1 Introduction

Deltas rank among the most economically and ecologically valuable environments on Earth (Nienhuis et al. 2020). With increasing anthropogenic influence, many of large deltas face severe pressure from urbanization, land-use changes, socio-economic transformation, climate change, and episodic natural hazards (Giosan et al. 2014; Musa et al. 2014; Lin et al. 2020). At present, sediment fluxes in many of the world’s large rivers have been drastically reduced (Syvitski et al. 2009), which is affecting delta morphology and may lead to erosion (Syvitski et al. 2009; Tessler et al. 2015). Deltas are increasingly vulnerable to coastal hazards with expected accelerated sea-level rise (Nerem et al. 2018). A better understanding of how deltas grow can help address these threats and challenges.

The Yellow River is the most sediment-laden river in the world, with an average suspended sediment concentration of 35 ~ 40 kg/m3 (Ren 2006; Milliman and Farnsworth 2011). The lower reaches of the Yellow River have become a suspended river on the ground due to the siltation, with an elevation even be more than 10 m in some sections of the river (Xie 1999; Yu 2002). The river with high suspended sediment concentration breaks through the embankment, causing the lower reaches and estuaries to swing and shift frequently. In the past 2600 years, there have been 26 major shifts in the lower 600 km long section of the Yellow River (Wang and Liu 2019). Two of the larger shifts of the Yellow River in the last 1000 years occurred in AD1128 and AD1855, respectively. The former was caused by anthropogenic dike excavation, from flowing north into the Bohai Sea to joining the Huaihe River and eventually fully merging into the SYS, whereas the latter was a natural shift, with the Yellow River returning to its former course and re-entering the Bohai Sea.

The massive accumulation of sediments brought by the southward shift of the Yellow River over 700 years formed the AYRD at the former Huaihe River mouth. The delta cannot grow without sediment supply. For a long time, the sediment flux to the sea of the Yellow River has been influenced by both human activities and climate change. The sediment flux reaching the sea after the seventh century AD was basically increasing, reaching its peak around AD1855 (Wu et al. 2020), which provided abundant sediment supply for the growth of the AYRD. In addition, during its growth, annual sediment load of the abandoned Yellow River had been altered accordingly by a series of anthropogenic activities such as levee breaches, embankments, and river management in the basin. The sediment flux of original Huaihe River into sea was 100 ~ 200 million tons per year (Zhang 1998), whereas there was a large sediment flux reaching the sea after the Yellow River joining the Huaihe River. It is estimated that during the first 300 years of the Yellow River southward shift (AD1194 ~ AD1495), annual sediment load of the abandoned Yellow River channel was about 0.6 ~ 1.0 billion tons per year (Zhang 1998; Ren 2006) of which about 40% ~ 57.6% (Chen et al. 2018) accumulated on the river channel and the alluvial plain, and the rest of the sediment sank into the sea. During the next 360 years (AD1495 ~ AD1855), the annual sediment load of the abandoned Yellow River channel increased to about 1.0 ~ 1.6 billion tons per year due to the embankment of the north bank of the river and the closure of the channel on the north side of the river (Zhang 1998; Ren 2006), correspondingly enhancing the sediment flux into the SYS. The sediment flux of the abandoned Yellow River here for more than 700 years has been influenced by a combination of natural and anthropogenic factors, thus also influencing the growth process of the delta.

Due to the exploitation of oil–gas resources or the overall development program in the Yellow River Delta, the study of the Yellow River Delta located in Bohai Bay has made great progress. Many scholars have conducted thorough studies on the evolution of the Yellow River Delta, sedimentation patterns, depositional environments, and the geological aspects of delta erosion and disasters (Bornhold et al. 1986; Xue et al. 1995; Saito et al. 2000; Liu et al. 2014, 2019; Wu et al. 2015; Zhang et al. 2018). In contrast, the sedimentary evolution of the AYRD remains more questionable due to fewer long-length cores and lack of precise chronological framework in the terrestrial part of the AYRD. What are the characteristics of the growth-evolution process of this delta? And what are the main factors influencing the growth of delta in different stages. The choice of the AYRD, a representative delta of growth, is conducive to further understanding of delta growth processes by exploring the evolution of the delta in time and space.

2 Materials and methods

2.1 Study area

During more than 700 years from 1128 to 1855 AD, the mouth of the Yellow River shifted southward to the SYS, forming the AYRD along the Jiangsu coast. The terrestrial AYRD is located in the north Jiangsu Provence, and its apex is located in Yuntiguan site, covering an area of about 7000 km2 (Zhang 1984). The SYS is mainly influenced by the semidiurnal tides, with a tidal range of about 1 ~ 4 m on average (Liu et al. 2013). The western SYS is dominated by a convergent-divergent tidal current field centered on the Jianggang in the southern part of the study area (Li et al. 2001).

The study area and the distribution of sampling stations are shown in Fig. 1. In order to study the development and evolution history of the AYRD, two cores (X02 and X03) were collected on the left and right sides of the mouth of the original Huaihe River (outside of Yuntiguan site, which is the apex of the terrestrial AYRD), and one core (X04) was collected on the front edge of the terrestrial AYRD. Borehole cores were obtained by the rotary drilling rig method in 2019, and their recovery rate was basically above 95%. The segmented cores are extended proportionally to the planned collection length to eliminate the effects of compression. The information about the cores is shown in Table 1.

Fig. 1
figure 1

Location of study area and cores distribution

Full size image
Table 1 Statistics of cores information in the study area
Full size table

2.2 OSL dating and the development of dating framework

Eighteen samples were collected from three cores for Optically Stimulated Luminescence (OSL) dating. All OSL sample preparations and measurements were performed under dim red light in order to avoid optical bleaching effects. After removing the outer layers of the samples, the inner cores of the samples were treated with 30% hydrogen peroxide (H2O2) and 10% hydrochloric acid (HCl) in order to remove organic matter and carbonates, respectively. After treatment with H2O2 and HCl, the raw sediment samples were then rinsed and wet-sieved to retain the grain-size fractions of 45 ~ 63 μm. The grains were then etched using 30% H2SiF6 for 3 ~ 4 days and 40% HF for 40 min, respectively, and then washed with HCl and water to isolate the quartz grains. The purity of the isolated quartz was checked by infrared stimulated luminescence measurements to ensure no feldspar contamination was present in any of the samples (Duller 2003).

All luminescence measurements, beta irradiation and preheat treatments were carried out using an automated Risø-TL/OSL DA-20 DASH reader equipped with a calibrated beta (90Sr/90Y) source with an EMI 9235 QA photomultiplier tube (PMT). Neutron activation analysis (NAA) was used to determine the uranium, thorium and potassium contents of these samples. The water contents (weight of water/weight of dry sediment) of these samples were measured in the laboratory. The dose rates (DR) and final ages were calculated using the Dose Rate and Age Calculator ('DRAC0) (Durcan et al. 2015).

The single-aliquot regenerative-dose (SAR) protocol (Murray and Wintle 2000) was applied to aliquots of grains that were 2 mm in diameter in order to determine the equivalent dose (De) of the samples. The growth curves were fitted using single saturating exponential functions. All OSL ages are reported relative to AD2019. The experiments were performed at the State Key Laboratory of Estuarine and Coastal Research, East China Normal University, Shanghai.

T = \(\frac{{D}_{e}}{DR}\) , where T is the sediment burial time, and De is the equivalent dose, and DR is the annual dose rate.

The Bacon model is a stepwise autoregressive gamma procedure based on the Bpeat model concept by Blaauw and Christen (2005, 2011). An optimal age-depth curve is finally obtained with 95% confidence interval. The Bacon procedure uses Markov Chain Monte Carlo (MCMC) for iterative operations, which can greatly reduce the age error range with higher reasonableness and accuracy (Blaauw and Christen 2005).

2.3 Grain size analysis

Grain size analysis of 2 cm sub-samples was conducted using the Mastersizer-2000 laser particle analyzer, which has a measurement range of 0.02 ~ 2000 µm with a relative error of < 3% for repeated measurements. The experiment was completed at the Key Laboratory of Coast and Island Development, Nanjing University, Nanjing. The matrix formula of McManus (1988) was used to calculate the sample statistics of the grain size distribution, that is, mean grain size (Mz). The sand-silt–clay triangle diagram proposed by Folk et al. (1970) was used to classify the sediments.

3 Results

3.1 Lithostratigraphic characteristics

Core X02, 1520 cm long, mainly appears grey-brown, dark grey, brown-grey and green-grey. The composition is highly variable, with mainly clay, clayed silt and silty sand. It can be roughly divided into four sections according to lithology (Fig. 2).

Fig. 2
figure 2

Geologic column of cores X02(a), X03(b), and X04(c)

Full size image

0 ~ 400 cm: grey-brown, clayey silt with high water content, soft, strongly cohesive and plastic. There are black carbonaceous spots at 110 ~ 112 cm and 113 ~ 117 cm, and scattered shell fragments at 110 ~ 111 cm.

400 ~ 850 cm: greyish brown, silty sand with low water content, dense, non-cohesive, non-plastic.

850 ~ 1350 cm: mainly brownish grey and dark grey, clayey silt with medium water content, soft, weakly cohesive, plastic. The 1320 ~ 1330 cm section is interspersed with many scattered shell fragments, ranging from 1 to 5 mm in diameter.

1350 ~ 1520 cm: greenish grey, stiff clay layer with a lot of concretions, ranging from 0.5 to 8 cm in diameter and with brownish yellow rust spots; very low water content, slightly consolidated, strongly cohesive but not plastic.

Core X03, 1520 cm long, is mainly brown and brownish-grey, silty to clayey silt. It can be divided into two sections according to lithology.

0 ~ 245 cm: mainly brownish-grey and greyish-brown, clayey silty sand with low water content, soft, strongly cohesive and plastic. There are bioturbation phenomena at 142 ~ 145 cm and 207 ~ 215 cm.

245 ~ 1520 cm: mainly brownish-grey and grey silt with medium–high water content, soft, strongly cohesive and plastic. There is bioturbation at 420 ~ 440 cm and scattered shell fragments at 443 cm. It is intercalated with grey-black, high water thin (5 ~ 20 mm thick) sand bandings or lenses at 1380 ~ 1420 cm.

Core X04, 2060 cm long, shows mainly brownish and brownish-grey silt and sand, with black carbonaceous spots or bandings at 120 ~ 400 cm. It can be roughly divided into two sections according to lithology.

0 ~ 960 cm: brownish and brownish-grey silt with low water content, soft, weakly cohesive and plastic. Black carbonaceous spots or bandings occur in many layers, such as 125 ~ 128 cm, 177 ~ 178 cm, 188 ~ 193 cm, 197 ~ 198 cm, 201 ~ 205 cm, 391 ~ 392 cm and 865 ~ 866 cm. Clayed silt or clay bandings occur in many layers, for example, in 683 ~ 691 cm and 943 ~ 944 cm.

960 ~ 2060 cm: brownish-grey sand with high water content, dense, almost non-cohesive and non-plastic.

3.2 Dating framework and deposition rate

Six samples of core X02 at 2.55 m, 6.70 m, 10.55 m, 12.70 m, 13.48 m, and 15.00 m were selected, respectively. The ages of the samples were 0.56 ± 0.07 kyr, 0.80 ± 0.11 kyr, 0.80 ± 0.05 kyr, 1.09 ± 0.13 kyr, 19.64 ± 0.07 kyr, 33.80 ± 4.40 kyr (Table 2), with no age inversions. However, the dating results of X02 show that the upper and lower sections are distinct at 13 m depth: the ages of upper section were over the past 1000 years; the ages of lower section were during the 34 ~ 20 kyr period, indicating that the deposition rate of cores in this section is extremely low with a 15 cm-thick erosional bed here (Liu et al. 2010a) separating it from the overlying unit. In order for the Bacon model to run better and thus obtain an accurate dating framework for the last millennium, only four samples above 13 m depth were selected for the core X02 for dating simulations. The Bacon running result contains the 95% confidence ranges, median and mean ages for each depth, and red curve shows single "best" model based on the mean age for each depth (Fig. 3). So, we choose the mean age as the value for the dating framework and use it to calculate the deposition rates. It was found that the sediment age above 13 m in X02 ranged from AD941 (± 125) to AD1573 (± 134), and after AD1573 (± 134), it almost stopped receiving sediment supply.

Table 2 Optically stimulated luminescence dating results of cores X02, X03, and X04
Full size table
Fig. 3
figure 3

Dating framework of core X02, X03, and X04. The grey dotted line indicates the 95% confidence interval; red curve shows the single “best” model based on the mean age at each depth

Full size image

Similarly, the dating framework was established for core X03 and core X04. The chronological depositional sequence for core X03 ranges from AD983 (± 108) to AD1784 (± 94); the chronological depositional sequence for core X04 ranges from AD114 (± 289) to AD1856 (± 128).

Based on the established dating framework, relatively unique layers were selected in each core to analyze deposition rate variations in these sections (Table 3). It was found that deposition rates in core X02 ranged from 2.0 to 2.3 cm/yr, in X03 from 1.7 to 2.0 cm/yr, and in X04 from 1.0 to 1.4 cm/yr (Fig. 4). The overall performance is an increase in deposition rate after the Yellow River southward shift, with a larger rate in the early stages and a slightly lower rate in the later stages.

Table 3 The ages in unique layers and the deposition rates of cores X02, X03, and X04
Full size table
Fig. 4
figure 4

Variation of vertical deposit rate in the Abandoned Yellow River Delta. The vertical dotted lines mark the years AD1128 and AD1855, and the shaded areas represent the period when the clay content began to increase

Full size image

3.3 Grain size characteristics of sediments

In conjunction with the dating framework established in the previous section, changes in the grain size fraction of the sediments are observed over time. The results show a significant shift in the sediment sequence of the AYRD due to the fine-grained sediments of the abandoned Yellow River source. After AD1128, the grain size of the sediments becomes finer: the mean grain size distribution of X02 ranges from 5.7φ to 6.8φ, with a mean value of 6.6φ; the mean grain size distribution of X03 ranges from 5.7φ to 8.1φ, with a mean value of 6.9φ; the mean grain size distribution of X04 distribution ranges from 5.9φ to 6.6φ, with a mean value of 6.3φ (Fig. 5).

Fig. 5
figure 5

Stages change of sediments grain size from the Abandoned Yellow River Delta. a Core X02; b Core X03; c Core X04. The dots (connected by red dashed lines) indicate the year when the sand fraction began to decrease at the beginning of the Yellow River joining Huaihe River, i.e., AD1135 in core X02; the purple dashed lines indicate the AYRD entered a period of rapid development with an increase in the clay fraction, the first occurring in the middle to late 14th century and the second in the mid-16th to the mid-17th century, and that the second occurrence may be related to the increase of sediment flux; the pentagons (connected by blue dashed lines) indicate the final land-forming stage of the AYRD with an increase in the sand fraction. Below the red dashed line is the pre- stage, between the red dashed line and the purple dashed line (the below one) is the growth stage I, between the purple dashed line (the below one) and the blue dashed line is the growth stage II, and above the blue dashed line is the growth stage III

Full size image

In addition, the grain size fraction of the sediments changes in stages, with a general decrease in the sand fraction and an increase in the clay fraction, particularly in core X03 and X04. The depositional sequence of core X02 is somewhat unique in that it experienced a brief decrease in sand content after AD1135, followed by a sharp increase in sand content perhaps due to the scouring of riverbed sand. There were two increases in clay content in cores X02 and X03, in AD1376, AD1513 and AD1362, AD1637, respectively, whereas an increase in clay content occurred for X04 in AD1362. The first time of clay content increase is close in three cores which indicate the AYRD entered a period of rapid development. The second clay content increase only occurred in core X02 and X03 with time difference which may related to the increase of sediment flux. Although there are some minor differences, the three cores share common features with changes in three steps: (i) a sudden decrease in sand content just after the Yellow River joined the Huaihe River; (ii) a sudden increase in clay content in the sediments in the mid-late 14th and mid-16th to mid-17th centuries; and (iii) a slight increase in the proportion of sand in the sediments during the final stage of deltaic land-forming which continued for about 10 ~ 90 years.

Although the timing of the sedimentary record differs slightly from the event of the southward shift of the Yellow River (AD1128 ~ AD1855), the sedimentary sequences of three cores essentially show sedimentary changes during the growth of the AYRD, beginning as early as AD1135 (X02) and ending as late as AD1856 (X04).The timing of the sedimentary record is basically slightly delayed with the outward extension of the estuarine location, and also basically reveals the seaward extension of the AYRD.

The grain size fraction changes significantly over time, and the sediment sequence can be divided into four stages, before the shift of Yellow River (pre- stage), Delta growth stage I, II and III, with a corresponding change in sediment type. After the Yellow River southward shift, there was a basic change from two types of silty sand and sandy silt to two types of silt and mud (Fig. 6). The sediment types in the AYRD are mainly silt, mud and, sandy silt.

Fig. 6
figure 6

Changes in sediment type of Abandoned Yellow River Delta in recent 1000 years. a X02; b X03; c X04

Full size image

4 Discussion

4.1 Growth process of the Abandoned Yellow River Delta

A study of the coastline along the Jiangsu-Shanghai area in the last millennium revealed that X02, X03 and X04 have undergone a sea-to-land transition in that order (Fig. 7). Based on the precise dating framework established in this study, it can be found that the ages of the tops of cores X02, X03, and X04 are AD1573 (± 134), AD1784 (± 94) and AD1856 (± 128) in that order, which also implies a sequential postponement of the time of land formation. The timing of land formation in each core coincides with the shoreline migration in the AYRD, which laterally confirms the reliability of the dating framework.

Fig. 7
figure 7

Relationship between the shoreline changes of Jiangsu and Shanghai in the past thousand years (according to Tan 1982) and the location of the cores used in this study. The land-formation times of cores X02, X03 and X04 are AD1573 (± 134), AD1784 (± 94) and AD1856 (± 128) in that order, which coincide with the shoreline change process of the AYRD, laterally confirming the reliability of the dating framework

Full size image

From AD1128 to AD1855, large amounts of sediment accumulated at the estuary of the abandoned Yellow River to form the AYRD. The development process of delta is the transition from sea to land, which is mainly the deposition of sediments outside the estuary, with the water depth becoming increasingly shallow and eventually transformed to land. The sea-to-land transition is a process of sediment filling the negative topography in the vertical direction. According to the depth-age framework established by each core, the deposition rate of cores in each period was analyzed (Fig. 4). It was found that X02, which is in up-estuary, has the largest deposition rate, almost always above 2 cm/yr, with the fastest deposition rate of up to 2.29 cm/yr between AD1128 and AD1135. The deposition rate of X03 is basically between 1.71 and 1.97 cm/yr, and the deposition rate is basically greater than 1.96 cm/yr between AD1268 and AD1495. The deposition rate of X04 is relatively slow and variable, basically less than 1.4 cm/yr. Before AD1128, the deposition rate is less than 1.24 cm/yr. Between AD1128 and AD1495, the deposition rate increases continuously, from 1.29 cm/yr to 1.38 cm/yr, but decreases slightly after AD1495. The deposition rate varies at different locations from the estuary to the sea. In general, the deposition rate decreases seaward, mainly as a result of the land-derived detritus brought by the river filling up-estuary first (Wang et al. 2013).

During the growth of the AYRD, the sediment flux into the sea is constantly changing, so the growth rate of the AYRD is also constantly changing. As far as the research data are concerned, the Yellow River had two major channels to the sea before the 16th century, one was along the original channel to the Bohai Sea, and one joined the Huaihe River to the SYS, thus the sediment flux accounted for half of the total sediment of the Yellow River, and the annual sediment flux into the sea was 260 ~ 600 million tons (Zhang 1998; Ren 2006; Xue et al. 2011; Chen et al. 2018). Whereas after the 16th century, due to the construction of dikes on the northern bank of the river, the northern channel was cut off and the situation of north–south flow into the sea ended, from then on, all of Yellow River sediments were remitted to the Huaihe River basin, and the annual sediment flux into the sea was 430 ~ 960 million tons (Zhang 1998; Ren 2006; Xue et al. 2011; Chen et al. 2018). However, the rate of vertical deposition during the delta growth is inconsistent with the sediment flux from the abandoned Yellow River into the sea.

The rapidity of delta development is expressed horizontally as the rate of seaward extension of the shoreline. Between AD1194 and AD1578, the rate of seaward extension was 33 m/yr, whereas between AD1579 and AD1591, the extension rate increased to 1540 m/yr (Table 4), an increase of two orders of magnitude, mainly due to the increase in sediment flux to the sea during this period as a result of river regulation by Liu Daxia and Pan Jixun. Between AD1592 and AD1855, the rate of seaward extension of the AYRD slowed to around 205 m/yr, but was much higher than the rate of extension of 33 m/yr between AD1194 and AD1578, an increase of almost one order of magnitude in comparison.

Table 4 Extension of the abandoned Yellow River estuary (From Ye 1986)
Full size table

The rate of horizontal land formation in the AYRD is also consistent with the sediment flux from the abandoned Yellow River into the sea. The land-forming rate was 3.2 km2 /yr between AD1128 and AD1500, and it was almost 4 ~ 7 times higher between AD1500 and AD1855 (Table 5).

Table 5 Rate of land formation in the Abandoned Yellow River Delta (From Zhang 1984)
Full size table

From the perspective of sedimentary geology, there are obvious incised valleys in the estuarine area of the Yangtze River, with the estuary gradually filling up to form the Yangtze Delta (Hori et al. 2002). There are no large incised valleys outside the abandoned Yellow River estuary in northern Jiangsu, which has been a depressional receiving basin since the Cenozoic (He 2006) and was a trumpet-shaped estuary until the Yellow River southward shift, with wider and deeper estuaries, up to 7 ~ 8 km at the widest section (Zhang 1984; Li 1991). Before AD1578, the river channel was wide and the runoff was weak due to the river management strategy of “wide river and fixed dike” (Xu 2001), and the estuary was in an "unsaturated" state for a long time, which provided a wide space for sediment accumulation. Therefore, sediments from the early diversion period were deposited in the estuary area and mainly filled the estuary (Fig. 8). After AD1578, the river channel became narrower due to the river management strategy of “restrain water and attack sand”, and the sediment increased, runoff was enhanced. The estuary was largely filled, so that the sediments entering the sea spread out under the influence of strong run-off, and the rate of seaward shoreline extension and land formation increased. Therefore, the horizontal growth of the delta tends to be more influenced by sediment supply, whereas the vertical deposition rate tends to be more influenced by the original topography of the estuary.

Fig. 8
figure 8

Schematic model of the development of the Abandoned Yellow River Delta. It represents the accumulation process of sediment brought by the abandoned Yellow River, with the center of deposition continuously advancing offshore. Before the 16th century, there was less incoming sediment and the rate of shoreline extension seaward was slower, while the rate of proximal vertical accretion was faster, and the amount of sediment transported offshore was less. After the 16th century, there was relatively more incoming sediment, and the rate of shoreline extension seaward was faster, while the rate of vertical accretion became slower, and the amount of sediment transported offshore increased

Full size image

Fluvial deltas are defined as accumulations formed by the seaward material of a particular river, and thus the distal mud is considered to be an important component, and the distal mud depositional characteristics also reveal the stages of river delta evolution (Liu et al. 2014; Jia et al. 2018). Typically, fluvial deltas evolve in the sequence of estuarine bay, estuarine delta, subaqueous delta, and distal mud of the shelf. After the Holocene high sea level period, the estuarine bay was first filled, then the estuarine and subaqueous deltas were deposited, after which the sediments overflowed massively from the estuary and were transported to more distant sites by the alongshore of the shelf circulation, forming distal mud deposits (Liu et al. 2014). The main body of the Yangtze Delta began to form around 6 kyr B.P. (Hori et al. 2001; Gao 2007). For the first 4,000 years, the sediments of Yangtze River were deposited within the large estuary to fill the estuary itself. For the next 2,000 years, sediments escaped from the estuary and spread out onto the open shelf, stretching more than 600 km southward to form the distal mud. The Yellow River delta near Tianjin began to form around 2.34 ~ 5 kyr B.P. (Ren 2006), and the development of the distal mud about 500 km away from the estuary began almost simultaneously with the delta near the estuary due to lack huge estuarine bay. The place where the AYRD developed also did not have a huge estuarine bay, and the sediment supply during its growth period once reached the highest sediment load in the history of the Yellow River, about 0.6 ~ 1.6 billion tons per year (Ren 2006). Except for some sediments deposited in the alluvial plain inside and outside the river bank, the annual sediment flux to the sea was about 260 ~ 960 million tons (Ren 2006; Chen et al. 2018), so the delta grows rapidly. The sediment type of subaqueous AYRD is mainly clayey silt (Liu et al. 2010a) with finer sediment grains, and some researchers have also referred to it as the western muddy zone of the SYS (Liu et al. 2014). However, considering its location and investigating its growth process, no distal mud has developed in the AYRD. The main reason is that this growth history lasted only for more than 700 years before the supply of sediments was interrupted.

In general, before the 16th century, the abandoned Yellow River carried large quantities of sediments that were deposited rapidly on the outer side of the estuary, with rapid vertical accretion and the fastest deposition rate near the estuary (X02), basically at 2 cm/yr. As the estuary was largely filled in the early stage, the later stage was dominated by horizontal land-formation, with the fastest seaward extension rate up to 1.5 km/yr. Thus, the rate of land-formation is not only related to the sediment flux, but also related to the original submarine topography outside the estuary (Yin 1986; Liu et al. 2020). The above results also basically confirm the theory that the rate of land formation is inversely proportional to the vertical deposition rate when the flux of fluvial sediment is constant.

4.2 Factors influencing the formation of the Abandoned Yellow River Delta

Sediment supply is the material basis for the evolution of delta growth, and numerous cases of delta evolution show that the magnitude of supply not only determines the scale of delta development, but also affects delta morphology (Nienhuis et al. 2020; Rao et al. 2020). The removal of two dams on the Elwha River in the United States between 2011 and 2014 caused the erosion of 30 million tons sediment trapped in the reservoir, causing the Elwha Delta to grow by 0.6 km2 over a five-year period (Ritchie et al. 2018). The AYRD grew by 7000 km2 over 700 years (Zhang 1984), a growth rate almost a hundred times faster than that of the Elwha delta in comparison, which is clearly attributed to an adequate supply of sediment. The Krishna Delta on the east coast of India has changed its morphology from a wave-dominated cuspate delta to an outbuilding lobate delta over the last 500 years due to increased sediment flux to the sea (Rao et al. 2020). These cases illustrate that when sediment supply is large enough, it can mask the shaping effects of tide and wave dynamics on the geomorphology of the delta to some extent.

Adequate sediment supply is required for the growth stage as well as for the geomorphic maintenance of the delta. Under the background of the decrease of sediment flux into the sea, it is a major trend for the evolution of delta to transform from river dominated delta to tide dominated delta and wave dominated delta (Nienhuis et al. 2020). Even the Magra Delta, a small river dominated delta along the Mediterranean coast, is gradually assuming a typical estuarine morphology due to sediment depletion (Pratellesi et al. 2018). The role of marine forces such as tide and wave on the evolution of deltaic landforms began to come to the fore when the river-borne sediments decreased. The marine forces changed the geomorphology of the delta by adjusting the redistribution of incoming sediments. After AD1855, the mouth of AYRD eroded and retreated (Liu et al. 2010b), and the shoreline gradually flattened due to the return of the Yellow River to the north and the cut-off of sediment supply.

The slope of the underwater topography affects the amount of sediment deposited in the estuary. During the period of Yellow River southward shift, about 260 ~ 960 million tons of sediment entered the sea each year (Ren 2006; Chen et al. 2018). The large amount of sediment brought by the abandoned Yellow River continued to accumulate in the estuary and the sediment center continued to extend seaward (Fig. 8), eventually forming the AYRD. In contrast, the Ayeyarwady River estuary, where the modern Ayeyarwady River delivers about 360 million tons of sediment to the sea each year (Liu et al. 2013), has a large sediment flux to the sea, which is of the same order of magnitude as the sediment flux to the sea from the abandoned Yellow River. However, the results show that there is almost no modern sediment accumulation on the continental shelf near the Ayeyarwady River estuary, and huge amounts of muddy sediments are deposited in wedges in the Martaban Gulf (Kuehl et al. 2019), with a thickness of 60 m, extending seaward to about 130 m water depth (Liu et al. 2020). The main reason is that the difference of two submerged slope outside estuaries is too large (Liu et al. 2013, 2020). The excessive submerged slope is not conducive to the massive accumulation of sediments, thus making it difficult to form large deltas.

Changes in sea level can affect the location of sediment accumulation, or the “tolerable space” for sediment accumulation (Wright et al. 1971), thus affecting the growth and evolution of deltas. The formation of the AYRD is an exceptional event in the last millennium, during which the sea level is basically stable (Wang et al. 2014), so its growth and evolution can basically exclude the influence of sea level change.

5 Conclusions

1. The dating framework shows that all three cores in the AYRD record sedimentary sequences for one thousand years. The time of land-formation from land to sea direction (X02, X03 and X04) is sequentially delayed of AD1573 (± 134), AD1784 (± 94) and AD1856 (± 128), respectively.

2. The AYRD is dominated by fine-grained sediments, with silt, mud, and sandy silt as its main sediment types. After the Yellow River joined the Huaihe River in AD1128, the grain size components in the sediment were significantly transformed, which showed that the sand component began to decrease at first. In the middle and late 14th century, the proportion of clay component in the sediments increased and remained high. In the middle of the 16th century, due to the further increase in sediment flux from the abandoned Yellow River into the sea, the clay content in the sediments increased sharply in the following 100 years. During the final stage of the transition from subaqueous delta to terrestrial delta, a slight increase in the proportion of sand component is a distinctive feature in the sedimentary record, with this change lasting from 10 to 90 years in different core records.

3. The dominant factors affecting the evolution of the AYRD are the sediment flux into the sea and initial submerged topography. After the interruption of sediment supply due to the northern return of the Yellow River in AD1855, hydrodynamic erosion by wave and tidal forces from the outer delta began to dominate. Before the 16th century, the sediments were deposited mainly in the up-estuary and nearshore, with rapid vertical accretion. The deposition rate of the up-estuary (X02) was the fastest, basically greater than 2 cm/yr, and the deposition rate of the outer side (X04) was slower, at 1.2 ~ 1.4 cm/yr. After the 16th century, the horizontal land formation was the main focus, and the seaward extension rate increased rapidly, reaching 1.5 km/yr at the fastest. The development process of the AYRD reflects the following law: when the sediment flux of river into the sea is certain, the land formation rate is inversely proportional to the vertical accretion rate, and the faster the horizontal land formation rate, the slower the vertical accretion rate.