Longyearbyen Lagoon (Spitsbergen): Gravel Spits Movement Rate and Mechanisms

Abstract

Understanding lagoon behavior is crucial for both scientific research and engineering decisions, especially in delicate Arctic environments. Lagoons are vital to coastal areas, often bolstering infrastructure resilience. Since spring 2019, we have monitored the Longyearbyen lagoon (Spitsbergen), vital for coastal erosion defense and serving as a natural laboratory. The location’s well-developed infrastructure and accessible logistics make it an ideal testing site available at any time. It can be used for many natural scientific studies. The lagoon continually changes due to the primary action of waves and tides. This article focuses on gravel spit movement, accelerating in recent years to several meters monthly. Using methods of aerial and satellite images, laser scanning, and hydrodynamic measurements, we have delineated processes, rates, and mechanisms behind this movement. The measurements revealed an accelerating eastward movement of the lagoon spit, from 8 m in the first year to 86 m in the fourth year of observation. This can be explained by a combination of the reconstruction of the Longyearbyen riverbed and increased flow because of climate change. Notably, the expansion does not only occur in the summer months: from September 2022 to February 2023, the spit moved by 40 m, and then, by 19 m from February to June 2023. We found that the bed-load transport along the spit coupled with gravel slides are the primary drives of lagoon expansion and growth. We also investigated movements of groundwater in the spit and changes in gravel contents along the spit, influencing the water saturation of the gravel. Modelling these processes aids in forecasting lagoon system development, crucial for informed management and engineering decisions in Arctic coastal regions.

1. Introduction

Coastal lagoons are bodies of water that are partially isolated from an adjacent sea by a sedimentary barrier, but which nevertheless receive an influx of water from that sea. Lagoons do not receive a large freshwater inflow. One may distinguish between microtidal and macrotidal lagoons, open and closed lagoons, and intermediate states.

Some 13% of the world’s coastline is associated with lagoons [1]. However, they are more common in some Arctic regions. Lagoons comprise 40% of the coastline of the Chukchi Sea in Alaska, and they are essential elements of the region [2]. In the Russian Arctic, the presentation of lagoon coasts varies from 5% in the Barents Sea to 24% in the East-Siberian Sea [3]. In Svalbard, lagoons and barrier islands together with estuaries present 29% of coastscapes [4].

Because of their large ecological significance, lagoons have attracted the attention of researchers of various specializations. Several lagoons became test sites and known in scientific community. For example, work at the Kaktovik Lagoon (Beaufort Sea) mainly focuses on the discontinuities of pH variability and CO₂ flux at the air–sea interface [5,6]. Three lagoons in Kotzebu Sound (Chukchi Sea, Alaska) gave valuable information about energy flow and trophic dynamics in the Arctic ecosystem [7]. Thermokarst processes were investigated in five lagoons of Bykovsky Peninsula (Laptev Sea) [8]. The Valkarkai Lagoon is known by its high sea level and meteorological data series (1956–1993) on the remote Chukchi Sea coast [9]. The Vallunden Lagoon in the Van Mijenfjorden (Spitsbergen) was used many times for fields work at UNIS [10,11]. Research programs at most of the abovementioned sites are mostly oriented on biology, geochemistry, waves, and sea ice processes and do not touch spit movement.

Lagoons are of significant interest to coastal processes in the Arctic, especially in the context of global warming [12,13,14]. As Spitsbergen is an example of a place where the increase in temperatures is most powerful, the coastal processes here are attracting rising attention.

Local investigations devoted to spit morphodynamics and showcasing how the global warming trend is accelerating coastal geomorphic changes at the head of the deglaciated fjords are very relevant and important [15,16,17,18,19,20,21]. Strzelecki et al. [17] detailed the post-Little Ice Age (LIA) (1590–1850 AD) sediment fluxes at the coastal zone in Billefjorden, central Spitsbergen and analyzed the response of the gravel-dominated barrier coast to the decay of a glacier. They showed that glacier retreat resulted in the development of paraglacial sediment cascade, where eroded and reworked glacigenic sediments progressed through alluvial fans to the coast, thus feeding gravel-dominated spit systems. In colder phases of the post-LIA period, coastal zone development was subdued and strongly dependent on the efficiency of sediment transport via a longshore drift. The acceleration of coastal erosion and associated spit development was coincident with rapid climate warming from the 1980s. Kim et al. [15], on the basis of a repeated photogrammetry survey (2016–2019), described pronounced variability in the time and space of gravel spit complexes on the tidal delta plain at the head of Dicksonfjorden (Northern branch of Isfjorden), estimating that the youngest spits elongated by 22 m/yr and migrated landward by 4.3 m/yr, explained by longshore drift and, to a lesser degree, by overwash processes.

So far, detailed measurements are unique and insufficient to create even a local picture, let alone Arctic-wide changes, in the context of global warming and explaining how gravel spits move. Lidar and hydrological measurements in connection with spit movements were not presented.

We investigated the Longyearbyen lagoon as an important site for local community/infrastructure and scientific experiments, aiming to describe the topography, estimate the rate of expansion of gravel spits, and reveal the mechanisms of spit movement. This article presents the results of a comparison of aerial photographs and space images (1965–2023), the processing of data obtained via laser scanning (2019–2023), and hydrological measurements (2022–2023).

2. Geographical Framework

The investigated lagoon is in the High Arctic (coordinates DD MM SS: 78°13′25″ N 15°40′11″ E) in Adventfjorden, near Longyearbyen (Figure 1), the biggest settlement of Spitsbergen. Longyearbyen, with only two and a half thousand citizens, became famous as a tourist destination and the test site or start point for many Arctic studies due to its developed infrastructure and transport accessibility. The lagoon is only 600 m to walk from the Svalbard Science Park, including the University Centre in Svalbard (UNIS), Norwegian Polar Institute, and other research organizations. The marine environment of Adventfjorden has been studied in detail and described in a booklet [22]. In the same sense, Longyearbyen lagoon is remarkably interesting for the investigation of coastal process and conducting various full-scall experiments in natural conditions. Coastal processes are particularly important in the Arctic region with permafrost spreading on the background on ongoing climate change [23]. The consequences are visible and convenient to investigate in Longyearbyen, where warming is occurring up 7 times faster than whole-planet warming has been reported [24,25].

The lagoon was formed and developed under the influence of sediment transport from the Longyear River, waves, currents, and tidal processes. During the snowmelt period, a significant volume of sediment from surrounding slopes and glacial erosion is carried to the fjord and redistributed by the water motion in Adventfjorden. The sediments predominantly consist of local bedrocks, such as sandstones and shales, with the occasional inclusion of magmatic stones brought from the mainland for riverbed reinforcement [28].

A sewer pipe running through the central part of the lagoon disrupts natural processes. The pipe (shown by the purple arrow in Figure 1c) is buried at a shallow depth and becomes visible behind the gravel spit, surrounded by sediment accumulation at low tide in the fjord behind the spit (photo in Section 3.2. Laser scanning). Currently, the influx of fresh water from the Longyearbyen River to the lagoon is minimal, since the main flow is streaming directly into the fjord, while influx to the lagoon is prevented by a natural small dam. Despite this, the lagoon remains largely open to the fjord, and the salinity level of the water matches that of the fjord.

The location of the lagoon in the southeastern inner part of Adventfjorden determines the tidal and wave patterns that affect its spit. In the Longyearbyen/Adventfjorden area, the maximum tidal range measures 2.1 m, while the minimum range ranges from 0.33 m at its lowest to 0.82 to 1.17 m at its highest. Figure 2 illustrates the typical tidal fluctuations in Longyearbyen. Adventfjorden, a relatively small eastern branch of the expansive Isfjorden, opens widely into the Greenland Sea. In Isfjorden, the predominant tide is M2 with a period of 12.42 h. The maximum speeds of tidal currents are contingent upon local topography, ranging from 2 to 10 cm per second in broader expanses to several meters per second in the narrow straits of Spitsbergen fjords [29]. Notably, the more vigorous currents are often confined to shallow passages, such as at the entrance to the Dicksonfjorden (the northern branch of Isfjorden) and in the central sector of the Van Keulenfjorden (another fjord opening into the Greenland Sea) (Figure 1b).

Due to its geographical positioning, the greatest wave fetch occurs at the entrance to Adventfjorden from Isfjorden, particularly in the Vestpynten area (south part of Adventfjorden mouth—Figure 1b). Direct measurements using an acoustic Doppler current profilers (ADCP) indicate that maximum wave heights in this area typically remain below 1.5 m [31]. Towards the end of Adventfjorden, wave heights diminish considerably, as they dissipate in shallower waters. Nonetheless, even in these shallower regions, waves may still play a role in sediment transportation. Of paramount importance here is the velocity of water movement, which we have meticulously measured to unravel the mechanisms behind gravel spit migration.

While Adventfjorden has been predominantly ice-free in recent years, ice typically persists in the lagoon from November through May due to the comparatively tranquil water movement. Winds originating from the east tend to push ice into the lagoon, contrasting with Adventfjorden, where wind influences ice drift out of the fjord. The drag exerted by westerly winds, accumulating over several hundred meters of the lagoon, is insufficient to break up ice within the lagoon. During low tide, ice on the lagoon surface can bend by up to 80 cm. The mouth of the lagoon represents the most dynamic area, characterized by significant bending, periodic ice breaking, and floe movement, leading to erosion [32].

Ice floes, measuring up to 0.9–1 m in thickness, can be also transported from Adventfjorden to the lagoon area, where they become stranded north of the spit. These floes originate locally and represent calved pieces of the ice foot—a belt of ice frozen to the shore, primarily formed due to tidal fluctuations—and can accumulate near the beach, forming rubble. Larger floes and small icebergs originating from Isfjorden [33] do not reach the lagoon area, as they become stranded in the shallow waters to the west.

Tidal fluctuations, ocean currents, and sediment deposition from rivers collectively influence the dynamic shape of the lagoon and the shifting position of the spit. Primarily, the head of the spit steadily advances eastward, towards the inner reaches of Adventfjorden, starting from the mouth of the Longyear River.

3. Data and Methods

The obtained field data (content, instruments, methods) and additional materials used to assess the speed of lagoon spit movement and to reveal the physical mechanism are described in this chapter.

A comprehensive array of measurements has been conducted from 2019 to characterize and elucidate the movement of the lagoon spit. In this paper, we present the data up to February 2024. Figure 3 delineates the locations of equipment and measurement positions. To accurately represent the changing contours and form of the lagoon spit, we have depicted its morphology during the field campaigns in autumn 2022 and 2023, using color lines over readily accessible satellite imagery from Google Earth, reflecting the state of July 2023 [26], which serves as a base layer in Figure 3. This depiction serves as a valuable visual aid in capturing the dynamic evolution of the lagoon spit over time and measured parameters.

3.1. Aerial Photos and Satellite Images

A compilation of aerial photographs and satellite images showcasing the evolution of the lagoon has been integrated using ArcGIS Pro 3.4.0 software. Georeferenced images from the years 1936 [34], 2009 [27], 2021 [35], and 2023 [26] enabled the digitization of the coastal line and lagoon spit into polygons, as presented in Section 4.1. These images, along with their digitized representations, are shown in Appendix A.

Additionally, data from [32] was utilized to reconstruct spit movement using aerial photographs since 2005.

3.2. Laser Scanning

Since 2019, we have followed lagoon spit advancing by laser scanning with Riegl VZ1000. Riegl VZ1000 is terrestrial laser scanner with an accuracy of 8 mm, a precision of 5 mm, an efficient measurement rate up of to 122,000 measurements/s, and a maximum range of 1400 m at 70 kHz. It makes point cloud at a distance of up to one km [36].

During the ice-free season, the bottom of the lagoon is exposed at low tide, revealing wet clay deposits on the surface or shallow waters in both the internal and external parts of the lagoon. This unique feature enables the topography of the gravel spit to be accurately reflected through 3D point cloud technology and visualized following processing. Capturing the lowest water level, which occurs only several times per year, is crucial for this purpose.

To achieve this, we conducted 13 laser scanning sessions from a wooden terrace near the so-called bird observation house (blue construction in Figure 4), situated 10 m above the lagoon. This vantage point facilitated the rapid reflection of the lagoon and spit surface. However, the spit slope facing the fjord was not visible from this position. To address this limitation, we conducted measurements from scan positions directly on the spit itself, twice in September 2022 and October 2023 (refer to scan positions in Figure 3—orange and green circles). This approach enabled us to meticulously map the topography of the spit, providing detailed insights into its morphology and spatial characteristics.

The RiSCAN PRO 2.15 software [37] was employed for the initial processing of raw scans, which involved accurate adjustments, filtering, visualization, and export to common 3D formats. Section 4.2 provides an illustration of the program interface and the resulting data, showcasing point clouds for specific days and times. This software facilitated the efficient processing and analysis of the scan data, enabling us to derive valuable insights into the morphology and dynamics of the lagoon and spit.

3.3. Hydrological Measurements

Hydrological measurements included measurements of the tidal variations of seawater and groundwater levels in the spit, and measurements of sea current velocities near the spit (see location of instruments in Figure 3). The measurements of seawater and groundwater levels were performed with 3 pressure and temperature recorders SBE 39 Plus (SBE1, SBE2, and SBE3) with a sampling interval of 1 min [38] (Figure 5). Sensor SBE1 was installed inside a plastic pipe (test pipe) with a diameter of 10 cm placed in the spit at a depth of 1.5 m. The pressure sensor (PS) measured groundwater pressure at the bottom of the pipe. The pipe was protected on the top from rainwater (Figure 5b). Small holes in the lateral surface of the pipe were made to support atmospheric pressure inside the pipe above the water. The pipe was installed near the spit tip on 10 October 2022 (Figure 5c).

The second sensor SBE 2 was mounted on the leg of a metal frame used for the installation of the acoustic Doppler velocimeter ADV, SonTek Ocean Probe 5 MHz (Figure 6a). The length of frame near the bottom was 60 cm (Figure 6b). The third sensor SBE3 was deployed at the sea floor at a depth of about 10 m in Longyearbyen harbor, 1 km away from SBE1.

All sensors were deployed from the same computer with UTC time setting. The records were synchronized in time. The dates of the sensor deployments are given in

Table 1. Dates of deployment of pressure and temperature recorders SBE 39 and acoustic Doppler velocimeter ADV in locations ADV-I, ADV-II, and ADV-III.

Device	Measurement Start	Measurement End
SBE1	12 October 2022	6 August 2023
SBE2	12 October 2022	23 October 2022
SBE3	12 October 2023	17 June 2023
ADV-I	12 October 2022	23 October 2022
ADV-II	2 October 2023	3 October 2023
ADV-III	3 October 2023	4 October 2023

.

Using ADV data, we calculated the standard deviation of the water speed in each burst with the formula

$$SD\left[V\right]=\sqrt{{SD\left[VE\right]}^2+{SD\left[VN\right]}^2},$$

(1)

where $SD\left[VE\right]$ and $SD\left[VN\right]$ are the standard deviations of east ($VE$) and north ($VN$) velocities measured with a 10 Hz sampling frequency. Standard deviation $SD\left[V\right]$ characterises wave amplitude. Horizontal Reynolds stresses near the seabed were calculated in each burst with the formula

$$RS={\rho }_w\sqrt{{〈{VE}^{\prime }·{VV}^{\prime }〉}^2+{〈{VN}^{\prime }·{VV}^{\prime }〉}^2},$$

(2)

where symbol $\langle \rangle$ means averaging over 2 min, ${VE}^{\prime }$, ${VN}^{\prime }$, and ${VV}^{\prime }$ are the fluctuations of water velocities.

3.4. Sediment Samples and Criterion of Sediment Stability at Seabed

Sediment samples were collected in four places of gravel spit from the surface layer (approximately 10 cm depth to collect 3 kg of sediments) to characterize the sediment composition (Figure 7, Figure 8 and Figure 9). The samples characterize the essentially different positions in the relief and condition of sediment transport. Point 1 is the tip of the gravel spit, the most advanced part where forward movement occurs. Point 2 is a shallow plain near the spit, exposed from the water only at low tide, where an accumulation of fine sediments is visible. Similar conditions exist in the interior of the lagoon and at the bottom of the fjord with calm water. However, it was interesting to show the contrast between the dynamic tip and the calmer shallow water, locating only in several meters from each other. Points 3 and 4 are in the middle, already forming part of the gravel spit, on the ridge (not flooded at high tide), and on the beach (covered by water in high tide and constantly being under the influence of waves/currents), respectively. It should be noted that even this middle part of spit can change its position, moving mainly towards the shore, as can be seen when analyzing photo images from previous years and laser scanning results. That is, it is also very dynamic.

Figure 8 and Figure 9 visualize the samples. They were processed following the traditional sediment analysis protocol [39]. We use the set of sieves—1/16, 1, 2, 4, 6.3, 8.10, and 20 mm to obtain the diagram (Section 4.5). The shallow sample significantly changed after treatment, and it is shown on Figure 8. Initially, a wet silt sludge sample had formed upon drying into clots, 1 to 3 cm in size, which looked like stones (Figure 8b). These clots turned into fine dust after sifting (Figure 8c). The organic residues appeared as black sticks about 1–2 mm in size (Figure 8c bottom right corner).

The stability of granular material in a flow near the seabed can be determined by the Shields criterion [40]. The Shields approach is based on a uniform, permanent flow with a turbulence generated by the bed roughness. In our case, there is additional flow pulsations generated by waves. As it is shown in [41], the Shields criterion is also valid for conditions with currents plus waves. The criterion is based on the consideration of the critical value of the Shields parameter.

$$\Psi ={\displaystyle \frac{\tau }{({\rho }_s-{\rho }_w)g{D}_{50}}}$$

(3)

where $\tau$ is total water stress at the bed, including stresses due to current and wave motion estimated with Formula (2), ${\rho }_s$ and ${\rho }_w$ are the densities of soil grains and water, $D_{50}$ is the is the mean or average particle size grain diameter, and $g$ is the acceleration of gravity.

The criterion of soil stability at the bed is given by the formula.

$${\Psi }_{cr}=r{K}_1{K}_{2}f\left(Re\right),$$

(4)

where $Re=D_{50}\sqrt{\tau /{\rho }_w}/\nu$ is the particle Reynolds number, coefficients $K_1$ and $K_2$ characterize the influence of the bottom slope in normal and parallel directions to the flow direction, and $r$ is the damage parameter changing from 0.4 by occasional particle movement to 1 by frequent particle movement. Lines S corresponding to $r=1$ and $r=0.5$ showing critical values of the Shields parameter ${\Psi }_{Cr}$ versus the granular Reynolds number $Re$ are shown in Section 4.6. Coefficients $K_1$ and $K_2$ are estimated with the formulas

$$K_1\approx 1\pm {\alpha }_1/\sin \phi , K_2=1+O\left({\alpha }_2^2\right),$$

(5)

where ${\alpha }_1$ and ${\alpha }_2$ are the bed slope angles slope in the normal and parallel directions to the flow direction, and $\phi$ is the angle of repose. Signs “+” and “−” are related to upsloping and downsloping motions.

4. Results and Discussions

4.1. Longyearbyen Lagoon System Development 1936–2024

During the 5 years of our observations, the spit migrated eastward by over 200 m. It has protruded approximately 400 m from its location in 2009, reflected on all conventional maps of Longyearbyen (Figure 10). Moving away from the mouth of the river, the lagoon increasingly loses its connection and supply of river waters.

Reconstruction based on georeferenced images (Figure 11) reveals the dynamic and ever-changing nature of the lagoon system. These changes are depicted in Figure 11 by the shoreline and spit contours represented as polygons, with colors corresponding to different years. Original images and their digitalization are presented in Appendix A. There have always been several spits stretching from the Longyear River in an east–southeast direction. This natural phenomenon is typical of coastal environments where gravel spits often develop due to significant wave action and longshore drift. Wind and water currents play a crucial role in shaping these spits, contributing to their elongation or curvature depending on their direction.

In the 1936 image, prior to significant anthropogenic influence, one relatively straight and thin spit is observed with a hook formation near the river, while several spits with thickenings at the base of curved (hook) ends can be seen half a kilometer to the east. By the year 2009, the coastline underwent a transformation due to the construction of an embankment, replacing former sandbanks with structures. This embankment cast a shadow over the eastern system of spits, while three short spits near the delta (left part of the image) curved towards the shore in a south–southeast direction, a configuration now commonly depicted on maps of Longyearbyen. In the 2021 image, both the gravel spit in the shadow of the embankment and the spits along the coastline had significantly shifted eastward (by 230 and 140 m, respectively) and assumed an east–southeast direction. Those nearer to the delta exhibited a pronounced bend towards the south, whereas the eastern ones expanded more towards their ends. By 2023, the imagery depicted four short gravel spits near the delta, a substantial advancement and approach of the central spit towards the shore, moving eastward by 112 m. Consequently, the access of fresh water to the central lagoon has ceased, while the spits sheltered by the embankment remained relatively unchanged.

4.2. Gravel Spit Topography and Movement 2019–2023

The visualization of point clouds, obtained in 13 laser scanning sessions, reveals the evolving gravel spits topography and movements. Appendix B includes all 13 scans in chronological order. The first and last point clouds are opened and visible, and all are listed in Figure 12. The presented view from above with coordinate axes in meters with a palette in traditional topographic colors shows the height of gravel spits of about 2 m (from 32.1 to 34.2 in the project coordinate system, which corresponds to the WGS84 ellipsoid height) and the advance of the ridge of the spit (red-cinnamon color) by 180 m to the east in 4 years. This reflects the main features of the spit shape and gives an idea of the possibilities that scanning data processing provides for monitoring the movement of spits.

A spit ridge, not covered by water during high tide is narrow (10–20 m) and winding. The ridge has an end advanced part that moves to the east and is itself a head rising by 10–20 cm, separated by a depression that can be overwhelmed by waves during storms at high tide. Several branches 20–30 m wide (yellow) extend from the ridge into the lagoon.

4.3. Spit Expansion Rate 2019–2023

Based on laser scanning data, we can provide a detailed estimation of the speed of gravel spit expansion. By successively opening and digitizing the point clouds obtained during scanning, we tracked the movement of streamers over 4 years, as shown in Figure 13, for eight key scans. This visualization allows for a precise tracking of the spit’s growth over time.

Notably, the spit exhibited rather insignificant movement, shifting 8 m eastward from 30 September 2019 to 3 October 2020. Subsequently, in the following year (2020–2021), it advanced by 40 m, and a further 42 m from autumn 2021 to autumn 2022. Accelerating its pace, the movement doubled the following year, covering 86 m from 11 September 2022 to 5 October 2023.

Intermediate shots marked by green arrows reveal interesting trends. For instance, during the eight winter months from September 2021 to June 2022, the movement was notably subdued, totaling 18 m, compared to the 24 m recorded over the four warm months from June to September 2022. Noteworthy is the movement during the winter of 2023, where it exhibited remarkable activity. Specifically, from September 2022 to February 2023, the spit moved 40 m, followed by 19 m from February to June 2023, and a further 27 m from June to October. The trend persisted, with the spit moving an additional 30 m eastward from October 2023 to February 2024, according to GPS measurements.

In certain years, such as those before 2009 and, notably, in 2024, Adventfjorden was completely covered by ice, resulting in the appearance of small shore rubble on the spit beach. Beneath this ice cover, the velocity of tidal water motion escalates. This intriguing ice phenomenon in the region warrants further investigation, particularly since spit movement is not solely confined to warmer periods, when sediment influx from the river is most vigorous and coastal relief reshaping would be more expected and evident. Through our observations of the lagoon, we have determined that, at times during winter, the gravel spit can migrate eastward, sometimes even more prominently than during summer months. This underscores the complexity of the mechanisms driving spit movement and highlights the need for comprehensive research into the interplay between ice dynamics and coastal geomorphology.

4.4. Results of Hydrological Measurements

The main results of hydrological measurements are presented in Figure 14, Figure 15, Figure 16, Figure 17 and Figure 18 and explanations.

Blue and green lines in Figure 14a show, respectively, the seawater levels measured by SBE3 and SBE2 versus time. Blue lines were moved in the vertical direction to offset zero mean seawater level. The blue line was moved in the vertical direction to obtain a zero mean level of seawater. The green line was aligned by the upper level of seawater, since water levels measured by SBE2 and SBE3 should coincide with high tide. The green and blue lines are almost coinciding. The local minima of the blue lines are slightly below the local minima of the green line, since sensor SBE2 was above the sea level in low tide. The green lines in Figure 14a,b are the same. The yellow line in Figure 14b shows the groundwater level measured by SBE1 versus time. The yellow line was aligned by the upper level of seawater measured by SBE2. The local minima of the yellow lines are above the local minima of the green lines, because the location of the groundwater measurement by SBE1 (point PS in Figure 5a) was above the location of the pressure measurement by SBE2 (point PS on SBE2 in Figure 6a). The increasing thickness of the green line in Figure 14 is explained by the influence of the surface waves during storm weather when $t\in (100\mathrm{h},170\mathrm{h})$.

Figure 15 shows the sea and groundwater level aligned by the upper level of seawater. The maximal difference between the groundwater and seawater levels reached in low tide was about 0.5 m. The zoomed fragment (Figure 15b) corresponding to the last syzygy tide shows a good correlation in time between the local maxima of seawater and groundwater levels, demonstrating relatively the high permeability of the spit soil. Groundwater levels began to increase immediately after the seawater and groundwater levels became equal after low tide.

The results of ADV measurements are shown in Figure 16, Figure 17 and Figure 18. Figure 16a shows the mean water speed $V$ in each burst versus time measured at the distance d ≈ 15 cm from the seabed in location ADV-I. Figure 17b shows the mean water speed $V$ in each burst versus time measured at the distance d ≈ 10 cm from the seabed in locations ADV-II and ADV-III. Figure 17a shows the time intervals when ADV was covered by the water and measured water pressure and water velocity. Measurements were not performed in low tide when ADV was above the water level. All ADV velocity data shown in Figure 16, Figure 17 and Figure 18 are of a good quality. The blue points in Figure 16b and Figure 18 have coordinates of east and north water velocities measured with a 10 Hz sampling frequency. The yellow points in Figure 17b and Figure 18 have coordinates of mean east and north water velocities in each burst.

Figure 16b shows that mean and wave-induced water motions in location ADV-I were almost reciprocating, directed from the northwest to southeast. Figure 18a indicates that mean water motion was almost reciprocating in location ADV-II of the same direction, but wave-induced water motions occurred in directions from northwest to southeast, and from north to south. It is explained by the interaction of waves near the spit tip. Figure 18b shows the reciprocal mean motion of the water along the spit in location ADV-III, and wave-induced motion from northwest to southeast. The maximal mean speed of water reached 30 cm/s in locations ADV-I and ADV-II. The water speeds measured in location ADV-III were lower than 10 cm/s.

The standard deviations of water speed and Reynolds stresses calculated with Formulas (1) and (2) are shown versus time in Figure 16c,d for measurement location ADV-I and in Figure 17c,d for measurement locations ADV-II and ADV-III. The spectral analysis of ADV data showed dominating wave frequencies in the range 0.2–0.4 Hz in all locations. Higher wave frequencies up to 1 Hz were also found in some bursts. The swell frequency of 0.1 Hz was found in a few bursts. Swell amplitudes were smaller amplitudes of local wind waves. Figure 16c and Figure 17c indicate wave-induced water velocity amplitudes below 30 cm/s, 20 cm/s, and 10 cm/s, respectively, in locations ADV-I, ADV-II, and ADV-III.

The blue points in Figure 16b show the dominant direction of wave propagation from northwest to southeast under the angle of about 30° to the north from the marine side of the Adventfjorden in location ADV-I. The narrow cloud of blue points indicates the waves of other directions. The dominant direction of wave propagation is similar to the direction of the mean current in this point.

Figure 16d and Figure 17d show that Reynolds stresses are well-correlated with the standard deviations of water velocities, indicating wave amplitudes. Maximal Reynolds stresses reached and exceeded 3 N/m³ in location ADV-I. In location ADV-II, maximal Reynolds stresses were below 2.5 N/m², but the time of measurements was much shorter than in location ADV-I. In location ADV-III, Reynolds stresses were below 0.75 N/m². Reynolds stresses are used further to estimate the stability of gravel with the Shields formula.

The blue points in Figure 18a show that the wave system included waves from the southeast and south directions near the spit tip in location ADV-II. In location ADV-III, waves propagated from the southeast to the northwest from the lagoon to the inner side of the spit. Location ADV-III was in the wave shadow zone of waves propagating from the fjord behind the spit.

4.5. Sediment Analysis

Petrographically, the lagoon spit primarily comprises local bedrocks, predominantly coarse- and fine-grained sandstones (appearing grey or brown), and shale (appearing almost black due to dark clay), with occasional inclusions of magmatic stone (displaying red, white, or grey hues). The larger pieces typically consist of sandstone, which exhibits greater resistance to erosion. These rocks generally range from sub-rounded to well-rounded in shape, with occasional sub-angular pieces. The size spectrum varies from silt to pebble, occasionally including small cobbles of differing roundness in varying proportions. Grain size distribution curves are presented in Figure 19a.

Utilizing the traditional Wentworth grain size scale [42], we can further characterize the composition of the gravel spit as follows:

The slope of the spit tip primarily consists of pebbles (48%), granules (27%), and sand (24%), with a negligible amount of silt (less than 1%). Moving towards the shallow plane near the tip of the spit, located just 4 m from the first point, the composition shifts to predominantly silt (59%) and fine-grained sand (less than 1 mm) (34%), with minimal granules and rare pebbles (1%).

The composition of the spit ridge primarily consists of pebbles (48.6%, mostly small) and sand (40.9%, with 34% being less than 1 mm), with a smaller proportion of granules (10%). Transitioning to the spit beach, the composition shifts to mostly fine (1–2 mm) sand (58%), coarse (1–2 mm) sand (17%), granules (6%), and relatively small pebbles (14%). Occasionally, small oblong cobblestones about 7–10 cm in size are encountered on the ridge and beach. These larger cobblestones were not included in the calculations for granulometric composition. All sediments except spit tip also contained a small amount (less than 1%) of silt.

We also used previous sediment studies on the Longyearbyen coastline. Samples for grain size analysis were taken from the top 2 m layer of sediment on the shore near the spit. The initial grain size distribution curve is shown in Figure 19b by a solid line. We consider the mean grain size $D_{50}$ and set that 90% of the total particles are smaller $D_{90}$. According to the solid line in Figure 19b, the grain diameters $D_{50}$ and $D_{90}$ are ~5.5 mm and ~23 mm. The dashed line in Figure 19b is constructed from a solid line by removing sand particles with diameters smaller 1 mm. Diameters $D_{50}$ and $D_{90}$ are ~8 mm and ~23 mm for the dashed line. Further, it is assumed that the dashed line describes the particle size distribution with a gravel slope at the spit tip.

The sediment amount and inflow regime depend on local weather and vary significantly, experiencing climate change consequences. The volume increased notably in recent years and became more irregular, while global warming is the most pronounced in the Arctic and particularly in Longyearbyen [43,44,45,46,47]. That gives rise to questions on risk and city sustainability and pushes the preventive measure [48,49]. The Longyear River has been artificially channeled to create space for the settlement, and periodic efforts are made to enhance its bed and adjacent slopes. These alterations have led to a more controlled sediment flow regime [47,50,51]. Unfortunately, we did not find any precise data on the sediment transport dynamics in the Longyearbyen River in recent years, which could be very valuable for explaining the movement of the lagoon spit and for forecasting.

4.6. Physical Effects Influencing Spit Movement

The stability of granular material in flow was estimated with the Shields criterion, given by Formula (4). The inclined lines in Figure 20 were calculated with different diameters $D_{50}$ and stresses $\tau$ changing from 0.1 N/m² to 3 N/m². The values of water stresses correspond to the calculated Reynolds stresses in Figure 16d and Figure 17d. The grain density was measured and assumed to be equal ${\rho }_s$ = 2600 kg/m³. The water density equaled ${\rho }_w$ = 1030 kg/m³. One can see that grains of diameter greater 4 mm are stable at the horizontal bottom. Grains with diameters 0.5 mm, 1 mm, and 4 mm start to move when Reynolds stresses reach, respectively, 0.25 N/m², 0.48 N/m², and 3 N/m². Thus, sand is not stable under the local conditions, but coarse gravel with grain diameters greater than 5 mm is stable.

Van Rijn [41] noted that ${\Psi }_{cr}$ depends also on $D_{50}/D_{90}$ and ${h/D}_{50}$, where 90% of particles have diameters smaller than $D_{90}$, and $h$ is the water depth. The ratio $D_{50}/D_{90}$ expresses the grading of the bed material. Sand particles are more difficult to mobilize when they are hiding between larger gravels. It leads to a reduction in ${\Psi }_{cr}$ by $D_{50}/D_{90}>0.2$. The representative value of $D_{50}/D_{90}$ calculated from Figure 20 equals 0.24 for soil taken from top 2 m of the beach. The sand concentration of the spit tip is smaller. The value of $D_{50}/D_{90}$ reaches 0.35 on the dashed curve in Figure 20, constructed from the solid curve by removing particles with diameters smaller than 1 mm. The value of $r$ may drop to 0.4 in this case. Figure 20 indicates the mobility of gravel with diameters smaller than 8 mm in this case. Thus, the accumulation of gravel at the end of the spit can be explained by the combined action of tidal current and waves.

On 3 October 2023, we observed that ADV deployed near the spit tip was buried under the gravel when the seabed was water-free around the spit (Figure 21). The ADV velocity data were of good quality, and the distance of sampling volume to the bed was 8–10 cm in the last burst of deployment in location ADV-II at 5:37 UTC on October 3. Gravel slide occurred between 5:37 and 7:46 UTC on October 3. The spit tip was replaced on about 2 m due to the slide. Thus, we think that the gravel slides at low tide phases are the other physical mechanism of spit development. It was relatively simple to dig out ADV from the gravel because of the high porosity and liquefaction of the gravel. Legs sank in the gravel under human weight, while 3 m either side of the gravel slope, the spit surface was hard.

Wave runup over the spit may influence the larger velocities of water above the spit leading to the gravel and sand transport [52]. These effects were observed systematically during high tide. For example, on 5 October 2023, the spit was eroded by waves not far from the tip (Figure 22a). At the same time, water was never observed in the small region near the spit tip (Figure 22b). The location of the ADV battery on the spit tip is visible as an orange-brown bag in Figure 22a.

Laser scanning of the spit was performed at low tide phases on October 2 and October 5 (Figure 23). One can see that the spit tip was replaced to the east on 2–3 m, and the spit was eroded over ~30 m distance. The eroded place is extended from the spit tip at a distance of ~25 m. At the same time, the spit tip became higher at ~20 cm. The elevation of the spit tip cannot be explained by the direct action of shear stresses in the water, since overwash of the spit tip did not occur.

The possible explanation is related to groundwater motions and the action of submerged gravel on unsubmerged gravel on the spit tip. Groundwater motions are caused by tide-induced and wave-induced water level variations near permeable beaches [53]. Tide-induced groundwater motions in the coastal industrial zone of Longyearbyen, Spitsbergen, were observed and investigated by Marchenko et al. [54]. Analytical and numerical simulations of the tidal pumping effect were performed by Fomin et al. [55]. Li et al. [56] modelled swash and beach groundwater flow to predict beach profile change in the swash zone. Horn [57] mentioned the importance of groundwater outflow and infiltration on sediment transport in the swash zone. We assume that groundwater motion may influence the convergent creep motion of gravel to the tip center in the case of a small spit surrounded by water near the tip.

5. Conclusions

The goal of the presented work was to assess lagoon development with the precise laser scanning and hydrological measurements and uncover the physical mechanism of spit movement.

Our vision of physical mechanisms influencing the spit extension consists of the following. The conducted studies have shown the maximum current speeds caused by waves to the southeast. According to the wave and tidal currents, the dominant transport of sand and gravel is directed to the spit tip. Gravel spots on the spit surface move along the spit and are grouped on the spit tip. Gravel material is provided by the spit itself for expansion on clay and sandy beds. Sand bars are formed in the front of the spit tip. Sand is washed out of the gravel during low tide, providing better capacity for the gravel transport. The fast speed of spit extension is explained by the systematic gravel slides of the spit tip. Groundwater motions inside gravel and the pressure of submerged gravel influence the small growth of the spit in the vertical direction near the tip.

In our five-year longitudinal investigation, we have discovered the dynamics of gravel spit migration and movement speed and elucidated potential underlying mechanisms. The ongoing monitoring of the lagoon holds paramount significance from an ecological standpoint. Notably, the eastward advancement of the spit, notably exacerbated post-2020, engenders a sluggish and intricate exchange of aqueous masses between the lagoon and the fjord, thereby impeding natural filtration processes. Consequently, effluent-derived contaminants may accumulate, precipitating adverse environmental ramifications. This underscores the imperative consideration of such dynamics in the strategic planning of infrastructural undertakings. In its entirety, the Longyearbyen lagoon stands as a compelling locus for the investigation of sea ice dynamics and sediment transport phenomena, warranting sustained scholarly attention.