Under-ice Environment Research Articles

A real time data-driven dynamic glasius bionic neural network path planning algorithm for polar under-ice feature scanning by “Xinghai 1000” AUV

Using Autonomous Underwater Vehicle (AUV) to survey polar under-ice feature is crucial for studying polar marine environment and global climate change. Compared with seafloor terrain scanning, it faces several challenges, such as uncertain under-ice environment, lack of reliable positioning methods, and stringent constraints on recovery location. Thus, taking “Xinghai 1000” as analysis object, this paper proposes a real time data-driven dynamic glasius bionic neural network path planning algorithm (RDGPP), which consists of an online bi-level path planning module. In global planning module, dynamically hierarchical strategy is used to realize task area layering. On this basis, global coverage path is planned by combining neuron activity value updated which utilize real-time environmental information obtained by FLS sonar with multivariant strategy. Moreover, due to AUV recovery mode limitation, return strategy has been devised in multivariant strategy to ensure AUV automatic return to launch point, thereby reducing energy usage. In local planning module, Flagged bit and GBNN-driven A-star algorithm (FGA) is introduced to minimize deadlock escape time for AUV. Simulation experiments verify RDGPP algorithm advancement in key indicators such as turning times, running time, etc., and effectiveness is verified by “Xinghai 1000” under-ice experiment.

Robotic exploration beneath the ice : the challenges, risks and rewards of deploying an autonomous underwater vehicle in Antarctica

Measurements beneath Antarctic ice shelves are critical to our understanding of how the glaciers will change and melt. To access these regions, we rely on autonomous underwater vehicles (AUV), which are free-swimming robots. The Australian Maritime College in Launceston, Tasmania, is leading efforts to deploy these vehicles in some of the harshest environments on Earth. This paper provides an overview of the history and recent Australian efforts in deploying AUVs in under-ice environments and discusses recent advances in mission planning and the ongoing challenges to take measurements from beneath ice shelves.

Topology Optimization of Long-Thin Sensor Networks in Under-Ice Environments

Long-thin network topologies are useful in monitoring tunnels and bridges because of their linear structure. In this paper, we apply the long-thin concept to under-ice acoustic sensor networks in the Arctic Ocean. We develop an acoustic wave propagation model for the under-ice environments using the ray-tracing method. Specifically, we derive the reflection coefficients of plane acoustic waves penetrating from seawater to ice in a ray-based model. The impulse responses and transmission loss, including their dependencies on the communication range, are examined in the under-ice environment. Furthermore, to maximize the operational lifetime of the under-ice long-thin network, we optimize the network topology by integrating Monte Carlo simulations with a differential evolution technique. The optimization results are presented for long-thin sensor network scenarios under the Arctic sea ice. Computer simulations show that by optimizing the sensor locations in the long-thin topology, one can take advantage of the sound convergent and divergent effects of the under-ice acoustic channel. As a result, the energy consumption of the long-thin sensor network can be significantly reduced so that its operational lifetime can be extended in the Arctic environment.

MIMO Underwater Acoustic Communication in Shallow Water with Ice Cover

Although multiple-input multiple-output (MIMO) underwater acoustic (UWA) communication has been intensively investigated in the past years, existing works mainly focus on open-water environment. There is no work reporting MIMO acoustic communication in under-ice environment. This paper presents results from a recent MIMO acoustic communication experiment which was conducted in Bohai Gulf during winter. In this experiment, high frequency MIMO signals centered at 10 kHz were transmitted from a two-element source array to a four-element vertical receiving array at 1 km range. According to the received signal of different array elements, MIMO acoustic communication in under-ice environment suffers less effect from co-channel interference compared with that in open-water environment. In this paper, time reversal followed by a single channel decision feedback equalizer is used to process the experimental data. It is demonstrated that this simple receiver is capable of realizing robust performance using fewer hydrophones (i.e. 2) without the explicit use of complex co-channel interference cancelation algorithms, such as parallel interference cancelation or serial interference cancelation.

Reinforcement Learning-Based Multi-AUV Adaptive Trajectory Planning for Under-Ice Field Estimation.

This work studies online learning-based trajectory planning for multiple autonomous underwater vehicles (AUVs) to estimate a water parameter field of interest in the under-ice environment. A centralized system is considered, where several fixed access points on the ice layer are introduced as gateways for communications between the AUVs and a remote data fusion center. We model the water parameter field of interest as a Gaussian process with unknown hyper-parameters. The AUV trajectories for sampling are determined on an epoch-by-epoch basis. At the end of each epoch, the access points relay the observed field samples from all the AUVs to the fusion center, which computes the posterior distribution of the field based on the Gaussian process regression and estimates the field hyper-parameters. The optimal trajectories of all the AUVs in the next epoch are determined to maximize a long-term reward that is defined based on the field uncertainty reduction and the AUV mobility cost, subject to the kinematics constraint, the communication constraint and the sensing area constraint. We formulate the adaptive trajectory planning problem as a Markov decision process (MDP). A reinforcement learning-based online learning algorithm is designed to determine the optimal AUV trajectories in a constrained continuous space. Simulation results show that the proposed learning-based trajectory planning algorithm has performance similar to a benchmark method that assumes perfect knowledge of the field hyper-parameters.

Reflecting and scattering of acoustic wave from sea ices

In order to build an efficient underwater acoustic sensor network in the Arctic Ocean environment, transmission characteristics of under-ice acoustic channels need comprehensive understanding. The reflecting and scattering of acoustic waves from sea ices have great influences on under-ice acoustic channels. Both topology and structure of sea surface ices are very complex and variable. The physical dimension, acoustic property and interface roughness of sea ices depend not only on local environment, but also on climate and formation time. Therefore, it is of great significance to develop a model of reflecting and scattering of acoustic waves from sea ices for investigating the sound propagation in the under-ice environment. Assuming that sea ices are a multi-layered elastic solid medium and the ice-water interface is rough and satisfies the boundary condition of perturbation, we develop a system of linear equations to solve the coherent reflection coefficient of the incident sound wave from water to sea ice. The coherent reflection coefficient is a function of the frequency of sound wave and incident grazing angle, and is numerically evaluated. The influences of ice thickness and ice-water interface roughness on the coherent reflection coefficient are analyzed. Furthermore, the method of calculating scattering coefficient by using the power spectrum density of the scattering field is introduced. The scattering coefficient as a function of the scattering grazing angle is numerically evaluated. The influences of ice thickness and ice-water interface roughness on scattering coefficient are analyzed. The results show that both the coherent reflection coefficient and the scattering coefficient are dependent on the frequency of acoustic wave, ice thickness and grazing angle. The coherent reflection coefficient is close to 1.0 and the scattering coefficient is less than 0.01 when incident grazing angle is less than 15°. In addition, the frequency of acoustic wave and ice thickness have weak influences on them. However, the frequency of acoustic wave and ice thickness have significant influences on the coherent reflection coefficient and the scattering coefficient when the incident grazing angle is big, say, greater than 30°. In general, the thicker the ice is, the smaller the coherent reflection coefficient and the scattering coefficient are. The coherent reflection coefficient is less than 0.18 when the ice thickness is greater than 10.0 m and the frequency of acoustic waves is greater than 2 kHz. The ice-water interface roughness has great influences on both the coherent reflection coefficient and the scattering coefficient. The rougher of the ice-water interface is, the smaller the coherent reflection coefficient is, and the bigger the scattering coefficient is.

Future Challenges in Southern Ocean Ecology Research

The Southern Ocean is experiencing relentless change. The Antarctic and Southern Ocean community, represented by 75 scientists and policy-makers from 22 countries, recently met to formulate a collective vision on the priorities for Antarctic research for the next two decades and beyond. Here, we assess high-interest research areas related specifically to Southern Ocean life and ecology that, although not all retained as the 80 top priorities among the addressed scientific domains, are of considerable relevance to the biology and ecology of the Southern Ocean. As certain regions of the Southern Ocean ecosystems have witnessed abiotic and biotic changes in the last decades (e.g. warming, changes in sea ice and abundance of marine organisms), such an exercise was urgently needed. We concluded that basic biological information on the taxonomy of numerous organisms is still lacking in areas such as the deep-ocean floor or the under-ice environments. Furthermore, there is a need for knowledge about the response and resilience of Antarctic marine ecosystems to change. The continuation of a long-term commitment and the development and use of innovative technology to adequately monitor the Southern Ocean ecosystems is required. Highlighting the most important Southern Ocean research topics allow the identification of the challenges and future requirements in technological development, and both research and funding strategies for the various stakeholders

Sexual dimorphism in an under-ice spawning fish: the burbot (Lota lota)

Sexual dimorphic characteristics arise in response to differing selective pressures on the sexes and can be used to attract mates or signal reproductive readiness. How sexual dimorphism is expressed where visual cues may be of limited use is an underexplored aspect of reproductive ecology. The burbot (Lota lota (L., 1758)) is a common boreal fish that is not overtly sexually dimorphic. It spawns mid-winter in a light-limited under-ice environment. We examined a variety of morphological and reproductive characteristics in burbot from a northern lake over one full year to assess both seasonal and sex-based variation. Spawning occurred under ice in early February. Seasonal variation was more pronounced in females for many of the traits examined. Growth, fin lengths, swim bladder mass, and liver lipid concentration did not differ between the sexes. Male burbot had significantly higher body condition, larger gas glands, and smaller livers. Males also had significantly larger gonads than females, unusual for boreal fishes. The high gonadal investment of male burbot suggests that sperm competition may be intense in this species. This study demonstrates that sexual dimorphism can be subtle and is present in a seemingly monomorphic species—the burbot.

Minimizing position uncertainty for under-ice autonomous underwater vehicles

Localization underwater has been known to be challenging due to the limited accessibility of the Global Positioning System (GPS) to obtain absolute positions. This becomes more severe in the under-ice environment since the ocean surface is covered with ice, making it more difficult to access GPS or to deploy localization infrastructure. In this paper, a novel solution that minimizes localization uncertainty and communication overhead of under-ice Autonomous Underwater Vehicles (AUVs) is proposed. Existing underwater localization solutions generally rely on reference nodes at ocean surface or on localization infrastructure to calculate positions, and they are not able to estimate the localization uncertainty, which may lead to the increase of localization error. In contrast, using the notion of external uncertainty (i.e., the position uncertainty as seen by others), our solution can characterize an AUV’s position with a probability model. This model is further used to estimate the uncertainty associated with our proposed localization techniques. Based on this uncertainty estimate, we further propose algorithms to minimize localization uncertainty and communication overhead. Our solution is emulated and compared against existing solutions, showing improved performance.

Mapping radiation transfer through sea ice using a remotely operated vehicle (ROV)

Abstract. Transmission of sunlight into and through sea ice is of critical importance for sea-ice associated organisms and photosynthesis because light is their primary energy source. The amount of visible light transferred through sea ice contributes to the energy budget of the sea ice and the uppermost ocean. However, our current knowledge on the amount and distribution of light under sea ice is still restricted to a few local observations, and our understanding of light-driven processes and interdisciplinary interactions is still sparse. The main reasons are that the under-ice environment is difficult to access and that measurements require large logistical and instrumental efforts. Hence, it has not been possible to map light conditions under sea ice over larger areas and to quantify spatial variability on different scales. Here we present a detailed methodological description for operating spectral radiometers on a remotely operated vehicle (ROV) under sea ice. Recent advances in ROV and radiation-sensor technology have allowed us to map under-ice spectral radiance and irradiance on floe scales within a few hours of station time. The ROV was operated directly from the sea ice, allowing for direct relations of optical properties to other sea-ice and surface features. The ROV was flown close to the sea ice in order to capture small-scale variability. Results from the presented data set and similar future studies will allow for better quantification of light conditions under sea ice. The presented experiences will support further developments in order to gather large data sets of under-ice radiation for different ice conditions and during different seasons.

A model for under‐ice‐target acoustic scattering.

A model is described that has been developed to calculate the scatter of a high frequency acoustic pulse that originates from a stationary source and is scattered from a stationary target near the under-ice surface characteristic of the Arctic. Measured under-ice acoustic profile data are used to model the large scale under-ice surface roughness by an ensemble of ice keels. Additional measured data are used to model the small scale roughness of ice keels by an ensemble of ice blocks. The Helmholtz-Kirchhoff integral in the Kirchhoff approximation is used to calculate the scatter from an ice facet. The scatter from an ice keel is given by a coherent sum of the scatter from all the ice facets. A T-matrix is used to calculate the scatter from the stationary target. The first order terms in the multiple scattering process between the target and ice surface is also calculated, i.e., the scatter from the target of the free field scatter from ice facets and the scatter from ice facets of the free field scatter from the target. Numerical results show the extent to which the scatter from the target in the under-ice environment differs from its free field scatter.

SIPEX—exploring the Antarctic sea ice zone

Sea ice in the polar regions plays a key role in both regulating global climate andmaintaining marine ecosystems. The international Sea Ice Physics and EcosystemeXperiment (SIPEX) explored the sea ice zone around Antarctica in September andOctober 2007, investigating relationships between the physical sea ice environment and thestructure of Southern Ocean ecosystems. One of the main goals of SIPEX was to conductlarge-scale sea ice and snow thickness surveys for the validation of satellite-basedmeasurements. SIPEX scientists used a variety of techniques including helicopter-basedradar and laser altimetry, as well as a remotely operated underwater vehicle, to gatherbaseline data on Antarctic sea ice thickness and the under-ice environment. These datawill be invaluable for monitoring possible future changes in the sea ice aroundAntarctica.

Stable Carbon and Nitrogen Isotopic

The perennially ice-covered lakes in the McMurdo Dry Valleys, Antarctica, are part of the coldest and driest ecosystem on earth. To understand lacustrine carbon and nitrogen cycling in this end-member ecosystem, and to define paleolimnological proxies for ice-covered lakes, we measured the stable carbon and nitrogen isotopic composition of particulate organic matter (POM) and benthic organic matter (BOM) within the lakes of Taylor Valley. The δ13C compositions of seasonally ice-free edges of the lakes (moats) are enriched relative to under-ice organic matter. Thus, the organic carbon isotopic composition of buried sediments may be a proxy for sample position within the lake. In the moats, δ13C values are governed by limited CO2diffusion across benthic cyanobacterial cell membranes. During a high glacial melt (2001–2002) season, both δ13CPOM and δ13CBOM in the moats were more depleted than during previous low melt years. We propose that this occurred in response to higher [CO2](aq) and/or reduced growth rates resulting from turbidity-induced light limitation. Though moats and under-ice environments are usually poorly connected, during the 2001–2002 season, the enrichment of the δ13CPOM values at 6 m depth in the stream-proximal sites relative to deep-profile sites implies enhanced connectivity between these environments. The δ13C compositions of BOM and POM profiles in Lake Hoare and Lake Fryxell indicate that these lakes are dominated by benthic productivity. In contrast, in Lake Bonney, the similarity of the δ13C values of BOM and POM indicates the pelagic component dominance in the carbon cycle.

Antarctic krill under sea ice: elevated abundance in a narrow band just south of ice edge.

We surveyed Antarctic krill (Euphausia superba) under sea ice using the autonomous underwater vehicle Autosub-2. Krill were concentrated within a band under ice between 1 and 13 kilometers south of the ice edge. Within this band, krill densities were fivefold greater than that of open water. The under-ice environment has long been considered an important habitat for krill, but sampling difficulties have previously prevented direct observations under ice over the scale necessary for robust krill density estimation. Autosub-2 enabled us to make continuous high-resolution measurements of krill density under ice reaching 27 kilometers beyond the ice edge.

Effects of sea ice cover on the swarming behaviour of Antarctic krill, Euphausia superba

Much of the distribution range of Antarctic krill, Euphausia superba, is covered by permanent or seasonal sea ice. Sea ice extent has been implicated as a major factor affecting reproductive success of krill and krill dispersal, but little is known of the way in which ice cover may influence krill behaviour. This is largely because the under-ice environment is difficult to study. Ship-borne echosounders have, however, detected krill aggregations in midwater in ice-covered regions. We used 120-kHz echograms collected underway during three cruises that crossed ice-covered and adjacent open waters in the Bellingshausen, Weddell, and Scotia seas to compare morphological and next-neighbour characteristics of krill swarms within and without ice cover. No significant differences were detected between the horizontal and vertical extent of swarms or swarm next-neighbour distance in ice-covered or open waters. Distributions of swarm mid-depths did, however, differ significantly between ice-covered and open areas in all three seas, although the direction of difference was not the same in each instance: swarms in the Weddell and Scotia seas were generally shallower under ice than in open water, whereas in the Bellingshausen Sea the opposite prevailed.

Effects of sea ice cover on the swarming behaviour of Antarctic krill, <i>Euphausia superba</i>

Much of the distribution range of Antarctic krill, Euphausia superba, is covered by permanent or seasonal sea ice. Sea ice extent has been implicated as a major factor affecting reproductive success of krill and krill dispersal, but little is known of the way in which ice cover may influence krill behaviour. This is largely because the under-ice environment is difficult to study. Ship-borne echosounders have, however, detected krill aggregations in midwater in ice-covered regions. We used 120-kHz echograms collected underway during three cruises that crossed ice-covered and adjacent open waters in the Bellingshausen, Weddell, and Scotia seas to compare morphological and next-neighbour characteristics of krill swarms within and without ice cover. No significant differences were detected between the horizontal and vertical extent of swarms or swarm next-neighbour distance in ice-covered or open waters. Distributions of swarm mid-depths did, however, differ significantly between ice-covered and open areas in all three seas, although the direction of difference was not the same in each instance: swarms in the Weddell and Scotia seas were generally shallower under ice than in open water, whereas in the Bellingshausen Sea the opposite prevailed.

A multibounce, single‐scatter ray theoretic model for underice predictions

This study describes the propagated and scattered acoustic field from a point source to a point receiver in an approximate underice environment characteristic of the deep Arctic Ocean. Our model considers both the azimuthal and vertical redistribution of scattered energy caused by a rough surface overlying an upwardly refracting, linear sound‐speed profile. The model considers one nonspecular surface scattering per path (connecting the source with the receiver) coupled with any number of specular interactions, each of which suffers losses due to scattering into nonspecular directions, i.e., consistent with conservation of energy. The propagation model is based on ray theory, while the scattering model uses Twersky's scattering cross section for hard or soft planar surfaces roughened by hard or soft, hemispherical bosses. The model calculates intensity as a function of position and direction. Bistatic results at all frequencies show that for hard and soft surfaces the strongest scattered field is produced directly above the submerged source, and that the total scattered energy levels increase as frequency increases. Monostatic geometries and effects of distributions of boss sizes will also be considered. [Work supported by NRL and ONR.]