it as good or even very good is applicable when the R-value surpasses 0.70.As indicate by researchers, a coefficient below 0.1 is commonly interpreted as suggesting a minor relationship, while a coefficient above 0.9 is generally viewed as indicative of a robust relationship.When it comes to interpreting values that lie in between, subjectivity plays a significant role.Depending on the specific guideline used, a correlation coefficient of 0.65 can be categorised as either a "good" or "moderate" correlation category.Therefore, asserting that a correlation coefficient of 0.39 represents a 'weak' association, while suggesting that 0.40 signifies a 'moderate' relationship, is a subjective matter [17].

Conclusion
In conclusion, the HEC-HMS model was used to calibrate and validate the hydrological model for the Timah-Tasoh reservoir for the continuous process.This study used the R-value to evaluate the model's performance.Simulating the hydrograph for J8 and J18 were calibrated by modifying parameters such as curve number, initial abstraction, impervious surface, lag time and baseflow.The calibration result shows a satisfactory R-value of 0.4902-0.5139,and the performance of the model is acceptable during the validation period with R-value of 0.5047-0.5559.Therefore, the HEC-HMS model can be applied to the Timah-Tasoh reservoir to generate a satisfactory hydrograph.

Figure 1 .
Figure 1.Location of study area.

Table 1 .
Details of rainfall and discharge stations in sub-catchment in the Timah-Tasoh reservoir.

Table 3 .
Lag times values for routing.

Table 4 .
Simulated and observed peak flow for the continuous-process for calibration and validation.