Field work was conducted in south Andaman and Swaraj Dweep ( Figure 1 ; Table 1 ) in May 2023 (10 days) and Feb 2024 (10 days). Systematic coastal mapping was conducted along twelve beaches ( Table 1 ; Nine – south Andaman, three – Swaraj Dweep). Garmin GPS Montanna 680 was utilized to mark and map different morphological features along the different beaches (terraces, microatolls, rocky shoreline/grainy shoreline). Hand samples (loose/unconsolidated sediments) and cores (consolidated limestone) were collected from the beaches to further identify and map the coastal facies within the intertidal zone and shallow marine environments up to the fringing reefs.

20 g sample from the hand samples collected in field were dried and wet sieved and the mass loss after drying was calculated. Further, different sieves with varying mesh sizes: >2 mm, >0.500 mm, >0.250 mm, >0.125 mm, and >0.063 mm were used to separate the sample into different size fractions. Additionally, some thin-sections were prepared by Continental Labs (Thin section preparation lab in Uttar Pradesh, India), and studied under a Leica DM750 polarizing microscope. In this study, we focused on thin sections containing grain sizes larger than 0.500 mm diameter (to identify carbonate components). Sediment/rock composition has been classified based on texture (based on the matrix and grain percentage; Dunham, 1962). Here, sediments from the shoreline were classified depending on their mud content: –<10% grainstone, 10 – 26% packstone to grainstone, 26 – 42% packstone, 42 – 58% wackestone to packstone, 58 – 74% wackestone, 74 – 90% mudstone to wackestone and >90% mudstone (Petrovic et al., 2022; Menon et al., 2024). Further, in situ coralgal massive structures were categorized according to (Embry and Klovan, 1971), providing a more precise facies classification of all allochthonous and autochthonous component. Facies map for the south Andaman and Swaraj Dweep were created using QGIS 3.34.6 LTR^©.

This study utilized a custom X8 configuration multi-rotor drone (Bajrang Drone Series), referred to as a UAV from here-on, equipped with 20 MPX camera to collect visible RGB imagery over the coastal area and shallow (< 1 m) water zones within south Andaman and Swaraj Dweep. The UAV was equipped with a gimbal to keep the camera in a horizontal position at all times, thus reducing the data acquisition errors.

Five UAV surveys were conducted to collect >1500 digital photographs covering five areas, that are Badabalu Beach and Burmanallah in south Andaman (Loc. 4 and 7, Table 1 ), and Radhanagar, Kalapathar, and Govindnagar Beach in Swaraj Dweep (Loc. 10,11, and 12 – Table 1 ). The UAV was flown at an above-ground height of ~80 m over the area, with the camera in nadir orientation with a 90% overlap.

Agisoft Metashape Professional 1.8.4 was used to reconstruct dense 3-D models resulting in five photogrammetric digital outcrop models (e.g., Khanna et al., 2020). These were meshed and textured (by back-projecting the original image data) and converted to an orthomosaic to interpret spatial distribution of sediment facies across various sections of the shoreline the facies zone boundaries.

River profiles are a good proxy of base-level change and have been widely used to get an estimate of tectonic and climate history in different geographical settings (Pedersen et al., 2018; Stucky de Quay et al., 2019; Ma et al., 2020; Sonam et al., 2021; Su et al., 2022; Clementucci et al., 2023). Any change in the base level creates a potential (knick point) at the river mouth, which then propagates upstream along the river network through erosion (Howard, 1994). This process can be mathematically expressed as a mass balance

where, z is the river profile elevation, U is the uplift rate (assumed to be block uplift) and E is the erosion rate. The erosion rate term can be modeled as a linear advection term (stream power law), where the wave velocity (migration velocity of the knickpoints) is a product of the rock erodibility (K), upstream contributing area (A) and channel gradient/slope (S) (Flint, 1974; Whipple and Tucker, 1999).

The rock erodibility, K, is an empirical parameter in the Stream Power Incision Model (SPIM) and an integration of various physical variables affecting the resistance to erosion. The relative contribution of climate and tectonics to the erosion process is determined by the power terms m and n. Overall, the advection equation with the source term as the uplift takes the following form

The above equation can be further linearised by taking the exponent of the slope (n) as 1. Studies generally find that linearised inversion results with n = 1 lead to a better fit between measured and predicted river profiles (Pritchard et al., 2009; Goren et al., 2014). The solution to the above linear advection equation (n = 1) is given in the following integral form

where τ x is the travel time of an advection wave from the river outlet (where the uplift signal is getting generated x = 0) to any point of the river profile x = x, and is mathematically represented as

In Equation 4, s is a dummy variable for the integral. τ − 1 ( t ) is the inverse function of τ ( x ) , i.e., τ − 1 ( τ ( x ) ) = x . This inverse function maps the response time and position x, transforming the distance along the channel from the river outlet into its corresponding response time. Further, the t in the upper limit of the integral in Equation 4 becomes zero as we use the present (t=0) topography as our input data.

Here, we have assumed the uplift/base level fall to be space invariant. Then, Equation 4 simplifies to

The above equation, in matrix form, can be rewritten as follows

where, A is the forward operator matrix and contains the response time terms for all the nodes of the river profile (Goren et al., 2014). We can use the linear least square solution of the above equation to find the U term

where, δ is a damping factor used to account for the ill-posed and unstable nature of the problem and the solution respectively, I is the identity matrix and Up is the prior uplift rate, which gets updated through the least-square method (Goren et al., 2014). The present study utilized SRTM digital elevation models (DEM) with a 30 m resolution to assess the uplift rate in the south Andaman and Swaraj Dweep. The model’s outputs facilitated the estimation of relative subsidence and uplift rates, which are factors that would have gradually impacted the topography. The values reported in the literature were utilized to determine the model parameters (rock erodibility (K) and area exponent (m)) in the absence of any prior information regarding the study area. Assuming n = 1, the value of m generally ranges from 0.4 to 0.6 and is not significantly influenced by the uplift rate (Whipple and Tucker, 1999; Kirby and Whipple, 2012). A mean value of 0.5 for m was employed in our modeling analysis, which is reported for other geographical regions with similar lithology and climate (Montgomery and Gran, 2001). The erosion coefficient or erodibility value employed in this research was 4.75-5.25×10^-6 yr^-1, which falls toward the upper end of the documented soil erodibility values in Mount Talang, West Sumatra (Afrizal and Masunaga, 2013). Similar values (of the order of 10^-6 yr^-1) have been reported for comparable rock types as those found in our study, including metasedimentary or volcanic rocks (Montgomery and Gran, 2001; Yanites et al., 2017; Morales et al., 2023), and other volcanic islands (e.g., Hawaii Island; Han et al., 2014). After fixing the values for the model parameters (m and K), we found the maximum response time (for the channel heads in the river network) for our study area to be ~1 Myr. Hence, we discretized the river length into intervals, each of which equated to 100 Kyr for both south Andaman and Swaraj Dweep. Interestingly, the eustatic cycles exhibited a 100 Kyr interval over the past ~1 million years (Lambeck et al., 2002). Consequently, a 100 Kyr interval was selected to mitigate the influence of eustatic component when calculating long-term tectonic uplift rates. We used the QGIS Whitebox plugin (Lindsay, 2016) through QGIS for the hydrological analysis of the DEM dataset. Additional processing of the hydrological outputs as well as river profile analysis, were carried out using Python.

The modern distribution of facies mapped in this study within the intertidal to shallow marine environments of south Andaman and Swaraj Dweep indicates that corals, algae, molluscs and microfossils ( Figure 10 ) are the main sediment producers along with pebble to cobble sized clasts and the sediments derived from erosion of ophiolites and flysch. The sediments are further re-distributed along the coast due to the local hydrodynamics (influenced by tide, currents, waves and winds).

A dominant rocky shoreline characterizes the eastern margin of south Andaman, whereas the southern and western margins display both the rocky and sandy seashores. The source of the sediments can be distinguished by the color, i.e. erosion of the ophiolites/flysch (dark) or the carbonate sediments (lighter in color). The sand is of a mixed carbonate-siliciclastic origin adjacent to rocky shorelines. The clasts distributed along some of the beaches are formed due to the fragmentation of older mafic rocks, followed by subsequent reshaping and resizing due to local hydrodynamic activity (waves). The silty to muddy siliciclastic sediments brought in by creeks/channels indicate extensive transport from the hinterlands to the shoreline. Carbonate sediments dominate on beaches where no rocky exposures are observed. The sediments are predominantly biogenic, with major components being corals, molluscs, algae and microfossils. Majority of these components are sourced from the modern fringing reef and redistributed along the beach due to local wave and tide-induced currents. The presence of grainstone in the upper intertidal zone indicate moderate to high energy environments and the significant role of wave and tide induced currents in the transportation and redistribution of sediments. In contrast, the beaches of Swaraj Dweep are primarily composed of unconsolidated biodetrital grainstone facies and coralgal boundstone (coral carpets/terraces; Holocene fringing reef). However, in some places, rock layers (mudstone and calcareous siltstone) are observed adjacent to the shoreline. At Radhanagar beach, these rocks exhibit steep dips and are yellow in color. Along Kalapathar beach, the older rocks are greyish in color and contain abundant trace fossils. Most of these rocks belong to the Middle Miocene Inglish and Long Formation (Sharma and Srinivasan, 2007; Bandopadhyay and Carter, 2017).

Beachrock, which is primarily composed of sand to silt-sized carbonate sediments consolidated by carbonate cement, is also observed along the shoreline in Andaman (Chotabalu, Chidiyatapu, Wandoor and Burmanallah). These have formed by the lithification of beach sediments due to constant wave activity. Algal/microbial mats in the rims of the beachrock attest to the cementation due to microbial activity in a high-energy environment (Vousdoukas et al., 2007). A beachrock outcrop from Chidiyatapu, which dates back to 4.4-3.9 ka BP (Rajshekhar and Reddy, 2002), suggests that these facies were formed during the Holocene highstand period (Menon et al., 2023). Additional studies are required to constrain the age (from different locations in Andaman) and origin of these beachrock.

The coralgal boundstone facies occur as terraces/coral carpets, which is a major sea-level dating marker also called ‘sea-level indicator’ (Rovere et al., 2016a; Khan et al., 2019; Mann et al., 2019; Khanna et al., 2021; Menon et al., 2023). In south Andaman, the dominant exposed coral morphologies are massive to branching predominantly Porites sp., including Porites microatoll ( Figure 4E ), while in Swaraj Dweep Porites microatoll and platy Acropora sp. ( Figures 7C, D ) are predominant, along with occasional encrusting corals. The presence of these corals within and above the inter-tidal zones in south Andaman and Swaraj Dweep indicates that the sea-level was higher in the recent past, also termed as Holocene Highstand, that has been identified elsewhere in India too (Menon et al., 2023).

Further, the coralgal rudstone are characterized by gravel-sized coral and clastics (pebble to cobble mafic/flysch). The coralgal rudstone facies, with its gravel-sized (large mean size) upturned coral clasts along with volcaniclastic sediments and skewed distribution of grain size could provide evidence of previous overwash deposits (previous storm/cyclonic deposits) (Joyse et al., 2023). The coral clasts are upturned, thus providing evidence of high-energy events transporting the corals from the fringing reef. These deposits are confined to the intertidal zones (limited landward extent) and are dominantly constituted by the beach sediments, further providing evidence of storm deposits (Brill et al., 2020).

Owing to the active subduction in the region, the studied areas are situated within tectonically active and earthquake prone area (Mishra et al., 2007; Moeremans et al., 2014). Model studies about the earthquake cycle in subduction zones often consider that the upper plate’s deformation is entirely elastic and fully restored over multiple cycles of earthquakes (Nicol et al., 2009; Govers et al., 2018). This assumption does not consider any long-lasting vertical movement during the earthquake cycle, especially over thousands of years. However, significant coastal uplift has been documented in several other subduction regions (Such as in Andean margin ~0.1 mm/yr between 122 Kyr and 6 Kyr) over tens of thousands of years that passes through several earthquake cycles (Saillard et al., 2009; Mouslopoulou et al., 2016). So in general, on longer time intervals, uplift tends to dominate the vertical motion of terrestrial fore-arc regions through seismic and aseismic processes (Wilson et al., 2007; Mouslopoulou et al., 2016). Paleo-shoreline indicators, such as multiple tidal notches or uplifted terraces, are observed along the shoreline, however, these indicators may be affected by eustatic changes periodically, potentially masking the underlying tectonic movements (Laborel et al., 1994).

The average uplift rate determined in this study is 0.13 mm/year for south Andaman and 0.05 mm/year for Swaraj Dweep over the course of the past 1.4 and 0.8 million years, respectively (Considering K = 5×10^-6 yr^-1 and Δ = 100 Kyr) ( Figures 11 , 12A ). The eustatic sea-level fluctuated around 120-130 m per 100 Kyr cycles for the last 800 Kyr (Lambeck et al., 2014). However, if we zoom in to examine eustatic changes at smaller time intervals, we find numerous instances where the rate was exceptionally high. For example, if the last 20 Kyr of eustatic rates is considered, the sea-level change rates range from with an average of 6 mm/yr with as high as 40 mm/yr during melt water pulse 1A (MWP-1A; Deschamps et al., 2012), while the average tectonic-driven uplift rate was smaller by a few magnitudes (for e.g. 0.13 mm/yr for south Andaman and 0.05 mm/yr for Swaraj Dweep). As a result, although the exposed coral terraces observed along the coast are formed by a combination of eustatic and tectonic components, the eustatic component clearly outweighs any tectonic component in the formation of younger (millennial year timescale) coral terraces in the current study area.

The global mean sea-level is currently rising at a rate of 4.2 mm/yr and this rise has been generally attributed to the melting of ice sheets and the thermal expansion of water (Fox-Kemper et al., 2021; NASA Observatory). It is also essential to note that the sea-level rise/fall rates can vary spatially due to factors such as tectonics, global isostatic adjustments and local factors such as sediment compaction (Rovere et al., 2016b). Thus, the sea-level fluctuations could lead (among other factors that are salinity, temperature, nutrient supply etc.) the shallow marine carbonates to either backstep, downstep, or drown (Sarg, 1988; Schlager, 2005; Brigaud et al., 2014).

In this study, we map the current exposures of the coral terraces/carpets in the intertidal zones along south Andaman and Swaraj Dweep. The corals dated from Swaraj Dweep (Awasthi et al., 2013) has been found to be around 8.7 to 8.4 Kyr BP, while coral terrace from Interview Island has been dated back to ~7.5 Kyr BP and from North Cinque Island corals are dated to be ~700 years to ~1.7 Kyr BP ( Figure 1 ; Rajendran et al., 2008; Kunz et al., 2010). The exposed corals dated from Mayabunder has been found to be around 505 years BP ( Figure 1 ; Yu et al., 2022). Additionally, due to recent Earthquake in 2004, part of Andaman also got uplifted that also has exposed corals that have been dated to be modern in age (Meltzner et al., 2006). To identify which process, i.e. eustasy or tectonics, is predominant, we further model the uplift rates in south Andaman and Swaraj Dweep, that is an amalgamation of the coseismic deformation along with the interseismic and aseismic surface deformation (Mishra et al., 2021). It is identified that the long-term average uplift rates (0.06 mm/year south Andaman and 0.03mm/year for Swaraj Dweep for the last ~400Kyr) are few tens of times smaller than the eustatic sea-level rise rate. Thus, this study clearly establishes that the eustatic component is the major force driving the formation of Holocene Fringing reefs. The Holocene fringing reef has now down-stepped to the current location of the modern fringing reefs in the subtidal areas of south Andaman and Swaraj Dweep ( Figure 13 ).

Mapping of the Holocene fringing reefs is critical, as they not only provide natural barriers within the intertidal zone but also are going to be the substrate of the future reefs in the context of ongoing eustatic sea-level rise.

Our present study estimates a time-varying long-term tectonic uplift (base-level fall) rate using linear inversion of river profiles from south Andaman and Swaraj Dweep. The results conspicuously show the inherent non-stationary nature of tectonic activity in the forearc region. In both islands, there is a distinct period of tectonic dormancy (low uplift rate) followed by a more active tectonic phase of long-term uplift (nearly doubling of average uplift rates). The average uplift rate of ~0.13 mm/yr for ~1 Myr (south Andaman), and ~0.05 mm/yr for ~800 Kyr (Swaraj Dweep) has been identified. The uplift rate estimated from the inversion model is of the same order of magnitude (toward the lower end) as compared to the long term average uplift rates from other subduction settings such as Cascadia (~0.1mm/yr to ~0.9mm/yr), Makran subduction margin (~0.05 mm/yr to ~1.2mm/yr) and the Andean margin (~0.6mm/yr to ~1mm/yr) (Personius, 1995; Saillard et al., 2009; Pedoja et al., 2014; Normand et al., 2019; Padgett et al., 2019). In this study, older tectonic phases (older than 1 Myr) are a bit ambiguous to interpret owing to the nature of the input elevation data (lesser number of data points with more scatter). However, the rates (0.2-0.4 mm/yr) are within the error limits of the global average rate of base-level change (Pedoja et al., 2014).

These rates could be used as boundary conditions to predict how much uplift has happened of the coastal areas, which could be merged with the information on previous sea level highstands (eustasy driven) to identify the exact elevation of past sea-level highstand carbonates. It is already established that in tectonically active regions the previous highstand deposits are at higher elevations. For example, in the tectonically active region of the Indonesia Archipelago, coral terraces dated to the MIS 5e have been identified, with an average elevation of 34m, indicating the imprint of both eustasy and tectonic uplift (Pedoja et al., 2018). In the Gulf of Aqaba, due to its vicinity to a transform fault, the coastal uplift rate is around 0.15 mm/yr measured from 25 m above MIS 5e deposit (Bosworth et al., 2017). Farther, in the subsiding margin, the past highlands are found at the present-day sea level or below it. For example, in Hawaii, interglacial and older carbonates are reported from the below present-day sea level (Webster et al., 2007). Thus, the coral terraces in south Andaman and Swaraj Dweep can serve as significant sea level indicators for the late Quaternary.

Based on the average uplift rate deduced from this study over the last 400,000 years, the elevations of the MIS 5e, MIS 7, MIS 9, and MIS 11 terraces are expected to be approximately 10 ± 2.5m, 19.2 ± 4.8m, 25.6 ± 6.4m, and 32.8 ± 8.2m on the southern Andaman (with 0.13 mm/yr uplift rate), respectively ( Figure 13 ). At present, no such highstand deposits have been identified in our study area representing MIS 5e or older highstand deposits. This is likely due to erosion caused by prolonged exposure to a tropical climate and/or burial below the extensive vegetation cover.

Significant tectonic activity and the establishment of multiple paleotsunami events in the Andaman region indicate the possibility of future earthquake-induced tsunamis. The devastating effects of the 2004 earthquake serve as evidence of the destructive nature of such events, resulting in the damage of low-lying areas along the Andaman coast. These impacts included the uplift of coral terraces (Meltzner et al., 2006; Kayanne et al., 2007), destruction of crop fields and coastal mangroves (as observed in the field) and the displacement of coastal communities (Malik et al., 2019). Given the scenario of rising sea levels, the combination of cyclonic storms and tsunamis in the future could cause even greater damage in the forms of inundation, erosion of the low-lying areas and destruction of the modern fringing reefs.

The Andaman region falls within a subduction zone, making it tectonically active and prone to earthquakes (Mishra et al., 2021), which can potentially trigger tsunamis (Kohl et al., 2005). Numerous studies have focused on gaining insights into past tsunami occurrences by employing sedimentological and geochemical approaches and estimating their ages through accelerator mass spectrometry (AMS) or optically stimulated luminescence (OSL) dating techniques (Jackson et al., 2014; Dura et al.,2015; Malik et al., 2019). In one such study, Malik et al. (2019) collected samples from trenches and cores near the Badabalu beach area and identified a few paleo-tsunami events in the Andaman Islands over the past 8000 years. These events took place in 1881, 1762, and 1679 CE, between 1300 and 1400 CE, as well as in 2000-3000 BCE, 3020-1780 BCE and 5600-5300 BCE (Malik et al., 2019). A separate study Jackson et al. (2014) focused on examining sediment cores collected from a lagoon located on the southeastern coast of Sri Lanka in order to determine the paleo tsunami records during the Holocene period. Organic matter was dated from layers above and below the tsunami deposits to determine their ages. The midpoint ages of these recorded tsunami events have been estimated to 2700, 4200, 4500, 5000, 6200, 6400, and 6600 yr BP (Jackson et al., 2014). It is worth noting that all these tsunami events were triggered by earthquakes occurring along the Andaman-Sumatra subduction zone. These tectonic activities also result in changes in relative sea levels by subsidence or uplift along the coastline (Rovere et al., 2016b). In case of subsidence, the rise in relative sea level could lead to drowning of the reef, as the reefs fail to keep pace with this rise (Schlager, 1981). Subsidence can also occur due to other factors, including the liquefaction of underlying layers due to pressure change (Peltzer et al., 1998; Gahalaut et al., 2008) or viscous deformation in lower crustal or upper mantle area (Hu et al., 2004; Thatcher, 1983). On the other hand, coastal uplift results in a relative sea level drop that exposes parts of the fringing reefs, leading to their demise. Whether rising or falling, sea-level fluctuations significantly impact sediment production and distribution along the shallow shelf and coastal regions (Pomar and Ward, 1995; Kirwan and Megonigal, 2013). Lowland areas are significantly impacted by the rise in sea level, particularly in the subsiding regions located in the southeastern part of the Andaman (Meltzner et al., 2006).

The Andaman Islands have been prone to a significant number of severe cyclones because of their geographic location. Historical records, such as those found in the “Imperial Gazette of India” (Hunter, 1881), document instances of storms impacting Port Cornvallis (located on Netaji Subhas Chandra Bose Island) in 1792 and 1844, as well as Port Blair in 1864 and 1891. Throughout the twentieth century, the Andaman Islands experienced direct crossings or close passing of strong cyclones in 1911, 1914, 1916, 1921, 1961, 1976, and 1989, which have been recorded by the India Meteorological Department. One particularly noteworthy cyclone occurred in November 1989, when a storm entered the Andaman Sea from the Gulf of Thailand and intensified into a severe cyclonic storm, with wind speeds exceeding 140 km/h (Kumar et al., 2008). This exceptional event resulted in extensive destruction of coastal settlements due to the high wind speeds and generated surge of approximately 2 m, leading to coastal areas being inundated, as observed locally (Kumar et al., 2008). The coralgal rudstone facies mapped in this study within the Brichgunj area, with its gravel-sized (large mean size) upturned coral clasts along with volcaniclastic sediments, provide evidence of previous overwash deposits (previous storm/cyclonic deposits), that should be studied in details to provide better understanding on the processes depositing them. The sediment mostly would have been transported from the fringing reef by the erosion during the cyclonic/storm event.

Recently also, the Andaman Islands have experienced the impact of strong cyclones, including Bulbul in October 2019, Yaas in May 2021, and Jawad in December 2021, all of which caused widespread damage in the region (Tropical weather outlook. India Meteorological Department). These cyclonic events threaten coastal ecosystems, including mangroves and fringing coral communities, due to the flooded water’s increased salinity, turbidity, and destructive wave actions (Krauss and Osland, 2020). Thus, with the frequency and intensity of storms expected to increase due to human-induced climate change, substantial variations in sediment supply could have serious repercussions for coastline stability (Abram et al., 2008). Carbonate sediment production in Andaman is determined by the abundance of live fringing reef and the associated reef-dwelling organisms (molluscs, echinoids and other invertebrates); increased incidence of storms could lead to inundation and widespread coral mortality (Baird et al., 2018). In this context, the Holocene Fringing reefs might be useful in preventing coastal erosion owing to stronger wave action caused by storms/cyclones.