In this study, the impact of channel rehabilitation on the extent of floodplain map for Q500 and Q1000 discharges is investigated in the Ismetiye creek, located in Bursa, Turkey. The necessary data used is discharge and the elevation data for the main channel and the catchment area of Ismetiye creek. Both point elevation data and topographic contour lines are used to create the digital elevation model of the area in Arc Map. Subsequently, it is used to develop the floodplain model of catchment area for extreme discharges using HEC-RAS software for the two cases including before and after the rehabilitation of Ismetiye creek. Water depths and floodplain maps are compared for these two cases. In addition, flood hazard maps, flood velocity maps, and water surface elevation maps are developed and compared. Flow hydrographs in the main channel and overbanks are computed at specified stations along the river. It is observed that rehabilitation works contributed the flood mitigation by reducing overbank flows and flood extent considerably. Finally, some recommendations are proposed for future works.
Research Article
The Impact of River Rehabilitation on Flood Mitigation: A Case Study of Ismetiye Creek Bursa
https://doi.org/10.21203/rs.3.rs-1983417/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Floodplain map
HEC-RAS
Rehabilitation
Mitigation
1.1. General Introduction
Flood can be defined as water overflowing a usually dry land resulting in submerging of the land. Floods are the most frequent type of natural disaster. They are often caused by heavy rainfall, rapid snowmelt or a storm surge from a tropical cyclone or tsunami in coastal areas. Even though the loss to life due to flood events during the past century has declined, the economic losses due to floods have continued to rise. This is due to the increase in urbanization and coastal development (WHO 2022).
Turkey experienced rapid urbanization during the past years which is why Turkey is still unprepared for many natural disasters including floods.
In Turkey, floods are the second most destructive types of natural disasters. 30% of natural disaster constitute of floods. Flash floods caused by prolonged and severe rainfalls are commonly observed especially in coastal regions of Turkey. Flood impacts are felt more strongly in Turkey's largest cities, which suffer from regular floods as a consequence of the combined effects of extreme rainfall, residential and commercial usage of flood plain and poor drainage. Since 1970, a flash flood prevention program has been successful in lowering the number of flash floods in Turkey. However, significant structural and mitigation work are still required, particularly in metropolitan regions of the nation, to prevent flash floods and mitigate their effects (Climate Change Post, 2022).
Bursa, which is an industrial hub in the North-Western Turkey, is no exception. It is also affected by floods time to time. Bursa has experienced many floods during the past 3 decades. Recently in the month of June 2020, flash floods killed 5 people in Kestel, a district 30 Km’s from central Bursa. The flash flood was caused by heavy downpours, leaving the villages inundated by floodwaters. Before that, the most fatal floods seen in Bursa were 1989 floods in Setbasi Gokdere, 2001 floods again in Setbasi Gokdere, 2005 floods in Yildirim Hacivat and Delicay stream, and 2010 floods in Cilimboz stream (Bursa Haber 2010).
The population of Bursa increased from 148.000 in 1950 to 3 million in 2020 (Macrotrends, 2020). This caused increase in the demand for city’s water resources. This stress is further accompanied by global warming impacts. These impacts include high precipitation rates and rising sea water levels for low-lying coastal areas. As a result of such local variations in the hydrological cycle, there is an increased need for more accurate predictions of hydrologic models. Flooding is also a threat in the river bank areas of Bursa. Hence a reliable hydrologic model is required to predict the flood levels in light of a flood event.
1.2. Research Objectives
This research aims at:
-
Developing digital elevation model (DEM) of the study area in ArcMap for both natural and rehabilitated states of Ismetiye creek and using this model for developing a floodplain model of the catchment area for extreme discharges in HEC-RAS.
-
Comparing the results before and after the rehabilitation of Ismetiye creek and analysing the impact of rehabilitation work on flood mitigation.
-
Giving necessary suggestions based on the results of this study for further improvement of flood mitigation works.
1.3. Scope of Study
BUSKI has been carrying out rehabilitation works in urban areas in Bursa. In this research, the impact of rehabilitation of Ismetiye creek on the flood mitigation in Ismetiye catchment area in Bursa was studied through 2D flood modelling. Necessary data used in this study was elevation data and discharge, both collected from BUSKI. Using the elevation data, ArcMap was used to create DEM of the study area. This DEM was then exported to HEC-RAS to develop 2D floodplain models for Q500 and Q1000. These models were developed for both before and after creek rehabilitation. Eventually, the developed models were analysed and compared to study the impact of creek rehabilitation on flood mitigation.
2.1. Study Area
The catchment area is located in the 1/25000 scaled maps H22-A4 and H22-D1 of Marmara Region, Bursa. The area is located between 29°7ʹ0ʺE-40°17ʹ30ʺN and 29°4ʹ30ʺE-40°14ʹ30ʺN geographic coordinates. The area of the catchment around Ismetiye Creek is 52.44 Km2. Fig 1 shows a satellite view of the Demirtaş dam, Ismetiye creek and the Ismetiye region settlements. Ismetiye creek originates from Demirtaş dam and joins with Nilüfer river in the south. Basin area is generally covered with forests and agricultural lands but there also are some residential areas in the south. Demirtaş dam, which was put into operation in 1983, is located at the upstream of Ismetiye creek. The water source of Demirtaş dam is Ballık stream. Ballık stream arises at an elevation of 1000 m, flows westward and reaches Demirtaş dam at an elevation of 170 m. Demirtaş dam spillway outflow and Ismetiye creek intermediate basin flow cause potential floods for which the channel was rehabilitated. The length of Ismetiye creek (shown in blue) is 8140 m and the length of rehabilitated channel (shown in green) is 4020 m. Slope of Ismetiye channel in its natural state was 0.006. As per the Rosgen Level Ι Stream Classification Technique (Rosgen, 1994), the channel has a gentle slope (<2%). Thus, it can be classified as ‘C Stream Type’.
Ismetiye creek, in its natural state before rehabilitation, was more prone to flood risk. Fig 2 shows Ismetiye creek in its natural state.
The rehabilitation work was started by BUSKI in the year 2013 and finished in 2017.
Fig 3 shows Ismetiye creek in its rehabilitated state as of March, 2022.
Fig 4 shows Ismetiye creek at the beginning and end points of rehabilitation.
2.2. Methodology
2.2.1. Flood discharge calculation
In the calculation of design flow through the channel, two methods namely Mockus Method and DSI Synthetic Method were used by BUSKI in their report, keeping in view the project criteria and technical specifications deemed right by the State Hydraulic Works Department (DSI).
The results obtained from both Mockus and DSI Synthetic method were compared in the BUSKI report (2013). However, the results obtained from DSI Synthetic method were accepted and used for the design of rehabilitation channel. Q500 and Q1000 values as obtained from the DSI Synthetic method for a 20-hour hydrograph can be seen in Table 1.
Table 1. Q500 and Q1000 in the Ismetiye creek for a 20-hour hydrograph (BUSKI Report, 2013)
|
DEMIRTAS DAM SPILLWAY FLOW |
ISMETIYE INTERMEDIATE BASIN FLOW |
AFTER SUPERPOSITION TOTAL FLOW |
||||
|
T |
Q500 |
Q1000 |
Q500 |
Q1000 |
Q500 |
Q1000 |
|
(Hour) |
(m3/s) |
(m3/s) |
(m3/s) |
(m3/s) |
(m3/s) |
(m3/s) |
|
0.0 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
|
0.5 |
2.94 |
3.52 |
6.77 |
7.82 |
9.72 |
11.34 |
|
1.0 |
11.15 |
13.33 |
24.97 |
28.83 |
36.12 |
42.16 |
|
1.5 |
23.72 |
28.36 |
51.00 |
58.89 |
74.72 |
87.25 |
|
2.0 |
41.14 |
49.19 |
69.00 |
79.68 |
110.14 |
128.87 |
|
2.5 |
62.52 |
74.64 |
66.12 |
76.35 |
128.63 |
150.99 |
|
3.0 |
83.58 |
99.52 |
54.01 |
62.37 |
137.59 |
161.89 |
|
3.5 |
100.66 |
119.48 |
39.27 |
45.35 |
139.93 |
164.83 |
|
4.0 |
114.51 |
135.39 |
28.35 |
32.74 |
142.87 |
168.13 |
|
4.5 |
122.90 |
144.61 |
20.83 |
24.05 |
143.73 |
168.66 |
|
5.0 |
128.25 |
150.19 |
15.06 |
17.39 |
143.31 |
167.58 |
|
5.5 |
129.67 |
151.19 |
10.70 |
12.35 |
140.36 |
163.54 |
|
6.0 |
128.25 |
148.90 |
7.57 |
8.74 |
135.82 |
157.64 |
|
6.5 |
125.75 |
145.48 |
5.54 |
6.39 |
131.29 |
151.87 |
|
7.0 |
122.75 |
141.54 |
4.20 |
4.85 |
126.94 |
146.39 |
|
7.5 |
118.83 |
136.65 |
2.97 |
3.43 |
121.80 |
140.09 |
|
8.0 |
114.13 |
130.94 |
2.15 |
2.48 |
116.27 |
133.41 |
|
8.5 |
109.59 |
125.43 |
1.58 |
1.83 |
111.17 |
127.26 |
|
9.0 |
105.81 |
120.84 |
1.13 |
1.30 |
106.94 |
122.15 |
|
10.5 |
94.71 |
107.57 |
0.43 |
0.50 |
95.15 |
108.07 |
|
11.0 |
92.22 |
104.58 |
0.28 |
0.32 |
92.50 |
104.90 |
|
11.5 |
89.36 |
101.20 |
0.00 |
0.00 |
89.36 |
101.20 |
|
12.0 |
86.42 |
97.77 |
0.00 |
0.00 |
86.42 |
97.77 |
Table 1. Q500 and Q1000 in the Ismetiye creek for a 20-hour hydrograph (continued)
|
DEMIRTAS DAM SPILLWAY FLOW |
ISMETIYE INTERMEDIATE BASIN FLOW |
AFTER SUPERPOSITION TOTAL FLOW |
||||
|
T |
Q500 |
Q1000 |
Q500 |
Q1000 |
Q500 |
Q1000 |
|
(Hour) |
m3/s |
m3/s |
m3/s |
m3/s |
m3/s |
m3/s |
|
12.5 |
83.61 |
94.49 |
0.00 |
0.00 |
83.61 |
94.49 |
|
13.0 |
81.04 |
91.48 |
0.00 |
0.00 |
81.04 |
91.48 |
|
13.5 |
78.13 |
88.11 |
0.00 |
0.00 |
78.13 |
88.11 |
|
14.0 |
75.24 |
84.79 |
0.00 |
0.00 |
75.24 |
84.79 |
|
14.5 |
72.37 |
81.49 |
0.00 |
0.00 |
72.37 |
81.49 |
|
15.0 |
69.61 |
78.33 |
0.00 |
0.00 |
69.61 |
78.33 |
|
15.5 |
66.65 |
74.95 |
0.00 |
0.00 |
66.65 |
74.95 |
2.2.2. Topographic data preparation in ArcGIS
In our study, ArcMap was used to create DEM of the study area. Two sources of data were used to create the DEM. The first source was point data extracted as shape files from AutoCAD files obtained from BUSKI report (2013). The projection system used to extract point data from CAD files was European Datum 50/Transverse Mercator 30 (ED 50/TM 30). The second source was the scanned topographic maps with contour lines, also obtained from BUSKI. The projection system used for importing the scanned maps to ArcMap was European Datum 1950 UTM Zone 36N. A DEM with a resolution of 0.5 m for Ismetiye creek and its catchment area was obtained in ArcMap. This DEM was then transferred as a terrain to RAS Mapper window of HEC-RAS for modelling of the creek. In RAS Mapper, it was made sure that the DEM imported from ArcMap was georeferenced. This was checked by adding the google satellite view on the background of RAS Mapper interface.
A. Modelling for natural channel
The CAD files were set to appropriate coordinate system mentioned above and then imported to ArcMap. The point layer containing elevation data was converted to point shapefile in ArcMap. There was scarce data available towards the downstream end of the channel. By using slope of the natural channel (0.006), further elevation data was added along the channel path until the end of Ismetiye creek. In addition to this, more elevation data was obtained from the second source consisting of scanned maps of study area. Points were drawn over contour lines of the map and elevation corresponding to the contour line was assigned to them.
Thus, elevation data along the complete path of Ismetiye creek was obtained. This enabled us to model the complete channel until its confluence with Nilufer river. Fig 6 (a) shows point elevation data of Ismetiye channel in its natural state after inserting additional elevation data.
Using Inverse Distance Weighted (IDW) raster interpolation method, a DEM was created from the point data as shown in Fig 6(b).
The cell size used for creating DEM in this study was 0.5 m. DEM was exported to RAS Mapper by adding it in the terrain layer of RAS Mapper.
B. Modelling for concrete channel
Intermediate sections were inserted for the concrete channel with a 20 m for a span of 4020 m. Slope of the concrete channel was 0.008484. For the remaining downstream part of channel, assuming the same slope of the channel throughout, intermediate sections were inserted at 20-m intervals by interpolation. This was done to obtain a DEM throughout the Ismetiye channel so as to run the simulation along the path of the entire channel.
- To obtain the DEM of the rehabilitated channel, the elevation data of natural channel was replaced by the elevation data of concrete channel. This was done by replacing the elevation data along the channel path by the intermediate sections. Fig 7 shows the point elevation data and DEM of Ismetiye channel in its rehabilitated state. Eventually, the DEM was exported to RAS Mapper by adding it in the terrain layer of RAS Mapper.
2.2.3. 2D-modelling in HEC-RAS
The terrain files of the study area obtained for both before and after the rehabilitation of Ismetiye creek were used in HEC-RAS 6.0 Beta Version to run 2D simulations. Discharge values with 500-year and 1000-year return periods, Q500 and Q1000, for a 20-hour hydrograph were used as inflow boundary conditions.
The HEC-RAS two-dimensional computational module has the option of running the following equation sets: 2D Diffusion Wave equations; Shallow Water Equations (SWE-ELM) with a Eulerian-Lagrangian approach to solving for advection; or a new Shallow Water Equation solver (SWE-EM) that uses a Eulerian approach for advection. The 2D Diffusion Wave computational method is the default solver. In general, many flood applications work fine with the 2D Diffusion Wave equations. The Diffusion Wave equation set runs faster and is inherently more stable. However, there are applications where the 2D SWE should be used for greater accuracy. Most 2D modeling situations, such as flood modelling, can be accurately modeled using this solver, where inertial forces tend to dominate frictional and other forces (HEC-RAS User’s Manual 2020).
The following guidelines as shown in Eq. (1) and Eq. (2) are used for picking a computation interval for the Diffusion Wave Equations:
Steps followed in HEC-RAS for 2D simulations:
- The projection was set.
- DEM was imported from ArcMap to RAS Mapper.
- Modification was done to the terrain by adding buildings, bridges, roadways and channels.
- Flow area was defined over the entire terrain.
- Mesh was created and mesh size was set as 10 m.
- Boundary lines were specified at the upstream and downstream ends of the channel and a third boundary line was introduced at the downstream end to ensure flowing out of water from the catchment area.
- Manning’s n value was assigned for both natural and concrete channels. For natural channel, it was taken as 0.05 and for concrete channel, it was taken as 0.012.
- The layer was then imported to HEC-RAS menu.
- Boundary conditions, flow hydrograph for upstream (given in Table 1) and friction slope for downstream were specified.
- Unsteady flow simulation for a 20-hour hydrograph was performed.
Please note that the selection of Δx (average cell size) and Δt (computation interval) is a balance between achieving good numerical accuracy while minimizing computational time. In our study, Δx = 10 metres and Δt = 5 seconds, which was calculated from Eq. (2), achieved favourable results.
In the simulations, discharge values with return periods of 500 and 1000 years (Q500 and Q1000) for a 20-hour hydrograph were used. The following 4 cases were modelled:
Case 1: Before rehabilitation of Ismetiye creek with Q500.
Case 2: Before rehabilitation of Ismetiye creek with Q1000.
Case 3: After rehabilitation of Ismetiye creek with Q500.
Case 4: After rehabilitation of Ismetiye creek with Q1000.
3.1. Results of 2D Simulation Before Rehabilitation of Ismetiye Creek
There was a significant flooding throughout the catchment area of Ismetiye creek as shown in Fig 9. In the case of Q500, the time of maximum extent is 7 hours 12 minutes and the maximum depth is 5.0 m, also shown in Fig 9.
Before rehabilitation, the extent of flooding was more at the beginning of the channel. As we go downstream, there was overbank flow observed which resulted in the flooding of catchment area. The overbank flow also flooded the dry channel present in the south of the Ismetiye creek. Moreover, new settlement region located on the right overbank side of downstream end of the channel also observed flooding.
3.2. Results of 2D Simulation After Rehabilitation of Ismetiye Creek
In the case of simulation after rehabilitation of Ismetiye Creek, it was found that flooding decreased significantly as compared to the case of natural channel. Fig 10. shows the flood maps after rehabilitation for Q500 and Q1000 respectively. For Q500, the time of maximum extent is 7 hour 52 minutes and the maximum depth is 5.544 m. For Q1000, the time of maximum extent is 7 hour 52 minutes and the maximum depth is 5.794 m.
After rehabilitation, there was more flow inside the main channel and less on the overbanks. The new settlements in the downstream were not inundated at all. Also, the channel in south of Ismetiye creek was not flooded at all.
As expected, the extent of flooding observed for Q1000 was more than that of Q500. The maximum depth observed in case of Q1000 was also higher as compared to that of Q500. Before rehabilitation, maximum depth for Q500 was 5.0 m while for Q1000, it was 5.12 m. On the other hand, after rehabilitation, maximum depth in case of Q500 was 5.544 m while for Q1000, it was 5.794 m. Clearly, the maximum depths were higher after rehabilitation because of the fact that more water was retained in the main channel.
Fig 11 shows the comparison of velocity maps for Q500 before and after rehabilitation of Ismetiye creek.
In case of Q500 before rehabilitation, a maximum velocity of 4.44 m/s was observed at X= 2500 m in the main channel. High velocity of flow in a natural river bed would result in erosion of soil and deformation of the bed. After rehabilitation, a maximum velocity of 13.7 m/s at X=75 m was observed in the main channel. Thus, the velocity of flow in the main channel was much less before rehabilitation as compared to the case of after rehabilitation. This was observed due to the fact that the main channel before rehabilitation retained less water with high overflow. However, for the overbanks, velocity was observed to be higher before rehabilitation and lower after rehabilitation. This can also be attributed to the fact that there was high overflow before rehabilitation. Hence, it can be said that not only did the rehabilitation of channel decrease the overflow but also the decrease in velocity of overflow after rehabilitation reduced the risk of soil erosion near the banks.
For Q1000, maximum velocity of 4.47 m/s was observed at X= 2500 m in the main channel before rehabilitation. After rehabilitation, a maximum velocity of 14.8 m/s was observed at X= 75 m. As compared to the maximum velocities obtained in the case of Q500, the maximum velocities in the case of Q1000 were marginally higher. However, the locations of points of maximum velocities remained the same as that of Q500.
Flood hazard maps were created to compare the risk of floods before and after the rehabilitation of Ismetiye creek. Flood risk zones as per the guidelines of Federal Emergency Management Agency (FEMA) 2018 were defined as shown in Table 2. The magnitude of flood hazard is calculated by Eq. (3) (FEMA 2018).
Table 2. Flood risk classification scheme (modified from FEMA 2018)
|
Hazard (m2/s) |
Flood Risk |
Hazard Rating |
|
< 0.2 |
Very Low |
0 |
|
0.2-0.5 |
Low |
1 |
|
0.5-1.0 |
Moderate |
2 |
|
1.0-2.0 |
High |
3 |
|
> 2.0 |
Extreme |
4 |
Hazard zones were assigned a hazard rating starting from 0 to 4 with 0 corresponding to very low risk zone and 4 corresponding to extreme risk zone. Fig 12 shows the comparison of flood hazard zones before and after rehabilitation of Ismetiye creek for Q500 and Q1000.
From the hazard maps of natural channel and rehabilitated channel, it was observed that hazard rating was maximum inside the main channel taking the values 3 (high risk) and 4 (extreme risk) along the channel path.
The catchment area in case of natural channel was more prone to flooding as expected and the extent of hazard was more. On the right overbank side, there were traces of yellow, blue and red colors which correspond to a hazard rating of 1,2 and 3 respectively. The regions on right overbank side of Ismetiye with hazard rating of 2 (blue) and 3 (red) are in the vicinity of inhabited areas, making those areas susceptible to the danger of flooding.
However, after rehabilitation, the extent of hazard considerably decreased and a hazard map with green (hazard rating = 0) was observed throughout the catchment area except for the main channel and its immediate surroundings where some spots with yellow (hazard rating = 1) and blue (hazard rating = 2) were observed. However, these spots lie in uninhabited regions.
As expected, the extent of flood hazard was more in case of Q1000 as compared to that of Q500. Due to high velocity of flow in case of rehabilitated channel, flood hazard was more near the banks as compared to the case of natural channel. Before rehabilitation, new settlements towards right over bank side of the downstream end of Ismetiye channel were exposed to inundations. However, there is no flooding seen in these areas after rehabilitation of the channel.
Cross-sectional lines were drawn at X= 75 m, 1500 m, 3000 m and 4000 m along the channel in order to obtain flow hydrographs for Q500 and Q1000. Separate lines were drawn for the main channel (MC) and each overbank. Therefore, hydrographs for the MC, right overbank (ROB) and left overbank (LOB) were calculated separately for each cross-section. Fig 13 shows the cross-sections at X= 75 m and X=1500 m for Q1000.
Flow hydrographs for X=75 m and X =1500 m at Q=500 m3/s and Q=1000 m3/s are depicted by Fig 14.
Fig 15. shows the cross-sections at X= 3000 m and X=4000 m for Q1000.
Flow hydrographs for X=3000 m and X = 4000 m at Q=500 m3/s and Q=1000 m3/s are depicted by Fig 16.
The peak in MC before rehabilitation was attained at 84.19 m3/s and after rehabilitation, it was attained at 168.6 m3/s. At X=4000 m, before rehabilitation, it was also observed that flow hydrograph peak was constant after a certain point. This indicated that maximum flow capacity of natural channel was reached. Table 3 and Table 4 show the peak discharge values for main channel and overbanks.
Table 3. Peak discharge values for natural and concrete channels in case of Q500
|
|
Natural Channel |
Concrete Channel |
||||
|
X (m) |
QMC (m3/s) |
QROB (m3/s) |
QLOB (m3/s) |
QMC (m3/s) |
QROB (m3/s) |
QLOB (m3/s) |
|
75 |
91.27 |
25.81 |
26.45 |
143.65 |
0 |
0 |
|
1500 |
101.35 |
20.21 |
5.27 |
143.17 |
0 |
0 |
|
3000 |
83.9 |
50.84 |
6.24 |
143.6 |
0 |
0 |
|
4000 |
82.93 |
50.71 |
4.82 |
143.7 |
0 |
0 |
Table 4. Peak discharge values for natural and concrete channels in case of Q1000
|
|
Natural Channel |
Concrete Channel |
||||
|
X (m) |
QMC (m3/s) |
QROB (m3/s) |
QLOB (m3/s) |
QMC (m3/s) |
QROB (m3/s) |
QLOB (m3/s) |
|
75 |
104.58 |
29.86 |
31.36 |
168.42 |
0 |
0 |
|
1500 |
110.53 |
23.77 |
6.64 |
168.14 |
0 |
0 |
|
3000 |
85.5 |
73.29 |
8.44 |
168.3 |
0 |
0 |
|
4000 |
84.19 |
73.45 |
6.91 |
168.6 |
0 |
0 |
Here is a generalization of results that can be concluded from the above graphs valid for both Q500 and Q1000:
- It is obvious from the graphs that flow in main channel was more for rehabilitated case as compared to the natural case. This is because of the fact that channel capacity increased. The erosion of banks and deposition of sediments lowered the channel capacity in natural river. This resulted in higher peaks of hydrographs at the shown sections across the main channel after rehabilitation.
- On the other hand, before rehabilitation, due to high overflow in the main channel, the peaks of flow hydrographs were comparatively lower.
- Likewise, for the overbanks, the flooding was more in case of natural channel as compared to the rehabilitated channel which is also reflected by higher peaks of flow hydrographs.
These results clearly indicate that the rehabilitation work of Ismetiye creek confined most of the flooding to the main channel preventing the overbanks from risk of inundation. Thus, it is safe to say that rehabilitation work on a channel in its natural state makes the catchment areas of the channel less prone to the risk of flooding.
4.1. Summary and Conclusion
In this research, the affect of rehabilitation on Ismetiye creek in Osmangazi, Bursa was thoroughly studied. The main aim of this study was to compare the flooding of Ismetiye creek before and after its rehabilitation. After rehabilitation of the creek, channel bottom was made of concrete and the sides of channel were made with stone walls. The rehabilitated channel was 4020 m in length. The cross-sectional width of rehabilitated channel ranged between 12 m to 18 m.
The main goal of modelling Ismetiye creek was to identify the difference in risk of flooding before and after its rehabilitation. The model for simulation was developed using ArcMap and HEC-RAS. Initially, elevation data obtained from BUSKI in the form of AutoCAD files was imported to ArcMap. Subsequently, this elevation data was used to create a DEM through IDW raster interpolation. DEM was then exported as a terrain to RAS Mapper menu of HEC-RAS. The terrain file was then modified by adding secondary channels, roadways, buildings and bridges over the main channel. 2D simulations were run for Q500 and Q1000 to obtain the flooding in catchment area.
As of 2022, Ismetiye creek was rehabilitated by BUSKI for a stretch of 4020 m leaving the remaining part of the creek in its natural state. However, the simulations were carried out assuming that the entire channel was rehabilitated. From the 2D simulations for both Q500 and Q1000, it was found that Ismetiye creek before rehabilitation was more exposed to flooding with a high extent of water inundations as compared to the case of after rehabilitation. A maximum velocity of 4.8 m/s was observed for the main channel. Velocity on the overbanks was higher before rehabilitation which would result in erosion and deformation of the soil in case of flooding. Before rehabilitation, the overflow from main channel also reached the dry channel in the south of main channel resulting in the flooding of dry channel as well. From the results of simulations run for the natural channel, flood hazard was found to be significant for Q1000 but most of the catchment areas were found to lie in low-risk zone with a few exceptions where the flood risk was high. However, these areas were found to be in agricultural land. With regard to risk of flooding in residential areas, Demirtas Barbaros, Doganevler and Cumhuriyet neighborhoods of Ismetiye creek catchment area were exposed to high water inundations.
On the other hand, the 2D simulations with Q500 and Q1000 for the rehabilitated channel showed decrease in flooding as the extent of water inundations reduced considerably. Also, there was no flooding in the south due to the overflow from main channel. For Q1000, the maximum depth observed in the main channel was 5.120 m before rehabilitation and 5.794 m after rehabilitation. Thus, more water was retained in the main channel after rehabilitation, resulting in very less overflow which is also reflected by the flood hydrographs shown in the results for various sections. For instance, in case of Q500 at X = 75 m, QMC was found to be 91.27 m3/s for natural channel while as it was 143.65 m3/s for concrete channel. Thus, channel capacity increased for the concrete channel. Additionally, in case of Q500 at X = 75 m, QROB was 25.81 m3/s and QLOB was 26.45 m3/s respectively, for the natural channel. However, for rehabilitated channel, these values were found to be 0 m3/s due to no overflow.
The maximum velocity observed for the concrete channel was 14.47 m/s. Since most of the flow was retained in the main channel after rehabilitation, velocity of flow was very high for concrete channel as compared to the natural channel. After rehabilitation, a very small amount of flooding was observed in Ismetiye creek catchment area for both Q500 and Q1000. The entire catchment area was found to lie in a very low risk flood hazard zone. All the inundated areas were found to be in the agricultural lands lying in close vicinity of Ismetiye creek. There was no risk of flooding at all for any of the residential areas lying in the vicinity of Ismetiye creek. Thus, it can be concluded that complete rehabilitation of Ismetiye creek by BUSKI will entirely remove the threat of flooding in the nearby residential areas.
4.2. Recommendations
It should be emphasized that only a stretch of 4020 m of Ismetiye creek was rehabilitated while the rest of the channel still remains in its natural state, leaving some neighborhoods in Demirtas area of Osmangazi district like Doganevler and Cumhuriyet exposed to risk of floods.
Hence the following recommendations can be suggested to the environmental policymakers.
-
Keeping the risk of minor flooding for the areas of Doganevler and Cumhuriyet neighborhoods in mind, completing the unfinished rehabilitation works of Ismetiye creek is essential.
-
In case the rehabilitation work of Ismetiye creek is not completed, flood insurance schemes can be introduced for the residential neighborhoods which are susceptible to risk of floods.
Acknowledgements
The authors would like to extend their thanks to Bursa Su ve Kanalizasyon İdaresi (BUSKI) for providing the necessary data.
Declarations
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.
Funding
No funds, grants, or other support were received during the preparation of this manuscript.
Competing Interests
The authors have no relevant financial or non-financial interests to disclose.
Author Contributions
Both the authors, Amir Iqbal Wani and Serdar Korkmaz, contributed to the study conception and design. Material preparation, data collection and analysis were performed by both authors. Both authors read and approved the final manuscript.
- Abdessamed, D. and Abderrazak, B. (2019). Coupling HEC‑RAS and HEC‑HMS in rainfall–runoff modelling and evaluating floodplain inundation maps in arid environments: case study of Ain Sefra city, Ksour Mountain. SW of Algeria. Environmental Earth Sciences (2019) 78:586.
- Acosta, J., Keith, R., Hollite, J., Logronio, R., James, G. (2017). Flood modelling using GIS and LiDAR of Padada River in south-eastern Philippines. GISTAM 2017 - 3rd International Conference on Geographical Information Systems Theory, Applications and Management. 301-306.
- Astite, S.W., Belabid, N., Mahmouhi, N., (2015). Cartography of flood hazard by overflowing rivers using hydraulic modelling and geographic information system: Oued El Harrach Case (North of Algeria). Revista de Teledeteccion, 44:67-79.
- Bricker, Jeremy, D., Schwanghart, Wolfgang, Adhikari, Basanta, R., Moriguchi, Shuji, Roeber and Volker. (2015). Performance of models for flash flood warning and hazard assessment: The 2015 Kali Gandaki Landslide Dam breach in Nepal. Mountain Research and Development, 37(1): 5-15.
- Bursahaber (2010).https://www.bursahaber.com/bursayi-allah-korudu-makale,953.html
- Chow, V.T. (1959). Open Channel Hydraulics. McGraw-Hill, New York.
- Climate Change Post (2022). River floods Turkey. https://www.climatechangepost.com/turkey/river-floods/
- Fatih, S., (2013). İsmetiye (kelesen) deresi islah uygulama projesi. Mühendislik taşkin hidrolojisi raporu.
- FEMA (2018) Guidance for Flood Risk Analysis and Mapping. Federal Emergency Management Agency (FEMA), USA.
- W.B. (2020). HEC-RAS river analysis system. 2D modelling user’s manual. https://www.hec.usace.army.mil.
- Gulbaz, S., Cevza, M., Aysenur.. Ugur. (2019). Flood modelling of Ayamama RiverWatershed in Istanbul, Turkey. J. Hydrol. Eng., 2019, 24(1): 05018026.
- Hasheyan, F., Khaleghi, M.R., Kamyar, M. (2015). Combination of HEC-HMS and HEC-RAS models in GIS in order to Simulate Flood (Case study: Khoshke Rudan river in Fars province, Iran). Research Journal of Recent Sciences, 4(8): 122-127.
- Iglesias, (2014). Flood disaster mitigation and river rehabilitation by Marikina City, Philippines. Asian Disaster Preparedness Centre. http://www.adpc.net.mz
- İsmail, H., Enes, Y., Fatih, O., İbrahim, D. (2021). Comprehensive flood event specification and inventory: 1930–2020 Turkey case study. International Journal of Disaster Risk Reduction, 56 (2021): 102086.
- Lucien, H., Gebremeskel, S., Smedt, F., Hoffmann, L., Pfister, L. (2014). Simulation of flood reduction by natural river rehabilitation using a distributed hydrological model. Hydrology and Earth System Sciences, 8(6): 1129-1140.
- Macrotrends (2020). Bursa, Turkey Metro Area Population 1950-2020. https://www.macrotrends.net/cities/22670/bursa/population
- Mondal, S. and Priyank, P. (2018). Examining the utility of river restoration approaches for flood mitigation and channel stability enhancement: a recent review. Environmental Earth Sciences (2018), 77:195.
- Nataran, S. and Nisha, R. (2019). An integrated hydrologic and hydraulic flood modelling study for a medium-sized ungauged urban catchment area: A case study of Tiruchirappalli city using HEC-HMS and HEC-RAS. J. Inst. Eng. India Ser. A (June 2020) 101(2):381–398.
- Neslihan, B. and Asli, K. (2020). Flood hazard assessment of a flood-prone intensively urbanized area – A case study from Samsun Province, Turkey. Geofizika, 37 (2020): 1–25.
- Neslihan, B. and Asli, K. (2021) Flood map production and evaluation of flood risks in situations of insufficient flow data. Natural Hazards (2021) 105:2381–2408.
- Rosgen, D. L. (1994). A classification of natural rivers. Catena, 22, 169-199
- Sammen, S., Thamer, A., Quitiba, G. (2019). Environmental consideration in flood mitigation and river restoration. IOP Conf. Series: Materials Science and Engineering 518 (2019) 022088.
- Tahmasbinejad, H., Feyzolahpour, M., Mehdi, M., Fatemeh, Z. (2012). Rainfall-Runoff simulation modelling of Karun River using HEC-RAS and HEC-HMS model, Izeh District Iran. Journal of Applied Sciences, 12(18): 1900-1908.
- Thakur, B., Parajuli, R., Kalra, A., Ahmad, S., Gupta, R. (2017). Coupling HEC-RAS and HEC-HMS in precipitation runoff modelling and evaluating flood plain inundation map Sacramento, California. World Environmental and Water Resources Congress, 240-251.
- USGS (2006). Flood Hazards- A National Threat. USGS Fact sheet 2006-3026. https://pubs.usgs.gov/fs/2006/3026/2006-3026.pdf.
- WHO (2022). Floods -World Health Organization. https://www.who.int/health-topics/floods#tab=tab_1
No competing interests reported.