1 INTRODUCTION

The water column in open oceans usually presents distinct turbid layer (Vsemirnova et al., 2012; van Haren and Hosegood, 2017). Nepheloid layer is a layer of water with high turbidity relative to waters below and/or above in the oceans (McCave, 2019). The nepheloid layers are a significant transport mechanism in delivery of sediments, since it can transport large quantities of sediments over long distances, from the continental shelf to continental slope and pelagic environments (McPhee-Shaw et al., 2004). Understanding nepheloid layers helps in assessing the ocean sediment dynamics and sediment redistribution. Moreover, such high turbidity water can contain a wide range of rich organic matter and particle reactive elements, including nutrients, carbon, and trace elements (Lorenzoni et al., 2009). Therefore, nepheloid layers may influence the faunal feeding, distribution patterns, and the functioning of marine ecosystems (Wilson, 2016; Masunaga et al., 2017). Additional knowledge about the origin of nepheloid layers would advance our understanding of physical, geological, sedimentary, and ecological processes occurring along continental margins (Geng et al., 2018).

Surface nepheloid layers are found in the euphotic zone and result from production of planktonic organisms or river discharge into oceans (Gardner et al., 2018). Typically, there is a rapid decrease in particulate matter concentration below the surface nepheloid layer (de Madron et al., 2017) and concentrations are usually low through most of the water column. Beneath the clear intermediate water, elevated particle concentrations near the bottom are termed the bottom nepheloid layer (Biscaye and Eittreim, 1977; Gardner et al., 2017). In addition to the surface and bottom nepheloid layers, another important class of nepheloid layers observed along continental margins are intermediate nepheloid layers (McCave, 1986). Previous studies have well indicated that the intermediate nepheloid layers are principally generated by interplay between bottom sediments and with upper ocean dynamics, such as internal waves (Bogucki and Redekopp, 1999; Puig et al., 2004; Bogucki et al., 2005; Bourgault et al., 2014), internal tides (McPhee-Shaw et al., 2004), benthic storms (Gardner et al., 2018), open-ocean convection (de Madron et al., 2017), horizontal advection (Pak et al., 1980), the South Atlantic Central Water (Inthorn et al., 2006), Deep Western Boundary Currents (McCave, 2019) and semidiurnal tidal currents (van Weering et al., 2001). Recently, bottom trawling was considered as a possible mechanism for intermediate nepheloid layer generation in the Foix submarine canyon (ArjonaCamas et al., 2019). So far, however, almost no causes below the seafloor have been recorded to the best of our knowledge (McCave, 2019).

Previously, observational (Puig et al., 2013; Cheriton et al., 2014; Geng et al., 2018; McCave, 2019), experimental (Helfrich, 1992; McPhee-Shaw and Kunze, 2002; Tian et al., 2019), and numerical (Bourgault et al., 2014; Arthur and Fringer, 2016; Masunaga et al., 2017) approaches have been developed to investigate the nepheloid layers. In our laboratory experiment which was intended to research the seafloor instability caused by gas hydrate dissociation, we discovered some unexpected phenomena associated with the nepheloid layers. The interesting experimental results suggest that nepheloid layer may be generated by gas eruption, which is a new mechanism for nepheloid layer generation.

2 MATERIAL AND METHOD

Research was conducted in a transparent acrylic tank (Figs. 1–2) whose length was 200 cm, width was 40 cm, and height was 50 cm. Some dots (Fig. 2), with horizontal spacing of 20 cm and vertical spacing of 10 cm, on the side wall of the tank were used to identify the motion of subaqueous masses. The tank was designed to investigate the seafloor instability caused by gas hydrate dissociation and earthquakes.

Gas hydrates are stable at low temperature and high pressure. When this condition is changed, gas hydrate will be dissociated into gas and water, causing seafloor instability (Jia et al., 2016; Zhu et al., 2020, 2021). To simulate gas hydrate dissociation in this experiment, gases were introduced to the tank via air intakes in the bottom. An air compressor introduces air into the system. A pressure regulating valve connected the air compressor to the tank, controlling the gas input to the tank during the experiments. Earthquakes usually worsen the seafloor stability. With a vibration table under the tank (Fig. 1), we can, to a certain degree, simulate earthquakes. Two motors (red cylinders in Fig. 1) were used to provide vibration in our experiments.

The seabed we constructed features a sandwichlike layering of different types of sediment on top of each other (Fig. 2). A 2-cm-thick sandy layer (in yellowish-brown, Fig. 2) was sandwiched between two clay layers (in gray, Fig. 2). The overlying clay layer is 10 cm thick and the slope inclined at about 16°. Gases were introduced to the sandwiched sandy layer via soft gas tubing whose outlets were at the same height as the sandy layer and the inlets were connected to a pressure regulating valve. We carried out two experiments. In the first experiment, we aerated the seabed without vibration. In the second experiment, we aerated the seabed and exerted vibration load.

3 RESULT

In our preliminary experiments, we did not find serious seafloor instability, such as rapid underwater landslides. Unexpected results, however, were observed.

3.1 Intermediate nepheloid layer

In the first experiment, we slowly aerated the sandwiched layer until gas erupted from the seafloor. After the eruption, the water became turbid and a bottom nepheloid layer formed near the seafloor. This bottom nepheloid layer moved downward along the slope and then surprisingly moved horizontally to form an intermediate nepheloid layer. This intermediate nepheloid layer moved laterally to far end of the tank (Figs. 3a & 4a).

Then we aerated the sandwiched layer again. Similarly, another intermediate nepheloid layer recurred at nearly the same depth by detachment of the bottom nepheloid layer and subsequent seaward advection. Before the creation of this second intermediate nepheloid layer, the first intermediate nepheloid layer was nearly horizontal (Figs. 3a & 4a). When the second intermediate nepheloid layer move forward, it did not merge with the first one initially. Interestingly, the second layer slid under the first one, lifting the first layer slightly (Figs. 3b–f & 4b–f). Motion of these intermediate nepheloid layers can be determined from the marks on sidewall of the tank. Finally, the two intermediate nepheloid layers merged into one.

3.2 Bottom nepheloid layer

In the second experiment, we aerated the sandwiched layer and exerted vibration. The vibration caused the tank to quake violently. Almost simultaneously, the tank was destroyed by vibration. Figure 5 shows the choppy water and leaking tank caused by the quakes. Even though the leaking tank stopped the experiments, we captured some interesting results. Similar to the first experiment without vibration load, a nepheloid layer occurred in the second experiment (Fig. 5). But unlike the intermediate nepheloid layer in the first experiment (Figs. 3–4), this nepheloid layer continued down the bottom slope of the tank. Moreover, the bottom nepheloid layer was thicker and cloudier than the intermediate nepheloid layer. The bottom nepheloid layer was generated by gas eruption as well as vibrations and resulted in a turbidity current.

4 DISCUSSION

The laboratory experiments in this work were intended to investigate seafloor instabilities, such as subaqueous landslides. Surprisingly, the gas eruptions generated nepheloid layers (Figs. 3–5). Almost all previous studies indicated that the intermediate nepheloid layers are generated by upper ocean dynamics and bottom trawling (Arjona-Camas et al., 2019; McCave, 2019). This is the first time experimental evidence that intermediate nepheloid layers were generated by gas eruptions.

4.1 Eruption events and gas hydrates

In marine environments, fluid eruption and/or sediment eruption is a relatively common process. Many of these eruptions are associated with gas hydrate. Solheim and Elverhøi (1993) reported geological evidence for eruptions in water depths between 320 m and 340 m in the Barents Sea and they attributed these eruptions to gas hydrate dissociation. Subsequent geochemical and geophysical investigations have further confirmed gas eruptions and identified hydrates as the active methane source (Lammers et al., 1995; Andreassen et al., 2017; Nixon et al., 2019). Andreassen et al. (2017) suggest the hydrate-controlled methane expulsion were probably widespread across past glaciated hydrocarbon provinces. A large eruption with about 2 million cubic meters of ejected sediment was identified in water depth over 2 000 m in the Gulf of Mexico. This eruption explained the relationship between mud diapirism and gas hydrates (Prior et al., 1989). Compared with violent eruptions, the hydrate-associated gas venting is a more widespread phenomenon and it has been observed in many places, such as the US Atlantic margin (Brothers et al., 2013; Skarke et al., 2014), the Arctic seafloor (Westbrook et al., 2009; Bünz et al., 2012; Berndt et al., 2014), the Black Sea (Riboulot et al., 2018), and the South China Sea (Chen et al., 2019; Zhu et al., 2019). An increase in the bottom water temperature or lowering sea level may lead to gas hydrate dissociation and possible eruptions.

As gas hydrate research expands around the world, it will be interesting to find whether hydrateassociated venting causes intermediate nepheloid layer in nature. Although pressurized air was used to simulate gas hydrate dissociation in our experiment, many events may be associated with gas aeration and eruption. A mud volcano is another important way to form eruptions (Judd, 2005). There are about 5 500 offshore mud volcanoes, mostly in deep water (Judd, 2005). An estimate by Milkov (2000) is 10³–10⁵ for known and inferred deep-water mud volcanoes. Graue (2000) reported one single eruption from a mud volcano in Azerbaijan ejected 5 million cubic meters of material and a mud volcano during its lifetime could eject 11.4 km³. The mud volcano and its eruption can be generated by deep gas accumulations, diapirs, fluidized mud migration, and gas hydrate dissociation (Graue, 2000; Milkov, 2000; Ben-Avraham et al., 2002; Mazurenko et al., 2003; Schmidt et al., 2005). Fluid migration is critical to the mud volcano formation and many deep-water mud volcanoes are associated with gas hydrates (Milkov, 2000). In addition to gas hydrate and mud volcano, other causes may lead to eruptions. Shallow biogenic gas also serves as source for eruption (Holmes et al., 1998; Cole et al., 2000). The repeated fluid expulsions in the Gulf of Lion, western Mediterranean Sea resulted from hydro-mechanical processes during events of rapid sea-level rise due to the interaction between highpressure regimes and principal in-situ stresses (Gay et al., 2017).

4.2 Nepheloid layer generation

Our experimental results show a new mechanism for intermediate nepheloid layer generation by gas eruptions. Yan et al. (2020) conducted a similar aeration experiment. In their experiment, an eruption occurred and the water became turbid, but the eruption did not generate an intermediate nepheloid layer. On one hand, the size of their tank was small compared with the volume of their violent eruption. The length of their tank was half the length of ours. On the other hand, sediments in their experiment blanketed the entire tank bottom and the air outflow was in the middle part of the tank, which leaving limited distance for possible advection. Last but not least, their entire seafloor was flat while half of our bottom was inclined. The slope contributes to the downslope motion of bottom nepheloid layer (or gravity current) and subsequent detachment (intrusion) before approaching slope toe, which is critical to generate intermediate nepheloid layer.

Material could be included in the intermediate nepheloid layer from two possible sources as long as the density of the turbid fluid is greater than the density of surrounding water. One source is the materials ejected during eruptions; the other may be associated with eruption-induced upper water column dynamics. As soon as the density of the turbid fluid is equal to the density of the fluid encountered downslope, the moving fluid detaches from the sloping bottom and moves laterally away from the slope as an intermediate nepheloid layer. This behavior is common in submarine canyons (Gardner, 1989; Palanques and Puig, 2018). Previous experiments investigating intermediate nepheloid layers have always been performed in stratified fluids (Browand et al., 1987; McPhee-Shaw and Kunze, 2002; Tian et al., 2019). However, the only water was involved in our experiments was tap water and it is the first experimental evidence for intermediate nepheloid layer generation in singlelayer fluid. Our experiments therefore indicated that stratified fluid would not be a necessary condition for intermediate nepheloid layer.

Note that scaled physical-model experiments have inherent limitations as analogs for oceanic conditions due to disparities in Reynolds number (Boegma and Stastna, 2019). However, the scaled physical-model experiments offer the most straightforward way to observe the physical processes of nepheloid layer generation. It is interesting and important to find if gas eruptions trigger intermediate nepheloid layer.

In addition to the newfound mechanism for intermediate nepheloid layers, our experiments also offered new experimental evidence for bottom nepheloid layer generation by earthquakes. In our experiments, the quakes, combined with eruptions, significantly increased the turbidity of the water, forming bottom nepheloid layer (Fig. 5). Thunell et al. (1999) reported that the increased particle concentration and downward particle flux followed an earthquake. In addition, earthquakes could contribute to eruptions even in the far field (Manga and Brodsky, 2006; Mazzini et al., 2007; Walter et al., 2009), which may further muddy the water.

4.3 Possible influence of eruption-related nepheloid layer

Recently, the undersea Tonga volcano violently erupted in January 2022 with mushroom-shaped ash, steam, and gas plumes plunging into the sky as high as 30 km (Manneela and Kumar, 2022). Floating debris, likely including pumice, hindered sea transportation (Global Volcanism Program, 2022). During another eruption in 2009, Tonga volcano is covered by the bright steam plume and surrounded by discolored water caused by suspended sediments reaching a maximum of about 10 km from the island (Global Volcanism Program, 2009). At present, reports of the intermediate nepheloid layers generated by submarine gas eruptions in nature have not been made due to a lack of direct observations or in-situ long-term monitoring. However, the intermediate nepheloid layer in water is to submarine eruptions what the ash cloud in air is to volcanic eruptions. Alarmingly, the 1815 Tambora eruption in Indonesia, the biggest volcanic eruption in human history, spewed millions of tons of dust, ash, and sulfur dioxide into the atmosphere, temporarily changing the world's climate and causing the 1816 "year without a summer" in Europe (Klingaman and Klingaman, 2013; Luterbacher and Pfister, 2015). The 1991 eruption of Pinatubo sent a cloud of ash 400 km wide to elevations of 34 km (Decker and Decker, 2006). Intermediate nepheloid layers could travel for a long distance as well. Thorpe and White (1988) found a deep intermediate nepheloid layer off the Porcupine Bank extended along the slope for over 100 km, and off slope to about 16 km. Wilson (2016) reported intermediate nepheloid layers stretch a distance of 25 km off the slope in NE Atlantic margin. Submarine eruption could feed an intermediate nepheloid layer for an extended period and cause it to extend a longer distance, which would affect the physical, geological, and biological processes from continental shelf to abyssal environments. The striking lateral transport of the dissolved iron and manganese plumes over more than 4 000 km from their hydrothermal sources along the US GEOTRACES East Pacific Zonal Transect (Resing et al., 2015; Fitzsimmons et al., 2017) challenged our understanding of the element cycles. In the future, direct observations and in-situ monitoring of eruptions and intermediate nepheloid layers will provide more information that can be plugged into computer simulations of larger events.

5 CONCLUSION

Our laboratory experiments show that gas eruptions can trigger intermediate nepheloid layers. Different from erosion-associated upper ocean dynamics, we suggest a new mechanism for intermediate nepheloid layer generation by eruptions. The stratified fluid was not a necessary condition for intermediate nepheloid layer. Sufficient space for advection and oblique slope for detachment are the key recipes during intermediate nepheloid layer generation by eruptions. In addition to the newfound mechanism for intermediate nepheloid layer, our experiments also offered a new experimental evidence for bottom nepheloid layer generation by earthquakes. Given the scale effects of laboratory experiment, it will be interesting and important to find if submarine volcanic eruption or hydrateassociated venting causes intermediate nepheloid layer by direct observations and in-situ monitoring.

6 DATA AVAILABILITY STATEMENT

The datasets generated and/or analyzed for the current study are available from the first author or corresponding author.

7 ACKNOWLEDGMENT

The authors thank Prof. Nick McCave at University of Cambridge and Prof. Wilford Gardner at Texas A&M University for their advice and comments. The authors thank the support from Shandong Provincial Key Laboratory of Marine Environment and Geological Engineering, Key Lab of Submarine Geosciences and Prospecting Techniques, Ocean University of China.