Search

Highlights • The 2019 eruption of Volcano-F was VEI 3, sub-Plinian, dacitic, and likely lasted less than a day. • Products are geochemically similar to 2001 and to a previously unattributed pumice raft in 1964. • Volcano-F has therefore had 3 explosive eruptions in 60 years and is significantly at unrest. • Volcano-F is one of several potentially hazardous calderas likely in the caldera recharge phase. • These calderas need monitoring and further study to understand the risk posed. Abstract Low volcanic explosivity index (VEI) eruptions are common occurrences in the Southwest Pacific but, as demonstrated by the 2021/2022 eruption of Hunga Volcano, submerged calderas in the region are also capable of producing much larger and more hazardous eruptions. As such, characterising smaller events from potentially hazardous systems is essential. The 2019 eruption of Volcano-F, a submerged caldera, would likely have gone totally undetected had it not produced a pumice raft that inundated beaches in Fiji and eventually washed up in Australia. New data, acquired 5 months after the eruption, reveal the development of a new vent and the accumulation of at least 3.1*107 m3 bulk volume (dense rock equivalent of 5.6*106 m3) of material on the seafloor. Between 30 and 70% of erupted material entered the raft, while the rest remained near to or was dispersed down-current of the vent. This previously unaccounted for material increases the volume estimate for the eruption, confirming it as a VEI 3 event and highlights the importance of considering not just the floating component of a pumice raft forming eruption for VEI estimation. Geochemical analysis reveals the eruption comprised a homogenous batch of dacitic magma, with compositional characteristics similar to that erupted from the same volcano in 2001, and an until-now-unidentified pumice raft in the Coral Sea in 1964. Volcano-F therefore appears to have had at least three explosive eruptions in the last 60 years, indicating it is

Anaerobic methanogenesis is dependent on key macronutrients (carbon, nitrogen and phosphorus) and trace metals (including iron, nickel and cobalt) to drive methane production. Reservoir derived methane emissions have correlated to eutrophication status, with elevated emissions associated with more eutrophic systems. Additionally, sediment organic matter can enhance methane emissions, particularly through the ebullition pathway. As such, it is critical to understand how organic carbon and nutrient inputs into reservoir water columns and sediments drive methanogenesis to improve flooded land greenhouse gas emission estimates. In this study we examine the methane potential of sediments in mesotrophic (Little Nerang Dam) and eutrophic (Lake Wivenhoe) sub-tropical reservoirs under different nutrient and organic carbon availabilities using biological methane potential (BMP) tests. BMP tests were conducted with sediments incubated under anaerobic conditions using replicate controls (reservoir bottom waters) or treatments (excess nutrient and/or organic carbon availability). The results indicated that these systems are carbon limited. The addition of organic carbon significantly increases anaerobic methanogenesis by 20-fold over controls. Analysis of sediment samples from the reservoirs showed that both reservoirs were replete in key macronutrient and trace metal content for methanogenesis. Finally, a comprehensive catchment monitoring program of Little Nerang Dam measuring catchment inflow events, lateral transport of forest litter, and bulk atmospheric deposition showed that catchment inflows and lateral transport of forest litter were strongly linked to rainfall and accounted for more than 99% of the total annual load. This suggests the frequency of rainfall events is a critical driver of organic matter inputs that drive reservoir methane emissions in the humid, sub-tropical region.

This paper presents a path planning methodology which enables Autonomous Underwater Vehicles (AUVs) to navigate in shallow complex environments such as coral reefs. The approach leverages prior information from an aerial photographic survey, and derived bathymetric information of the corresponding area. From these prior maps, a set of features is obtained which define an expected arrangement of objects and bathymetry likely to be perceived by the AUV when underwater. A navigation graph is then constructed by predicting the arrangement of features visible from a set of test points within the prior, which allows the calculation of the shortest paths from any pair of start and destination points. A maximum likelihood function is defined which allows the AUV to match its observations to the navigation graph as it undertakes its mission. To improve robustness, the history of observed features are retained to facilitate possible recovery from non-detectable or misclassified objects. The approach is evaluated using a photo-realistic simulated environment, and results illustrate the merits of the approach even when only a relatively small number of features can be identified from the prior map.

This paper provides an overview of demonstrated and emerging work applying Uncrewed Maritime Systems (UMS) to coral reef conservation. It blends theoretical analysis and current field results with a focus on recent developments in Australia. Here a combination of Autonomous Underwater Vehicles (AUVs) and Uncrewed Surface Vehicles (USVs) are being used to support restoration of coral communities on the Great Barrier Reef. In 2018 and 2019, UMS were evaluated to assist in the 'planting' of coral larvae during the reef's annual sexual reproduction periods. Scientific results on actual coral larval settlement and recruitment, and hence coral community restoration using these techniques is showing promise but are beyond the scope of this paper. This paper focuses on the technical challenges and opportunities around employing AUVs and USVs for reef restoration at scale. These include: reduction in platform cost, management of larvae supply, precision of larvae placement and user interface/autonomy. This paper examines the field use of the existing prototype robots from recent on-reef trials and hypothesizes solutions to increase efficiency of larval distribution.

Shipping emissions are known to affect communities in coastal locations, especially near harbours. This study monitored the air quality near the premier cruise ship terminal in Melbourne over a continuous period of 98 days during the peak cruise ship season in Australia. As shipping emission plumes are intermittent and fluctuate spatially, they cannot be detected accurately by a single fixed monitor. To overcome this limitation, we deployed seven units of the low-cost KOALA air quality monitor, which measures PM2.5 and CO concentrations in real time and then transmits the data via 3G to an in-cloud database, in a spatially distributed configuration, four at ground level and three on the upper balconies of two high-rise apartment blocks. The time profile showed numerous spikes in the PM2.5 concentration, some of which exceeded 200 µg m– 3 for periods of 5–10 min, coinciding with ship movements. On average, the spikes were ~4–5 times above the normal background value (~10 µg m–3). Because of their very short duration, these episodes did not significantly raise the 24-h averages at any of the locations; however, they increased the number of days on which these values exceeded the limit specified by the national air quality standard, resulting in more exceedance days for the monitored area than the nearest air quality station. Although the long-term health effects of elevated PM concentrations are known, few studies have been conducted on the risks of short-term exposures to extreme spikes.

This paper presents two novel approaches for improving image-based underwater obstacle detection by combining sparse stereo point clouds with monocular semantic image segmentation. Generating accurate image-based obstacle maps in cluttered underwater environments, such as coral reefs, are essential for robust robotic path planning and navigation. However, these maps can be challenged by factors including visibility, lighting and dynamic objects (e.g. fish) that may lead to falsely identified free space or dynamic objects which trajectory planners may react to undesirably. We propose combining feature-based stereo matching with learning-based segmentation to produce a more robust obstacle map. This approach considers direct binary learning of the presence or absence of underwater obstacles, as well as a multiclass learning approach to classify their distance (near, mid and far) in the scene. An enhancement to the binary map is also shown by including depth information from sparse stereo matching to produce 3D obstacle maps of the scene. The performance is evaluated using field data collected in cluttered, and at times, visually degraded coral reef environments. The results show improved image-wide obstacle detection, rejection of transient objects (such as fish), and range estimation compared to feature-based sparse and dense stereo point clouds alone.

Robotic inspection often relies on building custom platforms for each new deployment; this is a luxury that urban search and rescue (USAR) robots do not have when time is of critical importance. A significant factor for robots deployed in disaster areas is the varying size of voids and access ways in their path. These situations require platforms that can quickly reconfigure on location. With these challenges in mind, we present the NeWheel system: An in‐field reconfigurable robotic platform that allows mobility changes before, and during, deployment. The NeWheel platform also has the advantage of being small enough to be person‐deployable and to travel as checked luggage on a commercial flight. This field report presents the results and learnings from three field trips on Peel Island located off the coast of Brisbane, Australia. These field trips featured the deployment of the NeWheel system in multiple configurations to inspect and map inside historic dilapidated buildings. It demonstrates the potential of the NeWheel in buildings cluttered with debris and with unstable flooring whether they are historically important or in USAR contexts.

This paper presents a high-fidelity autonomous surface vehicle (ASV) and an above-water environment simulator developed for, and used by, the Queensland University of Technology (QUT) team at the 2016 Maritime RobotX Challenge. This simulator was developed to address a gap in an existing robotic marine vehicle simulation software by producing high-fidelity sensor data from a realistic operating environment and simulating ASV motion dynamics as a direct interface for hardware-in-the-loop simulation. The key functions required for the simulator to be an effective model of the QUT ASV included a high-fidelity camera, LiDAR (Velodyne HDL-32E), Inertial Measurement Unit (IMU), and GPS sensor simulation, as well as buoyancy and physics simulation for vehicle motion and control performance in a spatiotemporally varying modeled sea state. Another key feature of the simulator was the provision to directly interface the Robotic Operating System (ROS) to publish and subscribe to sensor topics (in real time) and allow hardware-in-the-loop simulation of the ASV systems for algorithm development. The fidelity and performance of the simulator was benchmarked against the existing best available alternative simulator, V-Rep, with results shown to demonstrate that the autonomous marine surface vessel simulator achieved improved computational performance and sensor fidelity when simulating above surface marine environments.

This paper describes a novel probabilistic sensor fusion framework aimed at improving obstacle detection accuracy and classification of various targets experienced in the Maritime RobotX Challenge. In both the 2014 and 2016 Maritime RobotX Challenges, it was found that detecting obstacles using LIDAR only and classifying obstacles using vision only can be challenged by environmental conditions, such as glare from the sun or by objects such as the spherical black buoys from the obstacle field that disperses LIDAR rays. In this paper, a new multimodal sensor fusion approach is proposed that combines data streams from perception sensors, such as LIDAR, RADAR, and cameras, to improve the robustness of detection and classification performance over a single sensor method. Using data collected from both the 2014 and 2016 Maritime RobotX Challenges, an evaluation of the perception framework is provided. The proposed detection and classification framework is now being transferred to the Queensland University of Technology (QUT) autonomous surface vehicle to improve overall mapping accuracy and task execution.

Emissions from flooded land represent a direct source of anthropogenic greenhouse gas (GHG) emissions. Methane emissions from large, artificial water bodies have previously been considered, with numerous studies assessing emission rates and relatively simple procedures available to determine their surface area and generate upscaled emissions estimates. In contrast, the role of small artificial water bodies (ponds) is very poorly quantified, and estimation of emissions is constrained both by a lack of data on their spatial extent and a scarcity of direct flux measurements. In this study, we quantified the total surface area of water bodies amp;lt; 10(5) m(2) across Queensland, Australia, and emission rates from a variety of water body types and size classes. We found that the omission of small ponds from current official land use data has led to an underestimate of total flooded land area by 24 %, of small artificial water body surface area by 57% and of the total number of artificial water bodies by 1 order of magnitude. All studied ponds were significant hotspots of methane production, dominated by ebullition (bubble) emissions. Two scaling approaches were developed with one based on pond primary use (stock watering, irrigation and urban lakes) and the other using size class. Both approaches indicated that ponds in Queensland alone emit over 1.6 Mt CO2 eq.yr(-1), equivalent to 10% of the states entire land use, land use change and forestry sector emissions. With limited data from other regions suggesting similarly large numbers of ponds, high emissions per unit area and under-reporting of spatial extent, we conclude that small artificial water bodies may be a globally important missing source of anthropogenic greenhouse gas emissions.