Deltaic plains have always exhibited complex interactions with climatic, sedimentary, and tectonic processes because they are formed during the transition between fluvial and maritime environments ( Orton & Reeding, 1993 ; Overeem, 2005 ). Also, deltas are highly susceptible to changes due to natural and human activities ( Syvitski & Saito, 2007 ). Since ancient times, deltas have been of great interest to humans due to their fertile lands, and their accessibility for transportation and trade ( Stanley & Warne, 1997 ; Fan et al., 2017 ). Therefore, deltas hold significant social, ecological, and economic importance ( Bergillos & Sanchez, 2017 ). Due to intense human interventions, the morphology of deltas worldwide has changed significantly from their natural state ( Fan et al., 2019 ). One of the most important human activities in recent centuries that have modified deltaic environments is damming, which has reduced the sediment load of large rivers over the past century ( Yang et al., 2006a ; Hood, 2010 ; Kondolf et al., 2014a ; Wang et al., 2011 ; Fan, 2018 ). By reducing the sediment load of rivers, damming causes instability in deltaic systems and promotes the erosion of coastal deltas ( Yang et al., 2003, 2006b ; Giosan et al., 2014 ). On the other hand, dams are key structures that affect sediment transport and increase sediment deposition in their reservoirs by reducing water flow velocity ( Syvitski et al., 2005 ). Sedimentation in dam reservoirs decreases their capacity and causes various economic problems ( White, 2001 ). Therefore, various methods are used to reduce reservoir sedimentation ( Kondolf et al., 2014b ). Sediment flushing operation (SFO) is the most common method regularly implemented to remove and reduce sediments stored in dam reservoirs ( Brown, 1944 ; Shen, 2010 ; Kondolf et al., 2014b ; Randle et al., 2015 ; Panthi et al., 2022 ), particularly the finest particles that accumulate near the dam ( Dahal et al., 2021 ). In SFO, deposited material in the reservoir is eroded by increasing water velocity and lowering the reservoir pool elevation through the release of water from the reservoir ( Brandt, 2000 ; Kondolf et al., 2014b ; Lai et al., 2024 ).

The Sefidrud Delta in the southern Caspian Sea has been chosen as the study area for the following reasons: i) Reports and evidence indicate changes in the Sefidrud Delta and its shoreline over the past few decades, with the last major evolution occurring around AD 1600 ( Lahijani et al., 2009 ); ii) The Manjil Dam was constructed on the Sefidrud River in 1962; and iii) Sediment flushing operations were conducted in the Manjil Dam reservoir between 1980-1981 and 1997-1998. While most recent studies have focused on shoreline changes in the Sefidrud Delta, much remains to be investigated regarding the morpho-sedimentary dynamics of the Sefidrud Delta. By dividing the Sefidrud Delta shoreline into different sections, Kazanci and Gulbabazadeh (2013) concluded that there is an imbalance between sediment discharge and deltaic progradation in the Sefidrud River. Haghani et al. (2016) identified rapid Caspian Sea-level rise between 1977 and 1995, along with sediments from the Manjil Dam flushing operation (1981-1998), as two main factors affecting the lagoons in the Sefidrud deltaic area. By analyzing 14 sections of the Sefidrud Delta shoreline using the Digital Shoreline Analysis System (DSAS), Toorani et al. (2021) concluded that the mismatch between Caspian Sea level fluctuations and changes in the Sefidrud Delta shoreline suggests other influencing factors. Other studies have mentioned various reasons for the evolution of the Sefidrud Delta or shoreline changes, including sediment accumulation in the Sefidrud channel, the direction of Caspian Sea waves, or the neotectonic movements of Caspian Sea faults following the 1990 Manjil Earthquake ( Kousari, 1986, 1992 ; Krasnozhan et al., 1999 ; Khoshraftar, 2005 ; Sarvar, 2008 ). The implications of river damming on deltaic system evolution have also been widely studied ( Ly, 1980 ; Stanley & Warne, 1997, 1998 ; Brandt, 2000 ; Le et al., 2007 ; Wang et al., 2007, 2011, 2017 ; Blum & Roberts, 2012 ; Bergillos et al., 2015 ; Bergillos & Sanchez, 2017 ; Li et al., 2017 ; Zaimes et al., 2019 ; Bussi et al., 2021 ; Li et al., 2021 ). These studies emphasize that river damming has significant morphological and ecological impacts ( Best, 2019 ; Dunn et al., 2019 ), one of which is a decrease in sediment load that leads to sediment starvation and erosion in deltas ( Meybeck et al., 2003 ; Wang et al., 2011 ; Chen et al., 2022 ). A global review by Syvitski et al. (2005, 2007, 2009) stated that over half of the world’s large rivers are affected by dams, leading to coastal erosion and delta shrinkage. However, compared to the effects of river damming on deltas, few studies have examined the short and long-term implications of sediment flushing operations on the morphological evolution of deltas. SFOs have generally been investigated in terms of their impact on water quality ( Kemp et al., 2011 ; Espa et al., 2016 ; Hauer et al., 2018 ), damage to aquatic life ( Grimardias et al., 2017 ; Panthi et al., 2022 ), sediment dynamics, and river characteristics ( Lepage et al., 2020 ). Pourafrasyabi & Ramezanpour (2014) stated that the turbidity of the Sefidrud River is primarily due to the SFO of the Manjil Dam. The sediments released from the dam concentrate and remain at the riverbed for several years, delaying the recovery of the river. Flooding events resuspend this sediment in the water column, increasing turbidity and affecting aquatic life ( Pourafrasyabi & Ramezanpour, 2014 ). Therefore, the SFO may have created a sedimentary legacy for the Sefidrud Delta. To analyze the impact of damming and SFO on deltaic systems, two primary data sets are needed for the hybrid approach of this study:

1) Long-term hydrometric data from the basin, including water discharge and sediment load. More complete and long-term data increases the ability to analyze the sedimentary characteristics of the deltaic basin and other coupled factors.

2) Identification of all morphological changes in the delta over time.

Remote sensing, with its ability to detect detailed changes through time (change detection), plays a crucial role in this analysis ( Ghanavati et al., 2008 ; Munasinghe et al., 2020 ). Satellite remote sensing has proven to be an effective technology for monitoring morphological changes due to its ability to provide spatially continuous observations ( Cracknell, 1999 ; Munasinghe et al., 2020 ). To investigate changes in the Sefidrud Delta caused by the dam and SFO, this study analyzed aerial photographs and satellite images (Landsat) from 1956 to the present, alongside water discharge and sediment load data, to examine the morphological evolution and its relationship with anthropogenic factors. This investigation is complex and significant in four ways:

i) Investigating deltaic systems is critical as these areas provide valuable insights into the interaction between fluvio-deltaic and marine sedimentation processes ( Trincardi et al., 2005 ; Bergillos & Sanchez, 2017 ).

ii) Evaluating the role of dams in sedimentation and delta development is challenging ( Hauer et al., 2018 ).

iii) Investigating the short- and long-term environmental effects of SFO is difficult ( Lepage et al., 2020 ).

iv) The extent to which sediment from upstream sources contributes to different delta components remains unclear ( Quang et al., 2023 ).

This study aims to address all of these aspects by applying a hybrid approach to examine the hydrological-morphological linkages in the Sefidrud Delta.

To detect changes in the Sefidrud Delta, aerial photographs from 1956 and 1971 were obtained from the Iran National Cartographic Center ( https://www.ncc.gov.ir ), and all available multi-temporal remote sensing data from Landsat MSS, TM, ETM+, and OLI satellite images (1972-2024) were acquired from the EROS Center ( https://earthexplorer.usgs.gov/ ) ( Table 1 ). After radiometric and atmospheric correction of satellite images using ENVI 5.6 software, three different techniques were applied:

1) Normalized Difference Water Index (NDWI): Used to detect changes in the Sefidrud Delta shoreline. The NDWI is commonly used to monitor changes related to water content in water bodies ( McFeeters, 1996 ). In this study, NDWI was applied to various satellite images as shown in Equation (1) to Equation (3) ( Yulianto et al., 2019 ).

$\text{NDWImss}=\left({\rho }_1-{\rho }_4\right)/\left({\rho }_1+{\rho }_4\right)$ (1)

${\text{NDWI}}_{\text{TM\&ETM+}}=\left({\rho }_2-{\rho }_4\right)/\left({\rho }_2+{\rho }_4\right)$ (2)

${\text{NDWI}}_{\text{OLI}}=\left({\rho }_3-{\rho }_5\right)/\left({\rho }_3+{\rho }_5\right)$ (3)

2) Tasseled Cap Wetness (TCW): Used to distinguish water and land areas within the delta, which is an effective spectral analysis method for this purpose ( Ouma & Tateishi, 2007 ; Mukhopadhyay et al., 2018 ). The TCW is presented in Equation (4).

$\begin{array}{c}\text{TCW}=\left(0.1509\times \text{Blue}\right)+\left(0.1793\times \text{Green}\right)+\left(0.3279\times \text{Red}\right)\\ +\left(0.3406\times \text{NIR}\right)-\left(0.7112\times \text{MIR}\right)-\left(0.4572\times \text{SWIR}\right)\end{array}$ (4)

3) Principal Component Analysis (PCA): Used to illustrate the evolution of the delta and for change detection during the study period. PCA is one of the most widely used and effective methods for detecting changes ( Cakir et al., 2006 ). The daily water discharge and sediment load observed between the 1961-1962 and 2017-2018 hydrological years at the Astaneh hydrological station were obtained from the Water Research Institute, Ministry of Energy of Iran ( Figure 2 ). The Manjil Dam plays a crucial role in preventing floods by storing water and reducing the sediment load from the Ghezel Ozan and Shahrud rivers ( Kazanci & Gulbabazadeh, 2013 ; Haghani et al., 2016 ). The maximum flood peak discharge is shown, with only one flood discharge recorded exceeding those observed before the construction of the Manjil Dam ( Figure 3 , Table 2 ).

The development and evolution of deltas are governed by a delicate equilibrium between fluvial sediment supply and various external factors ( Somoza & Santalla, 2014 ). Disruptions to this equilibrium can result in delta retreat and, ultimately, its disappearance ( Somoza & Santalla, 2014 ). This study examines the variations in water discharge and sediment load concerning the morphological changes of the Sefidrud Delta over three-time intervals between 1962 and 2024.

The construction of the Manjil Dam and its sediment flushing operations have been the primary factors influencing the morphological changes of the Sefidrud Delta over the past seven decades. By altering the hydrometric regime of the river, the Manjil Dam transformed the delta from a river-dominated bird’s foot into a wave-dominated cuspate shape, creating new shorelines and forming Kiashahr and Zibakenar lagoons. The large amount of sediment released during the flushing operation further transformed the delta, leading to the formation of a more pointy cuspate shape. Due to the similarity in sediment grain size distribution between the flushed sediments and the Sefidrud Delta (comprising clay, silt, and sand), the large volume of sediments released during the flushing operation has significantly contributed to the rapid expansion of the delta. The effect of the flushing operation on delta formation was so pronounced that, despite the rise in the Caspian Sea level, delta formation continued, and the delta’s morphology changed from a wave-dominated arcuate shape in 1981 to a fluvial and wave-dominated pointy cuspate shape by 1990.

The substantial volume of sediment supplied by flushing operations between 1980 and 1998 contributed to the rapid expansion of the Boujagh point bar sedimentary lobe, leading to a 45˚ river channel migration. The rapid progradation of the Boujagh point bar within the delta following the construction of the Manjil Dam highlights the contribution of upstream sediments to the development of various parts of the delta under the river’s altered hydrological regime. However, further investigation is required to understand all aspects of this process fully. The Kiashahr and Zibakenar coastal lagoons shrank significantly due to the influx of sediment. Additionally, the flushing operation contributed to the rapid formation of mouth bar complexes, resulting in the development of new shorelines and the formation of the Northern Kiashahr Lagoon. By the end of the flushing operation, the sediment load decreased by about 93.4%, leading to sediment starvation and erosion of approximately 1.166 km² of delta.

On the one hand, the Caspian Sea is non-tidal, allowing tides to be disregarded ( Farley Nicolls et al., 2012 ). On the other hand, the dominant wave direction in the Sefidrud Delta area is NE-SW. However, the morphological changes of the Sefidrud Delta do not align with this wave direction, indicating that sediment discharge from the Sefidrud River is the primary factor driving these changes ( Yamani et al., 2013 ). According to Figure 12 during the Manjil Dam flushing period, the Caspian Sea level (CSL) rose by approximately 2 meters, which would typically be expected to cause significant erosion and delta destruction. In contrast, due to the substantial increase in sediment discharge from the Sefidrud Flushing Operation (SFO), the delta expanded by approximately 26 meters per year ( Yamani et al., 2013 ). In comparison, during the same period, the Kura Delta in Azerbaijan (another delta in the southwestern Caspian Sea) experienced erosion of 10 to 15 meters annually due to the rise in the Caspian Sea level, with severe degradation observed in its eastern parts ( Yamani et al., 2013 ; EU4Environment, 2023 ). The findings of this research demonstrate that the most critical factor controlling the development or degradation phases of the Sefidrud Delta is the volume of sediments transported by the Sefidrud River.

While the decline in the CSL ( Figure 12 , after 1998) may have reduced the erosion rate of the delta, the reduction in fluvial sediment supply, coupled with the decline in water discharge ( Figure 2 , after 1998), has exacerbated the delta’s degradation.

Due to the growing demand for water and energy in Iran, the construction of Shahr Bijar hydropower dam on the junction of two main tributaries of Sefidroud (Zalaki River and Do-aban River) began in 2004. This dam with a 105 Mm³ reservoir capacity has aggravated the decrease of flow discharge and sediment load in the Sefidrud Delta since 2004 ( Figure 1(E) and Figure 2 ). Although the impact of this dam is much smaller compared to the Manjil Dam in reducing the water flow and sediment discharge of the Sefidrud River, it underscores the inadequate attention given to managing the sediment budget necessary for maintaining the Sefidrud Delta. The continuation of reduced sediment supply from upstream and a lack of adequate water discharge in the third time interval (1999-2024) has led the Sefidrud Delta to become a more flat, arcuate shape. The combined effects of reduced sediment supply and sea-level decline have driven the delta into a state of erosion and ecological degradation. Based on the stability observed between 1962 and 1980, the critical sediment demand for maintaining the Sefidrud Delta is approximately 14 Mt/yr. During the period from 1999 to 2024, the mean annual sediment load delivered to the delta was 3.16 Mt/yr, which is only 23% of the required sediment load to sustain the delta ( Table 2 ).

The contradiction between these coupled circumstances underscores that sediment supply is the primary factor driving the growth of the Sefidrud Delta:

Therefore, the government must effectively manage water discharge and sediment load from upstream and the dam to safeguard the delta. The evolution of the Sefidrud Delta over the past 70 years can undoubtedly be considered a symbol of the Anthropocene in the southern Caspian Sea. Suppose the flushing operation of the Manjil Dam is to be repeated. In that case, careful consideration must be given to its impact on the morphology of the Sefidrud Delta, as the active lobe of the delta is highly sensitive to changes in hydrological parameters. The significant impact of the Manjil Dam’s flushing operation on the Sefidrud Delta’s morphology underscores the need for long-term studies on its environmental effects, water quality, and the health of the region’s aquatic life. Flushing operations not only altered the morphological and hydrological characteristics of the Sefidrud Delta but also caused widespread environmental damage by filling coastal lagoons and shifting the river’s main channel.

The evolutionary process of the Sefidrud Delta underscores the critical role of sediment budget alterations as the primary factor shaping delta formation, influenced by damming, flushing operations, and sea-level fluctuations over the past seven decades. Rapid changes in the Sefidrud Delta morphotype and alternating constructive-destructive phases make it a representative example of the degradation caused by human-induced interventions on deltaic systems. Although the hydrological regime and morphological type of the Sefidrud Delta were initially impacted by the Manjil Dam, sediment supply for delta development was maintained. Therefore, the delta continued to grow by forming new mouth bars. However, after the cessation of sediment flushing operations, the necessary sediment budget for delta sustainability was no longer supplied, resulting in sediment starvation and erosion.

The formation of mouth bars has been a critical indicator of the Sefidrud Delta’s development, as delta growth has consistently been linked with the formation of mouth bar complexes, particularly after the construction of the Manjil Dam and the completion of the flushing operation. Since 1999, due to the lack of sufficient water discharge and sediment load, no new mouth bar complexes have formed.

While many deltas globally are eroding due to sea-level rise, the Sefidrud Delta has been severely impacted by human mismanagement and a drastic reduction in sediment supply, regardless of sea-level decline. Thus, effective management of the water discharge and sediment load of the Sefidrud River is critical for the preservation of the delta. The hybrid approach used in this research, combining continuous monitoring of physical changes in the delta through remote sensing and hydrological data, provides a practical methodology for studying deltaic systems worldwide. This approach can help detect both short and long-term changes in response to natural and anthropogenic factors that alter the hydrological regimes of deltaic systems. Expanding this model to other deltas can help simulate the time evolution of river delta formation processes and understand the compounded effects of destructive and constructive factors in deltas worldwide.

We believe that the current vulnerable state of the Sefidrud Delta is threatened by the continued reduction in fluvial sediment supply and the decline in the CSL. Excessive dam construction on the Sefidrud River’s main tributaries after 2004 has further exacerbated the reduction in water discharge and sediment transport capacity. As a final point, the legacy of the Manjil Dam and its flushing operation must be carefully investigated to fully understand its negative impacts on water quality and the destruction of habitats in the region.

My acknowledgment goes to Engineer Ahmad Najafi, director of the Remote Sensing Academy of Iran, and Professor Omid Abdi for their support during this study.