Effectivity of micro drainage system as part of macro drainage system: A case study of Kelapa Gading, North Jakarta

Flooding is one of the problems that Jakarta faces every year. To handle the issue, the government took action to build the East Flood Canal. The canal construction is expected to reduce flooding, especially in the east and north region of Jakarta. However, the construction did not make the area immediately free of flood risk. Previous study (Susanti, 2017) showed that even with the existence of the East Flood Canal, potential inundations may still happen, specifically in the Cakung Lama area. In 2017, there were still several inundation points in the Sunter area, especially in Kelapa Gading. This research aims to identify the cause of inundations in the area by evaluating channel capacity in the micro drainage system in Kelapa Gading using hydrological model HEC-RAS 4.1.0 and WinTR. The simulation results showed that several channels in the system cannot accommodate the inflow, hence causing inundation in some areas in the Kelapa Gading area. According to the simulation results, it can be concluded that the micro drainage system in Kelapa Gading is not effective to accommodate the inflow.


Introduction
Flooding occurs in Jakarta every year [1]. A large scale of floods occurs approximately once every five years in the Ciliwung River Basin [2]. To deal with the problem, one of the actions taken by the government was construction of East Flood Canal. The construction of the East Flood Canal is predicted to reduce the inundation area in Jakarta [3].
The development of the East Flood Canal does not directly make the service area into a flood-free area. The scale of flood impacts has significantly increased in the last few decades [4]. The result of the previous study on the macro drainage system in the service area of the East Flood Canal showed that the cause of the inundation in the area is not because of the system. The macro drainage system in the East Flood Canal is still able to accommodate its design peak discharge [5].
However, in recent years we can still find inundations in the service area of the East Flood Canal, such as in Sunter Basin, primarily in Kelapa Gading. To determine the cause of the inundation in the Sunter Basin, further study is needed. Continuing the previous research about the macro drainage System in the East Flood Canal, the purpose of this study is to determine whether the cause of the inundation that occurs was because of its micro drainage system by using HEC-RAS 4.1.0 and WinTR-20 to calculate its effectivity.

Research method
The selected study area is located in Kelapa Gading, North Jakarta. The data used in this study are mostly obtained from the previous study. Some data, such as channels crosssection, are obtained from field surveys. To achieve the geographical data of the study area, such as land cover, length, and area, a program with geographic information system is needed. ArcGIS 10.1, a software from ESRI, is used to find geographical data from the study area. ArcGIS is a commercial software of geographic information systems, and the spatial information and its characteristics can accurately and fully be output to users regarding texts or pictures according to specified needs [6].
The rainfall calculation is done based on rain gages that affect the study area. Rainfall data used were from two rain gages in Tanjung Priok and Kemayoran. The annual maximum daily rainfall was identified for 13 years (2005-2017). The relative weights for each gage were determined using Thiessen Polygon. The average areal precipitation will be used to design rainfall using Log Pearson type III and Gumbel distribution.
In this study, WinTR-20 ver 3.10 and HEC-RAS 4.1.0 are used. WinTR-20 ver 3.10 is a computer program to assist the engineer in the hydrologic evaluation of flood event for use in analysis of water resource projects. It is a physically-based event model which computes direct run-off resulting from any synthetic or natural rainstorm [7]. WinTR-20 ver 3.10 is an SCS-CN based program to simulate the rainfall run-off modeling that can represent the spatial condition of watershed properties. Since the spatial distribution has a significant effect on the hydrological responses [8], this model can describe the effect of spatial distribution of land use variability at Upper Ciliwung Watershed [9]. The outputs of WinTR-20 ver 3.10 are the peak flow for each sub-area, and this will be used as input on HEC-RAS 4.1.0. HEC-RAS 4.1.0 (Hydrologic Engineering Software-River Analysis System) is developed by Corps' Civil Works Hydrologic Engineering Research and Development Program. HEC-RAS 4.1.0 is a one-dimensional flow model which is capable of performing calculation of steady flow, unsteady flow, and movable boundary sediment and water quality analysis for a network of natural or constructed channels [10].
The result of the HEC-RAS 4.1.0 simulation will show the comparison between the water level based on the peak flow from WinTR-20 ver 3.10 with the existing channel profile. The result will show the area in which design flood discharge exceeds the existing channel capacity of the channel (overflow). Then, the effectivity of the micro drainage system is calculated. It can be said that the effectivity is 100% effective if the capacity of the channel exceeds its peak flow. If not, the effectivity can be calculated by Eq. 1: If the effectivity is less than 100%, then it can be said that the existing channel cannot accommodate the flood discharge and causes an inundation. From these results, it is hen seen whether the cause of the flooding area is caused by the micro drainage system.

Results and discussion
Simulations were conducted in two models, for Kodamar and Kelapa Gading area. Each area consists of several sub-areas. Kodamar consists of two sub-areas, SubK1 and SubK2. Kelapa Gading consists of eight sub-areas, Sub-area 1 to Sub-area 8 ( Fig. 1).  Based on Fig. 1, using Arc-GIS 10.1, the properties data of each sub-area can be acquired, as in Table 1. It is essential to determine the properties' data of each sub-area to determine the amount of surface run-off of each sub-area. The properties of each sub-area will be used as the input data of WinTR-20 to produce the peak discharge for the selected return period. The relative weight for each rain gage is determined using Thiessen Method. The result of this calculation can be seen in Table 2. We can see that 83.46% of the study area is affected by the rain gage in Kemayoran and 17.54% by rain gage in Tanjung Priok. The average areal precipitation for the Kelapa Gading area can be seen in Table 3. From the data above, the design rainfall with 2, 5, 10 and 25 year return period can be calculated using the Log Pearson Type III distribution and Gumbel distribution. Then, the Goodness of Fit test is conducted using the Chi-Square Test and Smirnov-Kolmogorov Test. The purpose of the test is to determine which distribution will be used in the next calculation. The result of the calculation can be seen in Table 4.  The results of both tests show that the calculation of both distributions is acceptable. But, for the next step, the Log Pearson Type III distribution will be used because its ∆Max is smaller than the Gumbel Method. The design rainfall will be used as the inputs for the WinTR-20 along with the properties of each sub-area. The output from WinTR-20 simulation will be a peak flow from each return period. It can be seen in Table 5.
The peak flow will be used as the inputs for HEC-RAS 4.1.0 simulation. The result from HEC-RAS 4.1.0 simulation will show which channels are still able to accommodate the design peak flow and which channels are not. Fig 2. The blue line represents the water level from the five-year return period peak flow, the red one is the 10 year return period, and the green one represents the 25 year return period peak flow. The results obtained from the simulation are then analyzed to determine whether the result is following what is happening in the real event. The result can determine the factor of the occurrence of the flood as to whether it caused by the micro drainage system or not. The next thing to do is to calculate the effectiveness of the channel by comparing the existing capacity of the micro drainage system with the peak flow of each return period.
A micro drainage channel should be able to accommodate its 10 year return period peak flow. From Table 6, it can be seen that six channels cannot accommodate its peak flow. From that, it can be concluded that the inundation in Kelapa Gading Area caused by the micro drainage channel cannot accommodate the discharge.  Table 5. Peak time simulation result for each sub-area and reaches.