The influence of aquatic buffer characteristic of Ciliwung River from the Katulampa Weir to Manggarai Water Gate on hydrograph characteristics of design flood

Increasing population in Jakarta leads to increased needs of settlements and their supporting facilities. To compensate, various methods are being undertaken, from utilization of open area to the riparian, especially in the Ciliwung River. This utilization results in the loss of riparian natural ability, which changes the river’s hydraulic condition. These riparian changes will significantly affect the roughness coefficient. The roughness coefficient is a value that is affected by channel irregularity, variation of cross-section, effects of obstructions, vegetation, and degree of meandering. Changes of the roughness coefficient will influence the velocity of the stream. The purpose of this research is to determine the effect of riparian characteristics to the roughness coefficient that affects the stream velocity, which has impact on the peak time of the hydrograph. The peak time shift can be seen in the flood hydrograph characteristics of the Ciliwung River from the Katulampa Weir to the Manggarai sluice gate. Identification of Ciliwung riparian conditions at that segment is conducted by river routing. HEC-RAS 4.1.0 [1] application is used to obtain a design flood hydrograph at Manggarai Water Gate. To know the influence of riparian characteristics to roughness coefficient, HEC-RAS 4.1.0 simulation is done using existing and natural condition to compare the hydrographs between those two conditions. It obtains a lower peak discharge and extended duration of the hydrograph.


Introduction
The urbanization rate in Indonesia is regarded as high, at 53.3% in 2015 [1], after which the population in Jakarta increased to 10,192,886 people [2]. The increase of population in Jakarta causes increased settlements and required complementary facilities. This creates changes in land cover from open space to construct. Another aspect affected by population growth is uncontrolled constructions. An example of uncontrolled construction is the conversion of river banks for settlements, especially at the Ciliwung River. Riverbanks utilization causes a deprivation of its natural aspects, consequently changing its hydrologic conditions. These changes will influence the flow rates and change the flood peak times.
In this research, the characteristic mapping of the river border from Katulampa Weir to Manggarai Water Gate was performed to understand the changes in peak flood time and discharge due to the change in the river bank characteristics.

Research methodology
The research was carried out at Ciliwung River, specifically the Katulampa Weir-Manggarai Water Gate segment. Data used in this research are river lengths, cross-sections, streamflow discharges, and embankment conditions. Cross-sections data were obtained from "Studi Penataan Ulang Sempadan Sungai Ciliwung Tahap II" image data [3]. Stream flows were obtained from hydrographs corresponding to the event which occurred on February 5 th , 2018. The chosen date was attributed to the big flood which occurred in Jakarta. The flow discharge from the flood is required to verify whether the Ciliwung River can withstand the flow discharge, preventing flooding from occurring in Jakarta. River border conditions data were obtained from direct observations using field measurements.
Roughness was determined through conversion of bank conditions to Manning roughness coefficient [4] using the formula: All data were then applied to HEC-RAS 4.1.0 for simulation. The results were presented with a downstream flood hydrograph. Two flood hydrographs, with existing and natural river border condition, were obtained in the research. From the two results, the differences in peak time and peak discharge were observed. Thus, the magnitude of influence from river bank conditions was obtained.

Results and discussion
This research was done using two river boundary conditions. The first was the existing condition, while the second corresponds to the natural condition. It was assumed that the stream entering the channel only from Katulampa Weir is open. For existing condition, data were obtained from field measurements at Ciliwung River. Sampling locations are presented in Fig. 1.

Ciliwung River Segment in Bogor
Ciliwung River Segment in Bogor Di i From the figure above, the field measurements area is shown in yellow pins. There are 36 points reviewed for field measurements. These points are scattered in Bogor City, Bogor Regency Depok City, and Jakarta City with the following coordinates in Table 1. The observed conditions for each location are soil conditions, slope, and its vegetation height. Determination of above values is based on the manning value in a book "Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains" [4].
The correction values used are in Table 2.
Riverbank conditions from each sampling location were converted as Manning coefficients using formula (1) in Table 3. The observed conditions from each location are soil conditions, slope, and its vegetation height. Stream flow discharge data used for the simulation were obtained from previous research that indicated the maximum peak value at February 5, 2018 are presented in Table  4. On February 5, 2018, there was a high intensity of rain in Bogor City and Jakarta City, causing high floading in the Jakarta City. The election on February 5, 2018, aims to find out how much the flow changes in these extreme conditions. Data were simulated using HEC-RAS 4.1.0. The simulation was done under two possible scenarios, which are natural and existing conditions. The results from both simulations are presented with a downstream flood hydrograph.
The following is the simulation results obtained using HEC-RAS 4.1.0 (Table 5).  Table 5 and Fig. 2 are the simulation results of HEC-RAS 4.1.0. In the above results, there are changes in the hydrograph between natural and existing condition are observed. The change takes place in peak flood time, which takes one hour longer in natural condition compared with existing condition. The shift in peak discharge was also obtained in which the degradation occurred from 39.36m 3 /s in the existing condition to 38.71m 3 /s in natural condition.

Conclusions
Based on HEC-RAS 4.1.0 simulation results, it can be concluded that the condition of the river banks can affect the streamflow of the river. The natural condition of the river reduces the discharge value of the flood. The river bank changes also prolong the peak flood time. This can be used as an early warning for incoming floods whenever high flows occur.