Ocean Observations Research Articles

An observing system experiment framework for the tropical Indian Ocean salinity: A case study using a constellation of three satellites

In this study impact of assimilating Sea Surface Salinity (SSS) from multi-satellites (SMOS, Aquarius and SMAP) on numerical ocean model simulations in the north Indian Ocean has been analysed under the observing system experiment (OSE) framework. Daily data sets of Aquarius, SMAP and SMOS, which were available for a common period of April–May 2015, are used to constrain the ocean model using ensemble optimal interpolation technique. Apart from the control simulation in which satellite data were not assimilated, a total of seven assimilation experiments using different combinations of satellite SSS were conducted. The impact of assimilation experiments is analysed by comparing the model-simulated variables with in situ observations. Assimilating satellite SSS results in a reduction in Root Mean Square Error (RMSE) in SSS (∼ 54%) and also in subsurface salinity (∼ 21%) over the control run. The impact of assimilating SMAP observations is maximum on model simulations with the errors reducing by ∼ 54%. Subsurface salinity improvement is better with three satellites with ∼31% improvement in RMSE in the halocline region, which was ∼11% more than single satellite assimilation. Assimilation of SSS also resulted in improved simulations of the model surface, subsurface temperature and mixed layer depth. Model results show the ability of SSS observations to complement other ocean observation networks. One important observation from this study is that while the impact of assimilating SSS observations from a single satellite was on par with the impact of assimilating SSS observations from two or three satellites in correcting simulated surface salinity, assimilation from more than one satellite had a larger impact in the salinity of deeper layers of the ocean.

Simulation of Thermal Infrared Brightness Temperatures from an Ocean Color and Temperature Scanner Onboard a New Generation Chinese Ocean Color Observation Satellite

Since 2002, China has launched four Haiyang-1 (HY-1) satellites equipped with the Chinese Ocean Color and Temperature Scanner (COCTS), which can observe the sea surface temperature (SST). The planned new generation ocean color observation satellites also carry a sensor for observing the SST represented by the payload in this paper. We analyze the spectral brightness temperature (BT) difference between the payload and the Moderate Resolution Imaging Spectroradiometer (MODIS) onboard Terra for the thermal infrared channels (11 and 12 µm) based on atmospheric radiative transfer simulation. The bias and standard deviation (SD) of spectral BT difference for the 11 µm channel are −0.12 K and 0.15 K, respectively, and those for the 12 µm channel are −0.10 K and 0.03 K, respectively. When the total column water vapor (TCWV) decreases from the oceans near the equator to high-latitude oceans, the spectral BT difference of the 11 µm channel varies from a positive deviation to a negative deviation, and that of the 12 µm channel basically remains stable. By correcting the MODIS BT observation using the spectral BT differences, we produce the simulated BT data for the thermal infrared channels of the payload, and then validate it using the Infrared Atmospheric Sounding Interferometer (IASI) carried on METOP-B. The validation results show that the bias of BT difference between the payload and IASI is −0.22 K for the 11 µm channel, while it is −0.05 K for the 12 µm channel. The SD of both channels is 0.13 K. In this study, we provide the simulated BT dataset for the 11 and 12 µm channels of a payload for the retrieval of SST. The simulated BT dataset corrected may be used to develop SST-retrieval algorithms.

Crustal and uppermost mantle structure near the Gloria Fault, North Atlantic, from ocean bottom seismometer surface wave observations

SUMMARY In this work, we present both 1-D and 3-D shear wave velocity (Vs) models of the oceanic crust and uppermost mantle below the Deep OCean Test ARray area, located ∼ 70 km north of the central section of the Gloria Fault, in the eastern North-Atlantic Ocean. The velocity models are inferred from the dispersion of surface waves recorded on ocean bottom seismometers. Dispersion measurements are obtained from the analysis of ambient seismic noise at short periods (< 14 s) and teleseismic surface waves at long periods (> 14 s) using the two-station method. The 1-D Vs model is inferred from the joint inversion of Rayleigh wave phase velocities and Love wave group and phase velocities. The 3-D tomographic model is obtained by inversion of 2-D Love wave group velocity maps as a function of depth, further constrained by the average of Love wave phase velocities obtained from ambient noise (4–9 s) and the average Rayleigh and Love wave phase velocities calculated from teleseismic data (14–44 s). The 1-D Vs model shows a sediment layer with a low velocity of 1.05 km s−1, similar to previous studies in the region. Below the sediments, we find an oceanic crust with velocities ranging from 3.3 to 4.5 km s−1. The model reaches an unusually high velocity of 4.9 km s−1 in a 20 km thick layer at depths between 16 and 36 km. We interpret this fast velocity layer as indicative of the presence of harzburgite, a residue of enhanced melting that might have been formed by the proximity between the Mid-Atlantic Ridge and the Azores mantle plume. At greater depths the velocity decreases, forming a low-velocity zone that reaches a minimum at ∼ 70 km depth, which we interpret as the maximum depth for the lithosphere–asthenosphere boundary. The 3-D model shows a structure that is mostly horizontally layered, with Vs isocontours at 3.5–4.5 km s−1 highlighting oscillations of the crustal structure with wavelengths of ∼25–30 km. These oscillations may be due to changes in the rate of mantle upwelling and magma supply rate.

Global ocean observations and applications by China’s ocean satellite constellation

Satellite remote sensing data form the basis of ocean observation and applications. China has established a satellite network platform comprising ocean color satellite constellations, ocean dynamic environment satellite constellations, and ocean observation and monitoring satellite constellations. This platform provides consistent and reliable ocean observation data crucial for marine scientific research, economic development, and early warning and forecasting. This paper comprehensively describes the development process and plans for China’s ocean satellites from their inception. It offers detailed technical specifications of ocean satellites and outlines the current applications of ocean water color satellites (HY-1), ocean dynamics and environment satellites (HY-2), and ocean surveillance and monitoring satellites (GF-3) in ocean parameter inversion, target identification and detection, and early warning and forecasting. In the future, to enhance the level of industrialization in ocean remote sensing in China, it is imperative to leverage the diversity and timeliness of ocean remote sensing data. Additionally, emerging technologies such as cloud computing and artificial intelligence should be harnessed, and the application potential of various satellite data resources should be explored.

Estimating Ocean Observation Impacts on Coupled Atmosphere‐Ocean Models Using Ensemble Forecast Sensitivity to Observation (EFSO)

AbstractEnsemble Forecast Sensitivity to Observation (EFSO) is a technique that can efficiently identify the beneficial/detrimental impacts of every observation in ensemble‐based data assimilation (DA). While EFSO has been successfully employed on atmospheric DA, it has never been applied to ocean or coupled DA due to the lack of a suitable error norm for oceanic variables. This study introduces a new density‐based error norm incorporating sea temperature and salinity forecast errors, making EFSO applicable to ocean DA for the first time. We implemented the oceanic EFSO on the CFSv2‐LETKF and investigated the impact of ocean observations under a weakly coupled DA framework. By removing the detrimental ocean observations detected by EFSO, the CFSv2 forecasts were significantly improved, showing the validation of impact estimation and the great potential of EFSO to be extended as a data selection criterion.

Ocean-atmosphere heat exchange seasonal cycle on the West Florida Shelf derived from long term moored data

Twenty-three years of surface meteorological and oceanographic data sampled from moored buoys are used to study the seasonal and interannual variations of ocean–atmosphere heat exchange and its influence on West Florida Continental Shelf (WFS) water temperature and stratification. The data are from the University of South Florida's Coastal Ocean Monitoring and Prediction System (COMPS), part of the Southeast Coastal Ocean Observing Regional Association (SECOORA). Observed are incoming short and longwave radiation, air and sea surface temperatures (AT and SST), barometric pressure, relative humidity, wind velocity, water column velocity profiles, and water column temperature at discrete depths. These data are used to estimate net shortwave and longwave radiation and sensible and latent heat fluxes via the COARE 3.6 algorithm. When combined, these radiative and turbulent heat flux influences are compared with the heating and cooling of the WFS water column and SST. On seasonal average, heating starts in February and lasts through August, with a maximum rate of change in May, while cooling starts in September and lasts through January, with the maximum rate of change in October. Also on seasonal average, SST varies from 18.4 °C in February to 30.4 °C in August at mooring C10 (at the 25 m isobath) and from 20.1 °C in February to 30.2 °C in August at mooring C12 (at the 50 m isobath), the differences in the seasonal range being due to increased ocean circulation influence in deeper water. Both the spring and fall transition onsets, February and August, respectively, occur when the sign of the net heat flux changes. The water column begins to stratify in March, peaking in June–July and lagging the surface heating by one or two months, then decreasing through September at C10 and October at C12. Stratification is also modified by persistent upwelling when the Gulf of Mexico Loop Current (LC) interacts with the WFS slope at its southwest corner near the Dry Tortugas. Interannual temperature anomalies from the seasonal cycle are also related to how the LC interacts with the WFS slope.

Satellite imagery in evaluating oil spill modelling scenarios for the Syrian oil spill crisis, summer 2021

The second-largest oil pollution incident in the Eastern Mediterranean Levantine basin, following the oil pollution crisis in Lebanon in 2006, is considered to be the oil leakage from the Syrian Baniyas power plant (summer 2021), during which 12,000 tons of oil were released. At the operational phase, the everyday predictions of oil drift were provided using the MEDSLIK and MEDSLIK-II models in the framework of an agreement between the Mediterranean Operational Network for Global Ocean Observing System (MONGOOS) and the Regional Marine Pollution Emergency Response Centre for the Mediterranean (REMPEC). In this work, we present a novel post-operational comprehensive model-based analysis, conducting a SAR validation in two model outputs: the MEDSLIK and the OpenDrift models. Each simulation is initiated with the operationally acquired EMSA-CSN and ESA SAR images. Moreover, the high-resolution met-ocean fields (CYCOFOS, SKIRON) are used to force the oil drift and transformation in both models. The spill was developed under the calm-wind conditions that prevailed during the incident. We found that the boundary sea currents developed on the periphery of the Lattakia eddies (anticyclonic and cyclonic) were responsible for the fast westward spreading of the oil spill offshore in the NE Levantine, the north-south pathway bifurcation, and re-landing of oil in the extended coastal area of Lattakia. Model outputs were validated against Synthetic-aperture radar (SAR) images with appropriate performance metrics, used for the first time, to assess the capacity of a reliable representation of oil spill drift. The intercomparison between the two oil spill models indicated that both models produce almost similar results, while their validation against the satellite SAR observations illustrates moderate accuracy.

Ocean 2D eddy energy fluxes from small mesoscale processes with SWOT

Abstract. We investigate ocean dynamics at different scales in the Agulhas Current system, a region of important interocean exchange of heat and energy. While ocean observations and some of the most advanced climate models capture the larger mesoscale dynamics (> 100 km), the smaller-scale fronts and eddies are underrepresented. The recently launched NASA–CNES Surface Water and Ocean Topography (SWOT) wide-swath altimeter mission observes the smaller ocean geostrophic scales down to 15 km in wavelength globally. Here we will analyse different eddy diagnostics in the Agulhas Current region and quantify the contributions from the larger mesoscales observable today and the smaller scales to be observed with SWOT. Surface geostrophic diagnostics of eddy kinetic energy, strain, and energy cascades are estimated from modelled sea surface height (SSH) fields of the Massachusetts Institute of Technology general circulation model (MITgcm) latitude–longitude polar cap (LLC4320) simulation subsampled at 1/10∘. In this region, the smaller scales (<150 km) have a strong signature on the horizontal geostrophic strain rate and for all eddy diagnostics in the Western Boundary Current and along the meandering Agulhas Extension. We investigate the horizontal cascade of energy using a coarse-graining technique, and we observe that the wavelength range where the inverse cascade occurs is biased towards larger mesoscale wavelengths with today’s altimetric sampling. We also calculate the projected sampling of the eddy diagnostics under the SWOT swaths built with the NASA–CNES simulator to include the satellite position and realistic noise. For the swaths, a neural network noise mitigation method is implemented to reduce the residual SWOT random error before calculating eddy diagnostics. In terms of SSH, observable wavelengths of 15 to 20 km are retrieved after neural network noise mitigation, as opposed to wavelengths larger than 40 km before the noise reduction.

Preliminary assessment of carbonic acid dissociation constants: Insights from observations in China's east coastal oceans

Based on observations in China's east coastal oceans, we conducted a preliminary assessment of 16 sets of carbonic acid dissociation constants (K1* and K2*) by comparing spectrophotometrically measured pH values at 25 °C with those calculated from total alkalinity and dissolved inorganic carbon. We obtained that K1* and K2* often performed differently within different salinity ranges, and that the constants of Millero et al. (2002) (M02) demonstrated the best performance for the salinity range of 24–35. In contrast, the often recommended constants of Mehrbach et al. (1973) refit by Dickson and Millero (1987) (DM87-M) and Lucker et al. (2000) (L00) would underestimate pH at salinities of 24–30. This was mainly associated with the higher product of K1* and K2* by DM87-M and L00 than by M02 at this salinity range. Also, we found almost no differences between pH values calculated with DM87-M and L00.

On the potential of using Distributed Acoustic Sensing (DAS) to build teleseism based ocean-tomography network

The sound from marine teleseisms rank among some of the loudest in Earth’s Oceans, typically radiated into the ocean near the epicenter and propagating across the entire ocean basin. Acoustic sensing on ocean observatories provide a critical path to better understand origins of these low-frequency sound waves known as T-phases, and structure of the ocean over the megameter propagation through the ocean. Of the emerging technologies implemented on these observatories, distributed acoustic sensing (DAS) has the potential to revolutionize marine seismoacoustics. DAS can be implemented on unused oceanic fibers on the continental slopes surrounding an ocean basin could readily enable a teleseism based ocean-tomography network with surprisingly little new infrastructure. In November 2021, a DAS system installed on two runs of fiber optic cable that connect to an ocean observatory measuring offshore volcanic activity, detected a few examples of teleseismic T-phases. Combined with other hydroacoustic ocean observatories in the Pacific Ocean, the T-phase origins are localized to bathymetric features over 100 km from the epicenter, an important consideration when utilizing these T-phases as sources of opportunity to study the ocean.

Detecting the Kauai source beacon with ocean observatories innitiative hydrophones

In March of 2023, a sound source off the coast of Kauai began regularly transmitting a 75 Hz signal into the ocean, six times a day, every fourth day. The regular transmission of this signal allows for the long-term measurements of ocean basin scale temperature changes using the technique of acoustic tomography, as well as further investigations of long-distance acoustic propagation. The Ocean Observatory Initiative (OOI) has, as part of its suite of oceanographic sensors, 11 hydrophones that are continually recording ambient sound. These hydrophones are in various diverse environments in the north-east pacific including bottom mounted in shallow environments, deep environments, on top of seamounts, and moored at 200 m below the ocean surface. In this talk, we present preliminary results of acoustic propagation modeling and data analysis towards receiving the Kauai signal with the OOI hydrophones. The diverse environments of the hydrophones allow us to investigate how different levels of ambient sound and environmental factors, such as proximity to the sound channel affect the ability to detect a positive reception of the Kauai source across the ocean basin. [Work supported by ONR.]

Forward looking sonar applications for unmanned vehicles

Unmanned vehicles have the potential to greatly expand our ability to observe dynamic ocean environments. Whereas distributed acoustic sensors can monitor discrete measurement points in time and space, autonomous maritime platforms provide increased flexibility. Acoustic principles are a key component of this functionality and forward looking sonar (FLS) systems can acquire acoustically derived data products while enabling better informed autonomy behaviors based on enhanced situational and environmental awareness. This paper explores applications of forward look sonar technology when integrated with underwater vehicles. Utility of both interferometric and two-dimensional acoustic data products are examined to illustrate the impact on technologies associated with navigational drift mitigation and ocean observation. Results from sonar platform integration and in-water acquisition for development of an underwater terrain-aided navigation (UTAN) particle filter, an acoustic enabled simultaneous localization and mapping (SLAM) algorithm, obstacle avoidance techniques, and seabed observations are presented.

The role of in situ ocean data assimilation in ECMWF subseasonal forecasts of sea‐surface temperature and mixed‐layer depth over the tropical Pacific ocean

AbstractThe tropical Pacific plays an important role in modulating the global climate through its prevailing sea‐surface temperature spatial structure and dominant climate modes like El Niño–Southern Oscillation (ENSO), the Madden–Julian Oscillation (MJO), and their teleconnections. These modes of variability, including their oceanic anomalies, are considered to provide sources of prediction skill on subseasonal timescales in the Tropics. Therefore, this study aims to examine how assimilating in situ ocean observations influences the initial ocean sea‐surface temperature (SST) and mixed‐layer depth (MLD) and their subseasonal forecasts. We analyze two subseasonal forecast systems generated with the European Centre for Medium‐Range Weather Forecasts (ECMWF) Integrated Forecast System (IFS), where the ocean states were initialized using two Observing‐System Experiment (OSE) reanalyses. We find that the SST differences between forecasts with and without ocean data assimilation grow with time, resulting in a reduced cold‐tongue bias when assimilating ocean observations. Two mechanisms related to air–sea coupling are considered to contribute to this growth of SST differences. One is a positive feedback between the zonal SST gradient, pressure gradient, and surface wind. The other is the difference in Ekman suction and mixing at the Equator due to surface wind‐speed differences. While the initial mixed‐layer depth (MLD) can be improved through ocean data assimilation, this improvement is not maintained in the forecasts. Instead, the MLD in both experiments shoals rapidly at the beginning of the forecast. These results emphasize how initialization and model biases influence air–sea interaction and the accuracy of subseasonal forecasts in the tropical Pacific.

Using ocean ambient sound to sense arrival time fluctuations due to temperature

Ambient noise interferometry is a technique that uses coherent ambient sound measured at two separate locations to estimate the Green’s function between the sensors. Specifically, if the ambient sound is diffuse, the cross-correlation between sensors converges to the Green’s function between the sensors. In this talk, we apply the technique of ambient noise interferometry to two Ocean Observatories Initiative hydrophones that are separated by 3.2km and bottom mounted at a depth of 1500 m. It has previously been shown that these hydrophones can passively estimate several multi-path propagation peaks reliably throughout the 8 years that they have been recording ambient sound. We present an algorithm that estimates the arrival time of these peaks using the empirical Green’s function and compare these estimated arrivals to simulated acoustic arrivals. We demonstrate that the arrival times estimated with ambient sound show good agreement with the simulated results and have clear annual fluctuations due to seasonal temperature changes in the water column. We will discuss the future possibilities of extending this work for inversion based oceanographic measurements. [work supported by ONR.]

A cabled ocean acoustic array to be installed in the Gulf of Maine

It is often noted that we know more about the planets than the processes of our oceans. A network of ocean observatories is improving our knowledge; however, measurements are still relatively scarce and do not provide enough data to validate and refine models of ocean currents, soundscapes, and biological activity. A cabled acoustic array is being developed for deployment in 2024 in the coastal waters of the Gulf of Maine to complement existing oceanographic monitoring infrastructure. The system will make measurements of the nearshore (6 km) environment and shed light on the connections between coastal measurements and the overall Gulf of Maine environment. The system is designed to synoptically collect acoustic, oceanographic, and biogeochemical data, while also remaining flexible enough in design to incorporate new sensors in the future. Measurements made with this ocean asset will serve as a baseline for pattern and trend analyses of changing environmental conditions in an area of high productivity, human use, and species diversity. The data management and visualization team are developing an infrastructure to provide public access to the data and its products. Public data access will permit researchers at all levels to advance our understanding of ocean acoustics and oceanographic processes.

Call for papers: Special Issue on Sensors, Actuators, and Energy Harvesting for Sustainable Ocean Observing Systems

Call for papers: Special Issue on Sensors, Actuators, and Energy Harvesting for Sustainable Ocean Observing Systems

Preliminary Results of Marine Gravity Recovery by Tiangong-2 Interferometric Imaging Radar Altimeter

This paper presents for the first time the results of marine gravity recovery using the ocean observation data acquired by Tiangong-2 interferometric imaging radar altimeter (TG2 InIRA) which demonstrate not only the balanced accuracies of the north and east components of deflection of the vertical (DOV) as envisaged, but also the improved spatial resolutions of DOV compared with that by conventional altimeters (CAs). Moreover, much higher measurement efficiency owing to the wide-swath capability and the great potential in accuracy improvement of marine gravity field are also demonstrated. TG2 InIRA adopts the interferometry with short baseline and takes small incidence angles, by which wide-swath sea surface height (SSH) can be measured with high accuracy. Gravity recovery experiments in the Western Pacific area are conducted to demonstrate the performance, advantages and capability of TG2 InIRA. SSH data processing algorithms and DOV calculation have been designed by taking the wide-swath feature into account, based on which, the gravity anomalies are then calculated using the inverse Vening Meinesz formula. The derived gravity anomalies are compared with both the published gravity models and the shipborne gravity measurements. The results show that the accuracy of TG2 InIRA is equivalent to, or even a little better than, that of CAs. The fused gravity result using equal TG2 InIRA data and CAs data performs better than those using TG2 InIRA data alone or CAs data alone. Due to the signal bandwidth of TG2 InIRA is only 40 MHz which is much smaller than that of CAs, much higher accuracy can be hopefully achieved for future missions if larger signal bandwidth is used.

Inputs, amplification and sinks of perfluoroalkyl substances at coastal Antarctica

The sources, biogeochemical controls and sinks of perfluoroalkyl substances, such as perfluoroalkyl acids (PFAAs), in polar coastal regions are largely unknown. These were evaluated by measuring a large multi-compartment dataset of PFAAs concentrations at coastal Livingston and Deception Islands (maritime Antarctica) during three austral summers. PFAAs were abundant in atmospheric-derived samples (aerosols, rain, snow), consistent with the importance of atmospheric deposition as an input of PFAAs to Antarctica. Such PFAAs deposition was unequivocally demonstrated by the occurrence of PFAAs in small Antarctic lakes. Several lines of evidence supported the relevant amplification of PFAAs concentrations in surface waters driven by snow scavenging of sea-spray aerosol-bound PFAAs followed by snow-melting. For example, vertical profiles showed higher PFAAs concentrations at lower-salinity surface seawaters, and PFAAs concentrations in snow were significantly higher than in seawater. The higher levels of PFAAs at Deception Island than at Livingston Island are consistent with the semi-enclosed nature of the bay. Concentrations of PFOS decreased from 2014 to 2018, consistent with observations in other oceans. The sink of PFAAs due to the biological pump, transfer to the food web, and losses due to sea-spray aerosols alone are unlikely to have driven the decrease in PFOS concentrations. An exploratory assessment of the potential sinks of PFAAs suggests that microbial degradation of perfluoroalkyl sulfonates should be a research priority for the evaluation of PFAAs persistence in the coming decade.

An evaluation of eight global ocean reanalyses for the Northeast U.S. Continental shelf

The Northeast U.S. continental Shelf (NES) extending from the Gulf of Maine to Cape Hatteras, is a dynamic region supporting some of the most commercially valuable fisheries in the world. This study aims to provide a systematic assessment of eight widely used, intermediate-to-high spatial resolution global ocean reanalysis products (CFSR, ECCO, ORAS, SODA, BRAN, GLORYS, GOFS3.0, and GOFS3.1) against available in situ and satellite ocean observations. In situ observations include water level from tide gauges, and temperature and salinity from various sources including shipboard hydrographic data, and moorings on the NES. Overall, the coarser resolution products exhibit limited skill in the coastal environment, with the high-resolution products better representing the temperature and salinity on the NES. Common biases are found in all reanalyses and in some regions within the NES; for example, biases in temperature and salinity are larger in the southern Mid-Atlantic Bight than in the rest of the NES. There is no single reanalysis that performs well across all parameters in all regions within the NES, but GLORYS and BRAN stand out for their superior performance across the largest number of metrics, outperforming other products in 22 and 25 of the 65 metrics examined, respectively. SODA is the top performer among the coarser resolution products (CFSR, ECCO, ORAS and SODA). The Gulf Stream and local bathymetry are critical factors leading to differences between the reanalyses. Conditions in summer are less well represented than in winter. In particular, the Mid-Atlantic Bight Cold Pool is not reproduced in four (CFSR, ECCO, ORAS, BRAN) of the eight reanalyses.

MSGNet: multi-source guidance network for fish segmentation in underwater videos

Fish segmentation in underwater videos provides basic data for fish measurements, which is vital information that supports fish habitat monitoring and fishery resources survey. However, because of water turbidity and insufficient lighting, fish segmentation in underwater videos has low accuracy and poor robustness. Most previous work has utilized static fish appearance information while ignoring fish motion in underwater videos. Considering that motion contains more detail, this paper proposes a method that simultaneously combines appearance and motion information to guide fish segmentation in underwater videos. First, underwater videos are preprocessed to highlight fish in motion, and obtain high-quality underwater optical flow. Then, a multi-source guidance network (MSGNet) is presented to segment fish in complex underwater videos with degraded visual features. To enhance both fish appearance and motion information, a non-local-based multiple co-attention guidance module (M-CAGM) is applied in the encoder stage, in which the appearance and motion features from the intra-frame salient fish and the moving fish in video sequences are reciprocally enhanced. In addition, a feature adaptive fusion module (FAFM) is introduced in the decoder stage to avoid errors accumulated in the video sequences due to blurred fish or inaccurate optical flow. Experiments based on three publicly available datasets were designed to test the performance of the proposed model. The mean pixel accuracy (mPA) and mean intersection over union (mIoU) of MSGNet were 91.89% and 88.91% respectively with the mixed dataset. Compared with those of the advanced underwater fish segmentation and video object segmentation models, the mPA and mIoU of the proposed model significantly improved. The results showed that MSGNet achieves excellent segmentation performance in complex underwater videos and can provide an effective segmentation solution for fisheries resource assessment and ocean observation. The proposed model and code are exposed via Github1.