Hydrological model of Ciawi dry dam using GPM satellite data

Ciawi Dam is the first dry dam in Indonesia. The benefits of the dam are to reduce flooding in the capital city of Jakarta, especially at a 50-year return period. The operation of the dam using 1 conduit unit aims to cut the flood peak. The flood discharge analysis that has been used so far is very limited due to the lack of existing rainfall data. So that calculations are carried out using corrected satellite data. In addition, flood discharge is also calibrated to the discharge at the Katulampa Weir. Analysis of the design rainfall shows the value of design rainfall with 50-year return (R50) is 210.59 mm and the 50-year flood discharge (Q50) is 317.13 m3/dt. The flood reduction value at the 50-year return period is 27%. If an alternative water gate is added, the flood reduction value can increase up to 72% with an opening of 30%.


Introduction
Flooding is a common problem in parts of Indonesia, especially in densely populated areas, such as urban areas.The losses that can be caused are significant, both in terms of material and human losses, so it is appropriate that the problem of flooding needs serious attention [1].One way to control floods structurally is to build dams.Dams are defined as buildings in the form of landfill, rockfill, concrete, and stone masonry that are built other than to hold and preserve the water [2].Some of the dam's functions include irrigation, hydroelectric power (PLTA), flood control, recreational facilities, fisheries, and others [3].
Ciawi Dam, or Bendungan Ciawi, is an earth fill dam built with the concept of a dry dam.It is the first dry dam in Indonesia, which primarily functions as a detention dam or flood control [4].Retaining dams are built to slow down water flow and prevent significant flooding [5].Based on a previous study by JICA in 2013, the construction of the Ciawi Dam was designed with the return period discharge of 50 years,it will reduce the flooding in the Ciliwung River that flows to Jakarta.However, because this dam does not have a water gate, the construction impact is insignificant against minor flood discharge.
The floodgates regulate and maintain river water discharge so that it does not overflow and cause flooding, especially downstream of the dam [6].Floodgates in urban areas play an important role because they are one of the instruments for controlling river water discharge and are an element of flood control [7].Based on this background, in this study, the authors will analyze the necessity of the construction of a floodgate within the Ciawi Dam so that the function of the dam is optimized and is expected to help prevent flooding, help manage river water discharge so that the river water level can be maintained.

Study Area
Ciawi Dam is located in Cipayung Village, Megamendung District, Bogor Regency, West Java.Geographically, the coordinates of the Ciawi Dam are located at coordinates 106° 52' 46" E and 6° 39' 34" S. By River Basin, it is located in the Ciliwung Cisadane River Basin in the Ciliwung watershed.The layout of the Ciawi Dam is shown in Figure 2, where the spillway is located on the left side of the main dam while the conduit is on the right side.The longitudinal section of the conduit is shown in Ciawi Dry Dam Figure 3, with an upstream base elevation of +504.20 and a downstream base of +478.20. Figure 4 displays the capacity curve of the Ciawi Dam with a maximum storage of 6.5 million m 3 .Some documentation of the built condition is shown in Figure 5. Ciawi Dam was built for flood control purposes with a 50-year return period.The flood control is carried out by passing water through a conduit with a square shape of 4.2 m x 4.2 m in size.There are two conduits, but only 1 unit is used during operation.The other unit is used during maintenance.

Data Collection
The rainfall data used consists of ground station data and satellite data.Data from the rain station is used to calibrate satellite data.Furthermore, satellite data is used as the basis for analysis.Table 1 shows the rain data used.Ciawi Dam is located on the upstream of Katulampa Weir with a distance of ± 7 km.The calculation of flood discharge at Ciawi Dam will be calibrated at Katulampa Weir.Therefore, to facilitate the calculation, the hydrological study is calculated up to Katulampa Watershed.Satellite data collection for the Katulampa watershed uses a grid system.The delineation results show that the GPM zone of the Katulampa watershed consists of 3 grids.Figure 6 shows the grid division for the location.Discharge data is required as a form of calibration.Discharge data is obtained from water level readings converted using the discharge curve equation.The discharge data used is the discharge data at the Katulampa Weir.Some of the discharge data used are shown in Table 2.For data analysis using Hec HMS software, some spatial data is required.The tool requires spatial data input, including sub-basin maps, slope maps, land use maps, soil type maps, and HSG maps. Figure 6 shows the spatial maps.Figure 6 displays some of the data used for the basis of the hydrological simulation.Figure 6a shows that the Sub Basin formed up to Katulampa Weir is 35.Figures 6b and 6c show that the slope is very steep with the dominance of plantation land use.Figures 6d and 6e show that the dominant soil type is Humic Andosol with Hydrologic Soil Group B.  instantaneous.In 2014, The Global Precipitation Satellite (GPM) was successfully launched.GPM is a national satellite rainfall observation project initiated by NASA and the Japan Aerospace Exploration Agency (JAXA), whose main objective is to provide high spatial-temporal resolution and precise rainfall data for global rainfall monitoring through satellite remote sensing technology.The GPM satellite alliance includes a central observation satellite (GPM Core Observatory) and several other cooperative satellites [8].In this study, the satellite rain data is GPM (Global Precipitation Mission) satellite rain data.The choice to use this data is based on previous studies that prove the accuracy of this satellite [9] [10].The length of data download is adjusted to the length of download of observational rainfall data between 2001 and 2021.GPM data has a spatial resolution in a 0.1°×0.1°grid, available for three 3hour daily data periods.GPM data is available for 3 hours, a daily data period, five days, and monthly.These data are available on the https://gpm.nasa.gov/data/imergpage.However, it can be downloaded at https://giovanni.gsfc.nasa.gov/giovanni/ to obtain area data by searching the domain.

Rain Data Correction.
In general, satellite rainfall data does not match the observed/measured rainfall data, so a correction is required.These differences may be caused by these errors, particularly those due to sensors [11] recovery algorithms [12], etc. Deviation correction is one step that aims to adjust the satellite rainfall values to the observed rainfall values.With the observed rainfall values.This study used the bias correction method to correct the rainfall data.for data with edits, generally, validation is done from the data [13].The method used in this step is the original mean square error (RMSE) [14] [9] with equations 1.
With : Yi = observation data (actual data) Ŷi = forecast data (satellite data) n = amount of data.

Design Rain.
The planned rainfall determined through frequency analysis was applied to 6 data sets: GRID 1, GRID 2, GRID 3, GRID 4, and Citeko.The return periods were selected as 2 years, 5 years, 10 years, 25 years, 50 years, 100 years, 1000 years, PMP, and Isohyet PMP.The selection of this long recurrence period aims to visualize the uniformity pattern of the design rainfall.Overall, the design rainfall in the four datasets showed high homogeneity.The calculation of design rainfall in this study using the GEV method.Generalized extreme values (GEV) are used to describe the temporal relationship between extreme rainfall intensities over different periods to describe quanta located at the ends of the distribution [15] [16] [17].

2.3.4.
Calibration.Model correction is the process of selecting combinations of parameters, or as well as the process of parameter optimization to improve consistency.Among the hydrological responses observed in this case are postal data from rain gauges, and simulated in this case are GPM satellite data [13].The rainfall correction analysis uses a y = f(x) regression equation.The relationship between satellite rainfall as variable x and observed rainfall as variable y results in the satellite rainfall correction equation [18].The magnitude of this relationship is often expressed quantitatively as a correlation coefficient, so it is very appropriate to use in this study because we want to know the relationship between GPM data and measuring data.If the correlation coefficient is high, it does not mean there is a similarity in the timing of hydrological phenomena (hydrological similarity).However, it indicates the simultaneousness of hydrological phenomena [19].

Rainfall Data Correction
GPM satellite rain data needs to be tested for correlation with existing ground stations and to analyze RMSE (Root Mean Square Error) for further calibration/correction of the GPM (corrected) data taken.
Based on the Technical Guidelines for Flood Discharge Calculation on dams [3], rain data obtained through the GPM method must first be tested for quality.The test is carried out by reviewing the parameters of suitability and compatibility with the data recorded at the rain post (in this section, monthly base data is used).From the table above, it is found that: x The rainfall stations used are Citeko, Gadog and Cilember because the data is sufficient (more than 10-year record).
x The correction factor applies to all grids because they are located in the same climate.
x Citeko Rain Station is the basis for correcting GPM daily rainfall because it has a high level of correlation and RMSE.
Based on the Technical Guidelines for Flood Discharge Calculation [3] as a guide for corrections made to daily rainfall data.The correction factor applies to all grids.The average error of daily rainfall data before correction is 0.0618, while the absolute error of rainfall data corrected using the above parameters is 0.0341, so the correction results are acceptable.A picture of the probability curve can be seen in Figure 7.

Design Rainfall
Rainfall is the most considerable annual rainfall with a certain probability of occurring in an area or rainfall with a specific return period probability.The method to calculate the amount of design rainfall

Calibration
Flood calibration was carried out using flood hydrograph data for 3 events, namely those that occurred in February 2007 (before the dam), January 2020, and October 2022.The calibration results are shown in Figure 9 and Table 5. with satisfactory results.Table 5 shows the results of flood discharge calibration at Katulampa Weir for each year.The initial abstraction value, ratio, initial discharge, recession constant, ratio to peak are values that are changed so that the flood discharge value is close to the observation value.The calibration results show the value of NSE very good, RMSE good while percent bias very good.

Flood Discharge
The calibration results showed acceptable values.Flood discharge calculations were then carried out based on the previously developed model.Figure 10 displays the flood discharge hydrographs for various return times.

Flood Routing
Flood tracing is a procedure to estimate the timing and hydrograph of a flood at a point in a river based on a known inflow hydrograph in the river upstream.In this case, flood routing is carried out in the reservoir through the spillway and conduit.The results of flood routing are shown in Table 6, with the largest reduction value at the 1000-year return period of 42%.  6.To increase flood reduction, there is an alternative to adding controllable gates at the upstream of the conduit.So that gate-opening operations can be carried out.Table 7 shows the results of flood routing with floodgates for Q100 and Table 8 for Q50.Both tables show that flood reduction is more significant if there is a gate operation, so the alternative of adding gates to the Ciawi Dam can be considered.

Conclusion
Based on the description above, several conclusions can be drawn as follows: 1.The value of the Ciawi Dam flood discharge after calibration at Q50, Q100 is 278.37 m 3 /s, 317.13 m 3 /s and 354.88 m 3 /s respectively.2. The flood reduction values at the two return times above are 27% and 32%, respectively.3. When using flood control gate, the flood reduction value is increase than the existing condition.
With a 30% opening at Q50 resulting a 72% flood reduction, and with 40% opening at Q100 resulting a 67% flood reduction.

Figure 1 .
Figure 1.Location of Ciawi Dam Within Watershed Area Ciawi Dam has a catchment area of 88.5 km 2 , with a main river length of 17.27 km.Some technical data is shown as follows: x Dam Type : Zonal Random Earth fill dam x Crest main dam : +551,00 x Conduit lower level : +504.20 x Riverbed : +500.00 x Height from Riverbed : 51.00 m x Height from Foundation Base : 55.00 m x Peak Width : 9.00 m x Peak Length : 334.50 m x Spillway Type : Side spillway x Spillway Crest : +546.75 x Spillway Threshold Width : 62.0 m x Volume at El. 546.75 : 4.79 x 106 m 3 x Inundation area of El. 546.75 : 34.30Ha x Dodging channel type : Conduit (rectangular closed channel) x Dimension : 2 x (4.2 x 4.2) m x Length : 398 m

Figure 5 .
Figure 5. Map of Hydrological Post and GPM Grid of Katulampa Watershed

Figure 6 .
Figure 6.(a) Sub Basin Map (b) Slope Map (c) Land Use Map (d) Soil Type Map (e) Hydrologic Soil Group Map

2. 3
.1.GPM Satellite Data.The application of satellite remote sensing can provide rainfall information that is more extensive, continuous, and almost instantaneous, more extensive, continuous, and almost

Figure 7 .
Figure 7. Annual Maximum Daily Rainfall Curve

Table 1 .
List of Rainfall Station

Table 2 .
Calibration Water Level List

Table 3 .
Table displays the correlation test and RMSE results of the rain post and GPM grid.[20] Correlation and RMSE of Rain Post with GPM Grid

Table 5 .
Comparison of Calibration Results

Table 6 .
Flood Routing Results

Return Period Flood (m3/sec) Flood (m3/sec)
Sluice Gate OperationAt Ciawi Dam, flood operations only rely on 1 conduit, with the effectiveness of flood reduction shown in Table

Table 7 .
Flood Routing of Q100 Results with Alternative Gates

Table 8 .
Flood Routing of Q50 Results with Alternative Gates