Current and Modified Flood Discharge Capacity of a Reach of Tigris River between Kut and Amarah Barrages

This study was conducted to examine the discharge capacity of the reach of the Tigris River between Kut and Amarah Barrages of 250km in length. The examination includes simulation the current capacity of the reach by using HEC-RAS model. 247cross sections surveyed in 2012 were used in the simulation. The model was calibrated using observed discharges of 533, 800, , discharged at Kut Barrage during 2013, 1995, 1995 and 1988, respectively /s 3 m and 300

m and 3000 1025 and its related water level at three gauge stations located along the reach. The result of calibration process indicated that the lowest Root Mean Square Error of 0.095 can be obtained when using Manning's n coefficient of 0.026, 0.03 for the Kut-Ali Al Garbi and Ali Al Garbi-Amarah reaches respectively, and 0.03 for the flood plain of the whole reach under study. The reach under study has two lateral inflow streams, UmAljury, which joins Tigris River at station 51km, and Aljabab, which joins Tigris River at station 57km. The discharge of Aljabab varies .

INTRODUCTION
The Tigris River is one of the two main rivers in Iraq. Its catchment area covers about 221000 2 distributed across Iran, Turkey, Syria, and Iraq. The total length of the river is about 1850km. The river originates at the Taurus Mountains in Turkey. It flows towards southeastern parts of Turkey along the borders strip of Turkey and Syria then enters Iraq at Fiesh Khabur City northern of Iraq. The river joins Euphrates River at Qurna City south of Iraq to form Shatt Al Arab River. The Rivers has five main tributaries that are: Khabur, Greater Zab, Lesser Zab, Adhaim, and Diyala. The flow hydrograph of Tigris River was changed dramatically as a result of the construction of several dams on the river main streams and its tributaries, the climate change, and increased water demand. The climate change has led to irregular rainfall periods, having severe drought for years followed by an extensive flood in short period. The expansion in construction of dams in Turkey and Iran decreased the incoming flow measured at the Iraqi borders. As a result, both of the water level and the flow velocity of Tigris Rivers were reduced. This allows reducing the capacity of cross-sections due to deposition of suspended load over the years, especially in the southern part of Tigris River. Under this condition, it's essential to estimate the current flood capacity of Tigris River and to make modification to the cross-sections to account for future expected extensive flood.
Several studies were conducted on the Tigris River focusing on bed sediment load, computing Manning's coefficient, water quality models, comparing cross-sections of the river with different periods and computing current capacity for the river by using mathematical models. Most of these studies were conducted on the Tigris River within Baghdad. There is a lack of studies amid to specify the capacity of Tigris River under normal operating conditions and during floods in the current conditions of the river's cross-section. Generally, this study is conducted to evaluate and modify the flood capacity of the Tigris River between Kut and Amarah Barrages. This reach of the river is of 250km in length, and it contains many growing alluvial islands and sidebars which led in narrowing the cross-sectional area of flow. The extreme recorded discharge of Tigris River downstream Kut Barrage is 3000m 3 /s recorded during 1988 and 200m 3 /s recorded during 2010.

DESCRIPTION OF THE REACH OF TIGRIS RIVER UNDER STUDY
The reach of Tigris River is located between Kut and Amarah Barrages,  The reach under study has two lateral inflow streams, UmAljury, which joins Tigris River at station 51km, and Aljabab, which joins Tigris River at station 57km. The discharge of Aljabab varies between 0 and 400m 3 /s, and the discharge of UmAljury varies between 0 and 50m 3 /s. The reach has many lateral intakes of irrigation canals distributed along the left and right banks of reach. These intakes are used to supply water to large agricultural areas of approximately 100,000 hectares. Amarah Barrage regulates the water flow of the main stream of Tigris River, Musharrah and Kahla Rivers located upstream of barrage. Table 1, )MoWR, 2017(, shows data of the flood escapes and intakes within the study reach.

DISCHARGES OF THE TIGRIS RIVER
The historical data of flow measurements for the Tigris River within the study area covers the period between 1988 and 2017 with some missing daily flow records. These data were provided by National Center for Water Resources Management, )NCWRM, 2017(. They were measured at three gauge stations along the Tigris River within the study reach, which are the Kut Barrage, Ali Al Garbi, which is located at 105km downstream Kut Barrage, and Amarah Barrage gages stations. Major flood events sets at different dates were obtained from the provided historical data records, including water level and discharge data at the three measuring stations, as shown by Table 2. The flood with a maximum discharge took place during 1988 of 3000m 3 /s on 30 March. Actions were taken to reduce the discharge of this wave by using the flood escapes and flooding some agriculture areas so that just 270m 3 /s of the wave reached Amarah City, )NCWRM, 2017(.

USED SOFTWARE
The

Best fit
After the computations of steady flow are completed, the output of the data results will be viewed including longitudinal and cross-sections profiles. Also, the computations showed other hydraulic components.

INPUT DATA
To perform the computations of the one-dimensional steady-state gradually varied flow within the 250km reach of Tigris River between Kut Barrage and Amarah Barrage by using the HEC-RAS(Version 5.0.3) model, the following data were defined to perform these computations: -Schematic layout of Tigris River: it was sketched in the positive direction of flow from Kut Barrage at the upstream to Amarah Barrage at the downstream end of the study reach. Flow values, Table 1, at each of these locations were balanced so that the continuity equation is satisfied.
-Cross-sections geometry: the 247 cross-sections were located and their data were defined. The data of cross-sections include: station along with the cross-section and elevation data, banks, levees and lengths of downstream reach, Manning's n roughness coefficient of the main stream and the right and left banks and coefficients of expansion and contraction of the main channel. Interpolation was conducted for cross-sections in HEC-RAS model to provide accuracy in computation of friction losses using maximum distance between two sections of 250m.
-Boundary conditions: in a subcritical flow, the model requires the boundary at downstream of studied channel. Whereas, in supercritical flow, the model requires the boundary at upstream of the studied channel. In addition, the HEC-RAS model requires the amount of flow at upstream of channel for successful run. The boundary conditions that were used in the model is the discharge at the upstream of study reach and rating curve at downstream of reach at station 251+00, which is cross-section number zero and is located at the upstream of Amarah, Fig. 2

DESIGN OF RUNS
The HEC-RAS mathematical model was implemented to simulate and analyze water flow through the Tigris River between Kut and Amarah Barrages based on the surveyed cross-sections and recorded data of the flow.
Calibration is needed to define appropriate values for the Manning coefficients for the main channel and floodplain. The Root Mean Square Error, RMSE, was used to test the compare the computed and the observed water surfaces Calibration of the model was achieved by using stage measurements along Kut-Amarah reach at three gaging stations. Four sets of data were used in the calibration. These sets are: set number 1 to 4, which represent the data with heigh flow varied between 3000 and 533m 3 /s discharged from Kut Barrage. Verification of the calibrated Manning's n is achieved by using one set of data that is set number 5. The scope of the simulation is to find the maximum flow that can safely pass through the Tigris River reach between Kut and Amarah barrages under current condition, with the actual crosssections, and under modified critical cross-sections into trapezoidal cuts for 3300m 3 /s discharge, which is the maximum flood wave observed downstream Kut Barrage in 1988. Different scenarios applied to the model by increasing the discharge at the upstream of the reach to find out the current capacity for the Tigris River within the study reach and critical discharges that can cause the inundation. Each scenario will contain two cases, one of them with maximum lateral inflow and the other one with zero lateral inflow. In all scenarios, the outflow from the intakes will be at their minimum as possible to identify the critical discharge in the main reach after the intakes. The identification of the flow that can pass safely between Kut and Amarah barrages is directly connected to minimum freeboard resulting from the simulations in the surveyed available cross-sections. It is assumed that flow passes safely when a minimum freeboard is higher than 0.5m. After computing the current capacity and defining the cross-sections causing chocking to the flow, modifications are conducted in three steps to find out the maximum discharge carrying capacity after each made modification. In each step simulation runs are conducted to evaluate the flow conditions along the reach. The first step is to modify cross sections by removing growing islands and sidebars. In the second step, cross-sections with high flow velocities more than 1m/sec, are modified. Finally, in the third step, modifications to cross-sections are conducted to increase the capacity of reach of the river to accommodate a discharge of 3300m 3 /s.

RESULTS AND DISCUSSION 7.1 Calibration and Verification
Through many trial runs of the HEC-RAS simulation model, it was found that the differences between observed and computed water surface profile can be decreased if the reach is separated into two parts having different values of Manning's n coefficient. The first part is the first 105 kilometers extended from Kut Barrage to Al Garbi City. The second is the remaining 145 kilometers extended from Al Garbi City to Amarah Barrage. Table 3 shows the obtained RMSE when using different values of Manning's n coefficients during the calibration process along the reach between Kut and Amarah Barrages. The process was conducted by using the four sets of data at the three gaging stations. These four sets of data discharged from Kut Barrage. /s 3 m flow varied between 3000 and 533 represent the data with a The lowest RMSE of 0.095 was obtained when using Manning's n coefficient of 0.026, 0.03 for the Kut-Ali Al Garbi and Ali Al Garbi-Amarah reaches respectively, and 0.03 for the flood plain of the whole reach under study. It is clear that the most recent observation among the sets of data, which is set number four, has the lowest RMSE and the simulated water surface elevations closely match that observed compared to other sets of data. This is referred to the date of surveying the cross-section and the date of recording each set of data. The cross-sections used in the simulation model were surveyed at a date closer to that of recording the data of set number four. In the other words, the geometric data surveyed during 2012, which was used in model, are more representative to the condition of the cross sections when taking the water level measurements during 2013 that other dates. The calibrated Manning's values were verified using one set of data that is set number 5. Table  4, shows the water surface profile for verification process. Table 4 presents a comparison between the observed water surface profile of set number 9 and that simulated by using the calibrated Manning's coefficients. The results of the verification process showed good agreement between observed and computed water surfaces with RMSE of 0.19,

Current Capacity of the Reach
Estimating the discharge capacity of the reach of Tigris River under study was conducted by increasing an assumed discharge at Kut Barrage and different scenarios for the lateral inflow and outflow until the critical discharges are reached that can cause the inundation. It was found be that discharges below 400m 3 /s are completely accommodated by the main channel of the reach. At discharges more than 400m 3 /s the water surface elevations reach the flood plain at different locations. This result was obtained for the two cases of lateral inflow and outflow. In first case, no lateral inflows while the second case there is lateral inflow. Fig. 3 and Fig. 4. show the water surface elevations along the reach at a discharge of 400m 3 /s for the first and second case, respectively. It is clear that the water surface elevations reached the flood plain of the reach at different locations. By increasing the flow out of Kut Barrage to a discharge of 1800m 3 /s, all water levels are kept below 1m than the levels of the left and right levees of the reach of Tigris River. This value of discharge is considered as a critical discharge before inundation occurred at which at least 1m of freeboard is ensured along the river reach. This result was the reach by analysis two cases of assumed lateral inflows and outflows along the reach of the river. In case one, all of the lateral inflows were assumed to be zero. Fig. 5 and Fig. 6 show the water surface elevations along the reach at a discharge of 1800m 3 /s for the first and second case, respectively, which show that the water surface elevations are kept at least 1m below the elevations of the levees of the reach of the river. While the analysis of the second case showed that the freeboard will be less than 1m from station 30+150 to station 91+000. At station 60+150 it is just 5cm freeboard.  When the discharge downstream Kut Barrage is increased to 2000m 3 /s with no lateral inflow, it was found that the freeboard will be less than 1m below the levees in many locations downstream of Kut Barrage as shown by Fig. 8. It must be noted that this discharge passed safely through the reach of Tigris River at that time. Fig. 9 and Fig. 10 show the water surface for the two cases, respectively. It is clear /s 3 m 3300 of elevations along the reach at a discharge that under the current conditions of the cross-section, the reach of Tigris River can't accommodate the flood discharge of 1988. The elevations of the water surface obtained with conditions of lateral inflow and outflows of case one are higher than the levees of the river reach. The elevations of the water surface obtained for case two have much higher extent than that of case one. This is an indication that the change in the hydrograph of the river and the reduction in the flow caused deterioration of the river capacity due to deposition of large amount of sediment for the period from 1988 and 2012.

Modification of Cross-sections of the Reach
It was shown previously that the current capacity of the reach of Tigris River between Kut Figure 11. Modified cross-section number 245, station 1+000. and Amarah Barrages is limited to 1800m 3 /s when there are no lateral inflows or 1700m 3 /s when there are lateral inflows along this reach. These flow discharges form nearly half of the flood discharge of 1988. More than thirty years of low flow within the reach with the lack of periodic maintenance are the main cause of deterioration of the capacity of this reach of Tigris River. Therefore, it is a very important issue that should be given utmost attention to recovering the capacity of the river to accommodate expected flood discharges. The strategy to recover the capacity of the river reach is to identify islands and sidebars along the reach and modify the cross-sections at their locations. Then specifying cross-section causing choking to the flow and modify these cross-sections one by one. The cross-section of 12 islands and 2 sidebars along the reach of Tigris River between Kut and Amarah Barrages were modified by reshaping into trapezoidal cross-sections.  Fifty-eight cross-sections were specified as cross-section causing choking to the flow along the reach of Tigris River between Kut and Amarah Barrages. These cross-sections of the river were expanded and reshaped into trapezoidal cross-sections. Fig. 12 shows a sample modified crosssection number 219 located at station 27+150. Results of the model runs with modified cross-sections showed that the water levels along the reach are lowered by an average of 20cm in addition to that 20cm when modifying the crosssections at the islands and sidebars so that the achieved lowering in water level is 40cm compared to current conditions of the cross-section.  A flow of 2100m 3 /s can be now safely discharged at Kut Barrage without any lateral inflow and 1900m 3 /s with lateral inflow from the tributaries. Fig. 13 and Fig. 14 show the water levels along the reach of Tigris River after modifying the cross-sections for a discharge of 2100m 3 /s without any lateral inflow and 1900m 3 /s with lateral inflow, respectively. All water levels along the reach are kept at 1m below the level of levees.  With the above modification of the cross-section the flood discharge of 3300m 3 /s during 1988 cannot be reached. Additional cross-section needs to be modified. These cross-sections were selected based on the model runs with a discharge of 3300m 3 /s. The cross-section at which the water level reaches or be above the levels of the levees of the reach. Additional 111 crosssections were selected to be modified. These cross-sections were modified by reshaping into trapezoidal cross-sections. Results of the model runs with modified cross-sections showed that the reach can safely pass the flood of 3300m 3 /s without any lateral inflow and 3000m 3 /s with lateral inflow from the tributaries. Fig. 15 and Fig. 16 show the water levels along the reach of Tigris River after modifying the cross-sections for a discharge of 3300m 3 /s without any lateral inflow and 3000m 3 /s with lateral inflow, respectively. All water levels along the reach are kept at 1m below the level of levees.

Distance, km
Ground level after modification Left levee Right levee Water surface level Figure 15. Water levels along the reach after removing 12 islands, 2 sidebars and 169 expanded cross-sections for the discharge 3300m 3 /s with fully inflow discharges, without any lateral inflow. Figure 16. Water levels along the reach after removing 12 islands, 2 sidebars and 169 expanded . with fully inflow discharges, with lateral inflow /s 3 m sections for the discharge 3000 -s s cro

CONCLUSIONS
-Under the current conditions of the river reach, the water surface elevations will reach the flood plain of the reach at different locations at discharges more than 400m 3 /s. -The discharge capacity of the Tigris River between Kut and Amarah Barrages was significantly changed compared to its capacity of 1988. The reach can now safely accommodated 1700m 3 /s discharged at the Kut barrage with lateral inflow and 1800m 3 /s when there are no lateral inflows while it was 3300m 3 /s during 1988. -In case of removing 12 islands and 2 sidebars within the studied reach, the water surface level will reduce by an average of 20cm and flow of 1900m 3 /s can be discharged safely at Kut Barrage without any lateral inflow and 1800m 3 /s with lateral inflow from the tributaries. With cost of 95 billion IQD. -In case of expand fifty eight narrow cross-sections in addition to removing the islands and sidebars, the water level will declined by an average of 20cm in addition to that 20cm when modifying the cross-sections at the islands and sidebars the reach can safely pass 2100m 3 /s discharged at Kut Barrage without lateral inflow and 1900m 3 /s with lateral inflow from the tributaries. With cost of 223 billion IQD. -In case of reshaping additional 111 cross-sections along the reach, the discharge capacity can be safely increased to 3300m 3 /s without any lateral inflow and 3000m 3 /s with lateral inflow from the tributaries. With cost of 741billion IQD.