Morphological analysis and evaluation of Givi-chay River conduit form (northwest of Iran)

NDWI

Normalized Difference Water Index

MNDWI

Modified Normalized Difference Water Index

AWEI

Automated Water Extraction Index

WRI

Water Ratio Index

Q5

Equal to the amount of flow that is equal to or greater than the mentioned amount in 5% of the days of the flow year

Q10

Equal to the amount of flow that is equal to or greater than the mentioned amount in 10% of the days of the flow year

Q15

Equal to the amount of flow that is equal to or greater than the mentioned amount in 15% of the days of the flow year

Q25

Equivalent to a flow rate which is a flow that is equal to or greater than 25% of the flow's annual volume

Q35

Equal to the flow rate, which is equal to or greater than the mentioned amount in 35% of the days of the flow year

Q45

Equal to the flow rate, which is equal to or greater than the mentioned amount in 45% of the days of the flow year

Q50

Equal to the flow rate, which is equal to or greater than the mentioned amount in 50% of the flow year

Q55

Equal to the flow rate, which is equal to or greater than the mentioned amount in 55% of the days of the flow year

Q65

Equal to the rate, which is equal to or greater than 65% of the flow year

Q75

Equivalent to a flow rate which is a flow that is equal to or greater than that in 75% of the flow year

Q85

Equivalent to a flow rate which is a flow that is equal to or greater than 85% of the flow's annual volume

Q90

Equal to the rate of flow which is equal to or greater than 90% of the flow year

Q95

Equivalent to a flow rate which is a flow that is equal to or greater than 95% of the flow's annual volume.

River and river processes are considered as the most significant geomorphic systems which are active on the earth's surface (Bag et al. 2019). Over time, many changes in the morphology and dynamics of a river system can occur (Kondolf & Piégay 2003). Metamorphosis of rivers can be gradual and continuous in the long-term approach, or it can be periodic and sudden if considered under certain circumstances in the short-run (Remondo et al. 2005; Garde 2006). The effects of river adjustment caused by natural factors require a much longer time-span to reveal. However, there are the few exceptions that natural factors such as river floods, landslides, or earthquakes can induce channel adjustments in a very short time (Chaiwongsaen et al. 2019). In contrast, human activities can have a significant and rapid impact on natural processes and trends, resulting in a compressed timescale for river adjustments (Rinaldi & Simon 1998). Since these disturbances cause substantial changes to flow and sediment structures, at present few rivers are in a natural or semi-natural condition (Surian & Rinaldi 2003). Consequently, the flow structure of many rivers has been dramatically altered from their natural flow structure (Nanson & Hickin 1986; Oorschot et al. 2018).

Therefore, it can be said that several factors are involved in river morphology and researchers have pointed to them in various studies. These factors include the following: lithology of wall and bed materials (Schumm 1985; Jain & Sinha 2005; VanLaningham et al. 2006; Petrovszki & Timár 2010; Pike et al. 2010; Ziyad 2014; Falkowski et al. 2017; Batalla et al. 2018), urbanization and sand harvesting (Gregory 2006; Surian et al. 2011; Mandal et al. 2016), anthropogenic effects such as flood deflection, waste dumping, residential construction, docks and bridges (Debnath et al. 2015; Bandyopadhyay & Kumar De 2018), dam construction (Brandt 2000; Snoussi et al. 2002; Grant et al. 2003; Richard et al. 2005; Gordon & Meentemeyer 2006; Graf 2006; Bisantino et al. 2010; Ollero 2010; Shin & Julien 2011; Ashouri et al. 2014; Casado et al. 2016; Yang et al. 2018; Chen et al. 2019; Hammerling et al. 2019), and climate change (Zhou et al. 2012; Zhang et al. 2017). Morphological responses may include subtle shifts in cross-sectional stream channel geometry or widespread landscape transitions, involving progressive or abrupt change over daily to millennial timescales (Slater et al. 2019). The changes in river morphology may cause many problems, such as changing the river course, flooding of nearby areas, and damage to hydraulic structures, as well as having some environmental impact (Chinnarasri et al. 2008). In order to sustainably manage river systems, it is necessary to further investigate the characteristics of variation in river morphology at various temporal and spatial scales (Minh Hai et al. 2019).

Givi-chay River is one of the permanent rivers of Ardabil, in the northwest of Iran. The main issues related to the Givi-chay River are bed changes over time, river instability, river departure and erosion along the river, and the existence of different river patterns in different parts of it has led to different processes in different parts of the river. In addition, the Givi Reservoir Dam has been constructed on the Givi-chay River to regulate river water to control floods and also to supply drinking water, industry, and land improvement in the region. The construction of the dam and its related consequences can provide a very good basis for intensifying the hydrological and morphological changes of the river. Researchers have not yet done a comprehensive study on this river. This study attempts to investigate the changes of geomorphological indices of the Givi-chay River over the period of time 2000–2019, and considering that dams affect the flow structure and sediment load (Nelson et al. 2013; Overeem et al. 2013), one of the factors that can strongly disrupt the stability of rivers is the construction of dams (Liaghat et al. 2017). Evaluation of hydrological conditions, especially regarding the construction of Givi Dam and its effects on conduit morphology, is also discussed. As stream power can be an extremely useful index of fluvial sediment transport, channel pattern, river channel erosion, and coastal habitat development (Barker et al. 2009), changing river power causes changes in sediment transport balance and sediment load rate, which causes morphological forms that are found in the riverbed; specific power and total power have been estimated for different return periods. In order to determine whether the increase in river specific power will have the same effect in different regions or whether the parameters of each section are the most important factor in determining the type of cross-section and the number of cross-section changes, a calculation of the shear stress was performed. Determining and identifying morphological parameters, hydrological indicators, power status, and shear stress and studying the geological conditions and land use status along the Givi-chay River route, while being aware of the river behavior, can be used to deal with the river and management, and the implementation of the reorganization plans in them was quite conscious and based on the rules governing the river.

In this research, a topography map with a scale of 1: 50,000, geology map with a scale of 1:100,000, and Google Earth and Landsat 8 images, including OLI sensor (2019), Landsat 7 including ETM+sensor (2000), bedrock maps and the Givi-chay River at a scale of 1:2,000, hydrological data from the two stations of Abegharm (upstream of the dam) and Firoozabad (downstream of the dam) and field data are used. The maps for determining the limits of the Givi-chay River bed and boundaries with a scale of 1:2000 were prepared by the regional water organization of Ardabil province, including physiographic information (measurement curves), the distribution of urban and rural points in the bed, and the river boundary is the location of existing constructions on the river bed and boundary (bridges, diversion dam). Also, to control the results obtained by quantitative methods it is used from field studies for confirmation and verification. ENVI 5.3, ArcGIS 10.5, HEC RAS, Smada, Excel and SPSS software were also used for image processing and data analysis.

Atmospheric correction was performed by FLAASH on satellite images first, and then using water indices, including NDWI (Green−NIR)/(Green+NIR), MNDWI (Green−MIR)/(Green+MIR), AWEI_ no shadow (4×(Green−SWIR1) −(0.25×NIR+2.75×SWIR2)), AWEI_shadow (Blue+2.5×Green−1.5×(NIR+SWIR1)−0.25×SWIR2)) and WRI (Green+Red)/(NIR+SWIR1), respectively, by McFeeters (1996), Xu (2006), and Feyisa et al. (2014); the river route was extracted from the images. According to the kappa coefficient (0.98 for 2000 and 2019) and overall accuracy (99.05% for 2000 and 99.42% for 2019), the river path extracted by the AWEI_sh index for both periods of 2000 and 2019 was confirmed and used. After river extraction according to water spectral indices, in the next step, the Givi-chay River route was initially divided into two upstream and downstream sections of the dam. Then, based on various variables such as geological and topographic control, flood plain width, and anthropogenic effects, according to field observation and use of Google Earth images and Landsat satellite, it was divided into four intervals for the more accurate and scientific study of the river: interval 1 (upstream dam interval, from the junction of the Hirochay and Arpachay to the dam), and the downstream dam intervals of interval 2 (lowland), 3 (mountainous) and 4 (sub-mountainous).

There is also a significant difference between the two stations in terms of normal discharge, moderate discharge, and annual discharge, which can lead to the frequency of sediment regimes in the canal and canal bed loading and erosion (Figure 4(c)–4(e)) and (Figure 4(g) and 4(h)).

The results show that the bights increased from 75 to 100 during this time, indicating morphological and instability changes in the river (Mohammadi et al. 2008; Esfandiary & Rahimi 2019). According to Table 3, the mean curvature coefficient for the first period in 2000 was 1.48 and decreased to 1.40 in 2019. But in other periods in 2019, the bending coefficient increased compared with 2000, with the bending coefficient from 1.23 to 1.25 in the second period and from 1.85 to 1.86 in the third period, it increased from 1/15 to 1/18 in the fourth quarter. In general, the lowest bending coefficient for each period is in the fourth interval and in a finite amount. In the first, second, and fourth intervals, most of the intervals in both study periods have a curvature coefficient of 1.5–1.05 and therefore the conduit plan is sinusoidal, but in the third interval more than 60% of the range has a curvature of 1.5 to 2 and therefore the interval pattern is in the form of a meander. In the second and fourth intervals, the standard deviation of the bending coefficient is low and in the second interval it is 0.19 in 2000 and 0.18 in 2019, and in the fourth interval it is 0.14 in 2000 and 0.12 in 2019. In general, they indicate the existence of similar arcs. In the first and third intervals, the standard deviation is relatively high for both periods, indicating non-similar arcs (Mohammadi et al. 2008). In the first and third intervals, a severe meander pattern is also seen in the intervals (Table 3).

The decreasing trend of wavelengths at all intervals in 2019 relative to 2000 indicates a decrease in the number of successive meanders, meaning that the number of meandering streams has increased, as the number has increased. The decrease in wavelength should be accompanied by an increase in the number of meanders, and in the present study, the number of meanders has also increased. The meandering of the rivers at all intervals confirms this. The wavelength and valley length decreased in the second period compared with the first period, indicating a decrease in erosion processes and the superiority of sedimentation processes and a downstream flow force. In general, the average arc radius declined throughout the years from 2000 to 2019. But when comparing the intervals with each other, in the third interval the radius of curvature of the curves in both periods is smaller than the other intervals and is an indication of the pressure and instability of the arc and the greater twisting energy for erosion. That is, the curvature of the flow path in the third interval is greater for shear and displacement. In the second interval (plain interval), due to the presence of erodible formations on the banks of the river, the freedom of action and power of the river movement is high, and as a result of faster movement and high erosion, the radius of the rings from the other study interval is increased. Therefore, the larger arc radius is due to riverbed erosion and lower radius due to mountains and lithology, and the change in river radius can be due to engineering actions and human interference along the river path and reduced river sedimentation due to construction of a dam that has severely affected the morphological behavior of the river (Surian & Rinaldi 2003).

According to studies (Magilligan 1992; Hafez 2000; Flores et al. 2006), it can be said that the factors affecting river power are the width of the channel and the slope. In the first interval, the slope of the riverbed is high. However, downstream of the Givi Reservoir Dam where the river enters the flood side, especially in the area of Saidabad, the slope of the riverbed decreases up until Firoozabad, causing the alluvial materials to be deposited and the width of the riverbed to rise. In the interval after Firoozabad, the river gets steeper because of the folds of the river banks, forming a rocky and bare riverbed. The area under study is made up of metamorphic and volcanic rocks alternately placed between sedimentary layers. The oldest area has been formed by biosparite lime and siliceous limestone up to a thick layer as the youngest layer, which is the sediments of present-day rivers. Tertiary rocks are mainly the result of extensive volcanic activity, especially in the Eocene, Oligocene and Miocene, followed by the deposited conglomerate, sandstone, marl, gypsum, and limestone layers. The present-day sediments include fine and coarse alluvial material, sediments, and river deposits. In the first period, the river flows into a valley bed, and in parts formed by erodible formations and at sections close to the dam, the river width is generally increased. In addition, in the parts that are composed of the porphyry and megaporphyritic layers of the Eocene, these layers lose their resistance when exposed to water and then they become erosional. However, the erosion rate of these rocks is low compared with the amount of erosion in alluvial layers or alluvial terraces. In the second interval and immediately after the Givi Dam, the river passes through the valleys overlooking Givi town, and in this area the width of the bed due to the nature of the bank decreases and the riverbed has coarse sediments and is covered with broken stones resulting from falling from the banks and being transported from upstream (sections 8 to 10). Turning away from the dam and crossing the Givi River, the Givi Plain downstream of the main flood plain is entered, with a width of between 500 and 1,000 m and more than 12 km to Firoozabad (sections 11 to 20) and most of the arable lands. The gardens in this area are located in the flood plain of the river, and due to the passage of the river through erodible formations such as marl, especially near Sekerabad and Mikailabad, widespread riverbeds and young gullies are observed (VanLaningham et al. 2006; Pike et al. 2010; Lecce 2013). Large volumes of flanking material (especially during floods) are eroded and loose flanks lead to the widening of canals and intra-channel ridges, and these sediments are clearly visible in bends, middle islands, and marginal lands and steep banks, which are constantly displaced and eroded by currents, and the bed morphological variations in this range are high. Decrease in width at section 14 is due to occupation by debris from gardens and construction. At the beginning of the third period, the Firoozabad area is in the continuation of the flood of the previous interval and by joining Sanghorchay, the river enters the mountainous part and the coastal areas have deep valleys with steep slopes along the river due to encountering high altitudes and rocky outcrops, and the alternative route has a meander and river changes that are subject to valley changes and the meandering state is seen throughout the valley (sections 21 to 28). The course of the river in this period corresponds to the Ean geological unit (an alternation of andesite lava, andesite basalt, and basalt with andesite-shaped glass tuff), Ngms (alternation of gray and red gypsum marls with gray sandstone in layers of micro-coagulation) and Ngc (red conglomerate). In the fourth period, the river width is reduced and the riverbed is covered with coarse sediments, which extends to Ghezelozan, in some areas due to rock material loss from canal-fill elevations and redirection. During this period, the formation of sections is subject to lithological resistance of the riverbed and no agricultural uses are observed (Schumm 1985; VanLaningham et al. 2006; Pike et al. 2010; Lecce 2013; Falkowski et al. 2017; Batalla et al. 2018). Therefore, the low river width in the fourth interval (sections 29 to 32) is due to lithological resistance and from Firoozabad to the River Ghezelozan river-crossing corresponds to the Ean geological unit with a north–south trend, and this periodic unit of andesitic lava, andesite basalt, and basalt is associated with andesitic glass tuff.

Therefore, considering the width and slope parameters of the river and the calculation of total power and specific power (Table 4) and based on the power of 300 watts, as erosion power, it can be said that in terms of total river power and the cross-sections of the first interval (upstream of the dam), it does not have much erodible power. But in the downstream part of the dam, in the second interval (sections 11, 12, and 13), it has a total power of 799.515 W/m² and then (sections 8, 9, and 10) a total power of 575.651 W/m². In the third interval, in sections 25 to 28, the total power is 351.787 W/m², and in the fourth interval the river power is 671.593 W/m² and beyond the erosion threshold. In terms of specific river power at the average annual flow rate, the river does not perform major erosion work at any of the sections, except where the stream is directly impacted by the river banks and performs scouring or cutting or sanding of the shores, or where the current hits a screw directly. In terms of the average annual flow, the lowest specific power of the river was in the 19th and 20th sections. The base flow (Q2.33) plays the greatest role in the shape, pattern, and variability of the river sections, and in the active sections, the riverbed undergoes major changes and changes due to the base flow (Magilligan 1992; Hafez 2000; Flores et al. 2006), the fourth interval sections and sections 9, 12, 8 and 13 will have the highest potential for cross-section deformation and shoreline deformation, and at different return periods the erosive strength of the river will be high, and sections 20, 19, 15, 18 and 17 will have the least potential. The specific power of the river in different return periods based on monthly peak and maximum instantaneous flow for sections is shown in Tables 5 and 6, respectively. According to the indices of width, slope, and depth of water, the highest amount of shear stress was also in the second interval and in sections 12, 13, 9, 8,11 and the fourth interval sections, and the lowest shear stress was in sections 19 and 20. Therefore, the specific power and shear stress are directly related.

One can say that the existence of vegetation in the riverbed and along the river increases the roughness of the flow path, wastes water energy, slows the flow, and reduces the shear velocity and stress of the flow. In other words, the water energy is dissipated at the site of the wall by the aerial organs of the plant. The plant stems through which the flow passes bear the highest tensile force and as density increases, the roughness coefficient also increases. Hydraulic resistance also causes the suspended sediments along the riverbed to be absorbed and deposited, thus controlling the width and increasing the stability of the walls. In the fourth interval, where the level of vegetation is very low, the shear stress is also high; this is while in parts of the river path with dense vegetation, the vegetation acts as a protection and prevents erosion of the margins, reducing shear stress (Van De Wiel & Darby 2007; Eaton & Millar 2017; Zhu et al. 2018).

According to the results, it can be said that with the construction of Givi Dam, the fundamental changes in river flow regime are in the form of decreased downstream discharge in a high water season, increased downstream discharge in a low water season, downstream reduced sediment, sedimentation of suspended matter, and increase in riverbed level. This causes the natural balance of the river to be changed. These changes have also affected the geometrical and geomorphological features of the river.

This study shows that the sections and regions with fine-grained and alluvial bed materials are more sensitive to discharge changes and sediment load; generally, in the Givi-chay River, in the sections that the river passes from alluvial barracks, the riverbed is unstable. The most important factor in the meandering of the river, in the plain region where meanders are free, is its alluvial formation type and low slope. In contrast, in the mountainous regions, the river changes follow the valley changes.

Considering the parameters of width, depth, and slope of the river and calculation of total power, specific power and shear stress, the maximum amount of total river power is in the second interval, in sections 11, 12, and 13. In terms of river-specific power, the fourth interval sections, and the 9th, 12th, 8th, and 13th sections, have the most potential for deformation and coastal deformation, and at different return periods the erosive power of the river will be high, in the 20th, 19th sections, while sections 15, 18, and 17 have the least potential. The highest amount of shear stress was observed in the second interval, at 12, 13, 9, 11, and the lowest at 19 and 20.

In particular, the changes in the Givi-chay River plan are the developing of existing meanders, river route displacement, increasing curvature, and creating of small meanders. Pattern formation and morphological changes of the Givi-chay River are affected by hydrological processes induced by the construction of the Givi-chay dam, lithological resistance of the riverbed and river banks, and human interventions such as encroachment on land, the riverbed, bridge construction, and sand mining. In general, river management and water resource strategies should consider the styles and magnitude of channel changes, to avoid or mitigate their adverse effects on present and future human activities. In order to adopt principled methods and sustainable environmental management, the results of the present study can be considered by the Regional Water Organization, Watershed Management and Natural Resources, farmers and residents around the Givi-chay River.

The authors would like to thank the regional Water Company of Ardabil for providing the hydrology data for this study. We also acknowledge the support from Mohaghegh Ardabili University.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.