Estimating the Volume of Sediments and Assessing the Water Balance of the Badra Basin, Eastern Iraq, Using Swat Model and Remote Sensing Data

The primary objective of this study is to employ the remote sensing data and Soil & Water Assessment Tool model to estimate sediment volume and assess the water balance of the Badra Basin (2,615km2) in eastern Iraq. Remote sensing data was utilized as the main input with the Soil & Water Assessment Tool model. These data involved a land use-land cover map that was constructed by the classification of the Landsat-8 satellite imagery for the year 2020, STMR digital elevation model, soil map was acquired from the Food and Agriculture Organization and climatic data were sourced from the NASA-funded prediction of Worldwide Energy Resource. The results discovered that about 40 % and 18% of the yearly rainfall are losing by evapotranspiration and filtration. The average amount of annual sediment transported was predicted at 120.47 tons /ha, 2018 recorded the highest value of transported sediment which is about 360 tons /ha. The volume of annual runoff was assessed at about 340.74 million m 3 . These results proved that the Soil & Water Assessment tool model has the ability to estimation the sediment and runoff volume. The climatic elements, especially rainfall, in addition to soil classes, topography, and land use-land cover had a significant impact on the amount of transported sediments and the volume of runoff.


Introduction
The study and estimating of the volume of the sediment are some of the important studies that cannot be overlooked (Shendge et al., 2018).Ackermann et al. (1973) analyzed approximately 1,100 lakes in the United States of America and stated that most of these small reservoirs were filled with sediments during periods of less than 30 years.While Rãdoane & Radoane (2005) analyzed about 138 water reservoirs in Romania with storage capacity ranged between 1-1230 million cubic meters and determined that there are about 30 water reservoirs that will lose more than 30% of its capacity in periods between 2-10 years.Thakkar & Bhattacharyya (2006) reported that India suffers from losing approximately 1.3 (billion cubic meters) of the total capacity of the storage due to reservoirs filling with sediment.The eastern region of Iraq, including the study area, is a very important source for water harvesting and storage, but the main problem is the high rates of sediment carrying, which caused many dam reservoirs to fail (Abdullah et al., 2019).Therefore, the accumulation of sediments in the reservoir is a dangerous problem that maybe threatens the sustainability of the reservoir and has dire consequences for the productivity of the reservoir during its operation period.
On the other hand, water resources in the world and Iraq, in particular, suffer from many problems, the most important of which are global conditions represented in the lack of rainfall and high levels of temperatures, which led to an increase in the phenomenon of drought and the expansion of desertification (Al-Ansari, 2013;Awadh, 2018;and Awadh et al., 2021).Surface runoff water is one of the primary water resources.Therefore, studying their quantities, frequency, and probabilities is one of the most important studies, especially through the availability of field information or space data.There are few numerical models that have the ability to simulate hydrological processes, describe the water balanced elements, and estimate sediment volumes on a watershed scale and at a reasonable time scale (Boithias et al., 2017).One of the very important of these models is the SWAT (Soil & Water Assessment Tool) model, which is a physically distributed model (Ngo et al., 2020) based on mathematical models of the continuous simulation type.
The model of the SWAT was chosen in this work for estimating the volume of sediments and rainfall-runoff simulation, because it is a promising simulation model that ongoing to evolve day after day and can be greatly benefited from, especially in the study of ungauged and cross-border basins such as Badra Basin.Furthermore, many researchers around the world evaluated the SWAT model and with various conditions with basin areas ranging from 1 to 630,000 km2 (Rossi et al., 2009;Ezz-Aldeen et al., 2012;Tyagi et al., 2014).Consequently, the SWAT model has arisen as one of the most widespread and had been broadly utilized in the environmental and/or hydrological field (Gassman et al., 2014).The main problem in the Badra Basin is the high rate of sediment load during the rainy seasons, which threatens the failure of the dam that was built in the Badra Valley in Wasit governorate.It is a concrete weir 800 meters long and 3m high (Saeed, 2018).The aim of this study is to estimate the volume of sediments and assessing the water balance of the basin of Badra for the period 1991-2020, the SWAT model has been used.

Study Area
The basin of Badra is a water basin sited in the Wasit (Eastern part of central Iraq) and at the Iraqi-Iranian border between latitudes 32°47'44" and 33°38'44" North and longitudes 45°48'42" to 46°41'18" East (Fig. 1), with a total area of about 2,616 km 2 , extends into Iranian territory for a distance of approximately 81 km, with an approximate area of 2,319 km 2 (88% of the basin area), while it extends into the Iraqi territory for an approximate distance of 48 km and covers an area of about 296 km 2 (12% of the basin area) until its meets with the Shuweijah Marsh.On the Iraqi side, the Badra Basin water flows from the drainage of the basins and the mountains located on the northern and northeastern sides, including the slopes of the Bashtako Mountains.

Data Set
The data of remote sensing supply important data for water resource mapping (Abdel Rahman et al., 2016).Four major file's datasets required by the SWAT model: (a) Digital Elevation Model (DEM) was acquired from SRTM (Shuttle Radar Topography Mission), version3 with 30m spatial resolution (Ferretti & Guarnieri, 2007), and were download from the website of SRTM (http://srtm.csi.cgiar.org/),Fig. 2. a.(b) The soil map, map of soil is shown in Fig. 2.b was sourced from the FAO (Food-and-Agriculture-Organization).Digital-Soil Map of the World (DSMW), version 3.6, (Sanchez et al., 2009).This data comes with about 7000 types of soil for the whole world at a spatial resolution of 10km and a database for the properties of two layers of the soil, the upper layer at a depth of 30cm, and the lower layer at a depth 30-100 cm.(c) the map of Landuse-Landcover (LULC) was produced by utilizing the Landsat8 satellite image, free of cloud which acquiring on March 3, 2019.

Fig.1. Location map of the study area
The LULC result of the Badra basin contains 5 classes (Fig. 3 and Table 1).( d (Manhi & Al-Kubaisi, 2021).The extension of the ArcSWAT is a graphical user interface for the model of SWAT (Srinivasan et al., 1998).
The extension of the Arc SWAT contains various functions required to create and run the SWAT model, the most important of which are watershed delineation, input file generator, editing modules, and others.The Arc SWAT extension can be obtained for free and has been developed to work within the environment of many programs, the most important of which are Map Window GIS, GRASS, and QGIS.

Sediment Volume Estimation
To estimate the volume of the sediment to every HRU, the model of SWAT uses the Modified Global Soil Loss Equation (MUSLE).This equation was developing by Williams, (1975) through replacing the rain-fall factor (R) with a surface runoff factor and considers both sediment movement and surface erosion in the basin to improve the prediction of the sediment yield.Models based on physical processes such as the SWAT model depend on erosion calculation utilizing mathematical representations of basic hydrologic and erosion operation involving soil separation and transport (Foster, 1990).The MUSLE is:  = 11.8 (  *   *  ℎ ) 0.56 *  *  *  *  *  (1) Where the Sd represented the volume of the sediment (metric tons), Qsur represented the volume of run-off (mm), qp represented the peak of run-off rate (m3/sec), Ahru represented the area of the HRU (ha), K represented the factor of soil erodibility, C represented the factor of cover and management, P represented the factor of support practice, LS represented the factor of topographic, and CF represented the factor of course fragment.

Swat Hydrological Component and Assessment the Surface Runoff
There are two major approaches for assessing runoff (runoff is the flow that occurs along the sloping surfaces) within the model of SWAT based on the type of rain-fall data: the primary method is the modified SCS curve number (USDA, 1972) which requires daily rainfall data, and the second method is the method of the Green & Ampt (Green & Ampt, 1911) which needs sub-daily of hourly rainfall data.The SCS-CN method was applied in assessing the surface runoff volumes for the Badra basin.The hydrological cycle that will be simulated by the model of SWAT depends mainly on the estimation of the volume of both sediments and runoff.The model of SWAT usages the equation of water balance to simulate surface runoff (Neitsch et al., 2011), which is: (2) Where the SWf represented the final moisture content (mm); SWi represented the initial moisture content on a day i (mm); Rt represents the rain-fall volume on a day i (mm); t is the time ( days ); Qt represent the surface runoff on a day i ( mm ); ETt represent the Evapotranspiration on a day i ( mm ); Pt represent the percolation on a day i ( mm ), and QRt represent the return flow quantity on a day i ( mm ).
The SCS-CN method was established to become an appropriate basis to evaluate runoff volume under diverse land use with variable soil types.This method is frequently recognized as the CN (Curve Number) method, the land-use, and properties of the soil are lumped in one parameter (White et al., 2008).The model of SWAT forecast the volumes of runoff and the rates of peak runoff for each HRUs (Hydrologic Response Unit) used SCS-CN equation: Where the Qsur is the daily run-off or rain-fall excess in millimeters, Rday represents the daily rain-fall depth in millimeters, S represents the retention parameter (mm).Ia represents the initial abstractions which are generally be similar to 0.2S, thus the equation turns into: The parameter of retention changes spatially due to variations in the landuse, soils, slope, management, and temporally from the changes in the soil water content.The parameter of retention is well-defined as: CN = the day curve number.

Setup of the SWAT Model
The initial stage in implementing the model of SWAT is to prepare and collect all the data required for the model, followed by a number of important steps.Fig. 4 is a schematic diagram of the most important of these steps.It is important to standardize the projection for all data collected.The metric projection (UTM) is better and more accurate than the geographical projection because it deals with metric units.
The watershed delineation of the main basin and sub-basins utilizing DEM data represents the first step in implementing the SWAT model (Tolosa, 2015).The size of the sub-basins can be determined more precisely by manually editing the inlet and outlet points of each sub-basin.The construction of HRUs is the second step after the watershed delineation process.HRUs are areas that are aggregated and constructed within each sub-basin by overlaying a number of layers namely soil layer, LULC layer, and slop layer which are derived from DEM (Fig. 5) with the use of an appropriate threshold to enable the user to distinguish unique homogeneous lands.The use of appropriate thresholds in the process of building the HRUs will contribute to enhancing the operation of the simulation and facilitating the calculations processes (Srinivasan et al., 2010).
Once the HRUs are generated, the SWAT model must be supplied with all climatic data.Before loaded the climatic data, all the climatic statistical data required for all elements (precipitation, temperatures, wind speed, relative humidity, and solar radiation) were calculated using the SWAT weather generator program and then connected to the SWAT database.

Watershed Delineation
The main initial step in implementing the model of SWAT is the automatic delineation of the main Badra basin and each sub-basin.The digital elevation model was used to track the initial flow directions as well as the out-let point for individually sub basin.To increase the accuracy of the delineation process, the main outlet point was used at the confluence of the Badra basin with Hor Al Shuwaija, Fig. 6.As for the sub-basins, they were identified based on a threshold boundary of 3000 hectares or 30 km 2 and with the help of manual adjustment of the locations of the main outlet for each sub-basin.The results of the delineation process of the Badra Basin showed that the drainage area of the basin amounted to 2,615.76 km 2 with 4 sub-basins were named sub-basin 1 to 4. The height of the sub-basins ranged from 312 meters in the southwest to 2566 meters in the northeast, and that about 79% of the basin area have an altitude of less than 1000 meters.Table 2 shows some of the most important morphometric properties with the longest flow path of the Badra sub-basins.

HRUs
After the delineation process is completed, each sub-basin area is subdivided into small units named hydrological response units (HRUs).The hydrological unit is defined as the smallest unit that is calculated within the SWAT model and has unique characteristics in terms of the nature of each of the soil types, landuse, and degree of slope.The calculation and analysis of each hydrological unit will ensure the increased accuracy of the model of SWAT in predicting the volumes of each of the sediments and the runoff for each sub-basin.Then the results are collected for all sub-basins to calculate the average volumes for the entire main basin.The results demonstrate that Badra basin is separated into 29 smaller hydrological response units depended on landuse data, type of soil, and slope.Table 3 shows a summary of the HRUs data for each sub-basin.In addition to the concentration-time and curve number values for three possible states of soil moisture, namely, the dry state, the medium-moisture state, and the high-moisture state.

SWAT Predictive of the Sediment Volume
The model of SWAT was utilized to calculate the volume of sediment transported and deposited in the Badra Basin.The results of the sediment volume simulations conducted during the period 1991-2020 are shown in Fig. 7, While the most important results of the simulation elements are recorded in Table 4.The average annual of transported sediment of the Badra basin over the period 1991-2020 was assessed at 120.47 tons /ha, the year 2008 was the year in which the highest value of transported sediments was recorded.The highest value of transported sediments was recorded in 2008, which was estimated at 360 tons/ha due to a large volume of rainfall that was recorded through 2008, which amounted to about 705.5 mm.The lowest value recorded for the transported sediments was in 1999 and was estimated at 13 tons/ha.At the monthly level, the highest rates of transported sediment were recorded during the months from November to April, and the peak was in March.These high sediment loads can be explained by the fact that these months have high levels of rain and runoff and thus an increase in the amounts of sediment carried.

Water Balance Predictive for the Badra Basin
The model of SWAT considered as one of the hydrological models that can simulate the movement of water at the level of sub-basins or reservoirs, ponds, and rivers within the main basins.There are a number of important hydrological elements that may affect the water flow within the boundaries of the main basin, and the most important of these factors are surface run-off, groundwater movement, and lateral flow minus the amount of water lost during the transport process.Therefore, the total water productivity of the basin consists of these hydrological elements.Although groundwater is considered one of the elements of the water productivity of the basin, at the same time it is considered one of the filtering (percolates) elements.
The results in Fig. 8 and Table 5 show the annual average of the elements of the water balance for the period from 1991 to 2020 and on a monthly basis for the Badra Basin.About 16% and 40% of rainwater are lost through both filtration and evaporation processes, respectively.The results demonstrate that surface run-off is the main contributor to the total flow within the Badra basin, with a percentage of up to 75% of the total flowing water.While both groundwater flow and lateral flow contribute about 25% of the total water flow.More than 59% of the water in the mainstream of the Badra basin is an outcome of groundwater inflow.
) The climate data were collected from the NASA Langley Research Center (LaRC) Prediction of Worldwide Energy Resource (POWER) project funded by the NASA (the Program of the Earth Science/Applied Science) for three stations; Fig. 3 shows the location of these stations; and for 31 years starting from 1 January 1989 to 31 December 2019.Five climatic elements are required to implement a SWAT model which are rainfall, minimum & maximum air temperature, wind speed, solar radiation, and relative humidity

Fig. 4 .
Fig.4.A schematic diagram shows the SWAT Model setup

Fig. 5 .
Fig.5.A slope map of the Badra Basin

Fig. 8 .
Fig.8.Schema shows the water balance of the Badra Basin

Table 2 .
The most important morphometric properties of the Badra sub-basins

Table 3 .
The summary of the HRUs data for each sub-basin, The CN1 represented the Curve Number (CN) at dry state, CN2 represented the CN at the average moisture, and CN3 represented the CN at wet conditions, TC represents the time of Concentration

Table 4 .
The results of simulations