Advances in the investigation and risk assessment of cold source blockages in nuclear power plants in China

Nuclear Engineering and Design, Journal Year: 2024, Volume and Issue: 420, P. 112998 - 112998

Published: Feb. 14, 2024

In recent years, the issue of cold source blockages in Nuclear Power Plants (NPPs) has gained prominence due to its potential induce disasters, posing economic losses and safety hazards. Despite extensive research, challenges persist effectively monitoring, providing early warnings, assessing risks NPPs. This article consolidates existing literature on investigation risk assessment blockages, categorizing types, outlining attributes, evaluating systems. The findings emphasize need for a comprehensive monitoring warning system, advocating integration various methods into multi-modal network. approach serves as blueprint an inclusive platform, fostering collaborative observation technological synergy enhance warning, case study at Haiyang Plant, analytic hierarchy process (AHP) was used establish system marine organisms, resulting identification highly risky organisms—Ulva lactuca, Sargassum horueri, Rhopilema esculentum, Nemopilema nomurai. aligns with historical instances, validating rationality AHP-based system.

Language: Английский

---

Monitoring of the Environmental Indicators in the Marine Ecosystem

Published: Jan. 1, 2025

Language: Английский

---

Diatom Lensless Imaging Using Laser Scattering and Deep Learning

ACS ES&T Water, Journal Year: 2025, Volume and Issue: unknown

Published: March 24, 2025

We present a novel approach for imaging diatoms using lensless and deep learning. used laser beam to scatter off samples of diatomaceous earth (diatoms) then recorded transformed the scattered light into microscopy images diatoms. The predicted gave an average SSIM 0.98 RMSE 3.26 as compared experimental data. also demonstrate capability determining velocity angle movement from their scattering patterns they were translated through beam. This work shows potential identifying other microsized organisms in situ within marine environment. Implementing such method real-time image acquisition analysis could enhance environmental management, including improving early detection harmful algal blooms.

Language: Английский

---

Edge computing at sea: high-throughput classification of in-situ plankton imagery for adaptive sampling

Frontiers in Marine Science, Journal Year: 2023, Volume and Issue: 10

Published: June 8, 2023

The small sizes of most marine plankton necessitate that sampling occur on fine spatial scales, yet our questions often span large areas. Underwater imaging can provide a solution to this conundrum but collects quantities data require an automated approach image analysis. Machine learning for classification, and high-performance computing (HPC) infrastructure, are critical rapid processing; however, these assets, especially HPC only available post-cruise leading ‘after-the-fact’ view community structure. To be responsive the often-ephemeral nature oceanographic features species assemblages in highly dynamic current systems, real-time key adaptive sampling. Here we used new In-situ Ichthyoplankton Imaging System-3 (ISIIS-3) Northern California Current (NCC) conjunction with edge server classify imaged into 170 classes. This capability together visualization heavy.ai dashboard makes decision-making at sea possible. Dual ISIIS-Deep-focus Particle Imager (DPI) cameras sample 180 L s -1 , >10 GB video per min. Imaged organisms size range 250 µm 15 cm include abundant crustaceans, fragile taxa (e.g., hydromedusae, salps), faster swimmers krill), rarer larval fishes). A deep pipeline deployed multithreaded CPU-based segmentation GPU-based classification process imagery. AVI videos contain 50 sec between 23,000 - 225,000 particle segments. Processing one through takes average 3.75 mins, depending biological productivity. heavyDB database monitors newly processed is linked interactive visualization. We describe several examples where imaging, AI, enable have transformative effect oceanography. envision AI-enabled high impact ability resolve responses important NCC, such as oxygen minimum zones, or harmful algal bloom thin layers, which affect health ecosystem, fisheries, local communities.

Language: Английский

---

Taming the data deluge: A novel end‐to‐end deep learning system for classifying marine biological and environmental images

Limnology and Oceanography Methods, Journal Year: 2023, Volume and Issue: 22(1), P. 47 - 64

Published: Nov. 10, 2023

Abstract Underwater imaging enables nondestructive plankton sampling at frequencies, durations, and resolutions unattainable by traditional methods. These systems necessitate automated processes to identify organisms efficiently. Early underwater image processing used a standard approach: binarizing images segment targets, then integrating deep learning models for classification. While intuitive, this infrastructure has limitations in handling high concentrations of biotic abiotic particles, rapid changes dominant taxa, highly variable target sizes. To address these challenges, we introduce new framework that starts with scene classifier capture large within‐image variation, such as disparities the layout particles taxa. After classification, scene‐specific Mask regional convolutional neural network (Mask R‐CNN) are trained separate objects into different groups. The procedure allows information be extracted from types, while minimizing potential bias commonly occurring features. Using situ coastal images, compared R‐CNN model encompassing all categories single full model. Results showed approach outperformed achieving 20% accuracy improvement complex noisy images. yielded counts were up 78% lower than those enumerated some small‐sized We further tested on benthic video camera an sonar system good results. integration which groups similar together, can improve detection classification marine biological

Language: Английский

---

Automated zooplankton size measurement using deep learning: Overcoming the limitations of traditional methods

Frontiers in Marine Science, Journal Year: 2024, Volume and Issue: 11

Published: Feb. 13, 2024

Zooplankton size is a crucial indicator in marine ecosystems, reflecting demographic structure, species diversity and trophic status. Traditional methods for measuring zooplankton size, which involve direct sampling microscopic analysis, are laborious time-consuming. In situ imaging systems useful tools; however, the variation angles, orientations, image qualities presented considerable challenges to early machine learning models tasked with sizes.. Our study introduces novel, efficient, precise deep learning-based method measurement. This employs residual network an adaptation: replacing fully connected layer convolutional layer. modification allows generation of accurate predictive heat map determination. We validated this automated approach against manual sizing using ImageJ, employing in-situ images from PlanktonScope. The focus was on three groups: copepods, appendicularians, shrimps. An analysis conducted 200 individuals each groups. method's performance closely aligned process, demonstrating minimal average discrepancy just 1.84%. significant advancement presents rapid reliable tool By enhancing capacity immediate informed ecosystem-based management decisions, our addresses previous opens new avenues research monitoring zooplankton.

Language: Английский

---

Dynamic oceanographic influences on zooplankton communities over the northern Gulf of Mexico continental shelf

Journal of Sea Research, Journal Year: 2024, Volume and Issue: 199, P. 102501 - 102501

Published: April 23, 2024

Dynamic influences of ocean processes on distribution, abundance, and diversity zooplankton communities were studied over the continental shelf in northern Gulf Mexico (GoM) from 2015 to 2017. Zooplankton sampling was conducted four summer cruises northcentral GoM. Sampling designed waters potentially influenced by Loop Current (LC) and/or Mississippi River discharge assess impacts these two mesoscale features abundance zooplankton. During three-year study, LC displayed distinct spatial-temporal variations penetration occurrence Environmental conditions (e.g., sea surface temperature, salinity, dissolved oxygen) varied between months years sampled, significantly different among (ANOVA, p < 0.001). The majority consisted calanoid copepods (65% ± 7.2%, mean SD), while non-copepod taxa primarily chaetognaths, polychaetes, tunicates, ostracods (23 9.2%). Species correlated with oxygen (p 0.05). Canonical correspondence analysis significant associations dominant groups (Monte Carlo Permutation Test, In addition, non-metric multidimensional scaling indicated that assemblages distinct, likely caused plumes during study period. As one few efforts examine dynamics at a low taxon level GoM regarding impact features, this revealed seasonal (i.e. summer) spatial patterns subjected dynamic physicochemical GoM, which will continue changing climate.

Language: Английский

---

Understanding the picture: the promise and challenges of in-situ imagery data in the study of plankton ecology

Journal of Plankton Research, Journal Year: 2024, Volume and Issue: 46(4), P. 365 - 379

Published: June 19, 2024

Abstract Planktons are a fundamental piece of all ocean ecosystems yet, sampling plankton at the high resolution required to understand their dynamics remains challenge. In-situ imaging tools offer an approach sample fine scales. Advances in technology and methodology provide ability make in-situ common tool ecology. Despite massive potential tools, there no standard approaches for analyzing associated data. Consequently, studies inconsistent data, even similar questions. This introduces challenges comparing across devices. In this review, we briefly summarize increasing use, novel applications Then, synthesize analyses used these studies. Finally, address major statistical with unique mechanisms discuss theoretical uncertainties, which arise from low-sampling volumes many tools. To fully unlock power ecological studies, researchers must carefully consider how analyze We recommendations processing data while also acknowledging large need developing new tool.

Language: Английский

---

Rapid Zooplankton Assessment: Evaluating a Tool for Ecosystem‐Based Fisheries Management in the Large Marine Ecosystems of Alaska

Fisheries Oceanography, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 4, 2024

ABSTRACT Ecosystem‐based fisheries management (EBFM) remains an aspirational goal for throughout the world. One of primary limitations EBFM is incorporation basic lower trophic level information, particularly zooplankton, despite importance zooplankton to fish. The generation abundance estimates requires significant time and expertise generate. rapid assessment (RZA) introduced as a tool whereby nontaxonomic experts may produce counts shipboard that can be applied in near real time. Zooplankton are rapidly counted placed into three broad groups relevant higher levels: large copepods (> 2 mm), small (< euphausiids. A Bayesian, hierarchical linear regression modeling approach was used validate relationship between RZA abundances laboratory‐processed ensure method reliable indicator. Additional factors likely impact accuracy predictions were added initial model: sorter, survey, season, marine ecosystem (Bering Sea, Chukchi/Beaufort Gulf Alaska). We tested models included random effect sorter nested within which improved fits both (Bayes R = 0.80) euphausiids 0.84). These also fit when fixed season 0.65). data predict each category results consistent with model training data: 0.80), 0.64), 0.88). Bayesian therefore able associated error accounting these effects. To demonstrate utility management, series from Bering Sea shelf shown vary relation warm cold conditions. This variability impacted commercially important fish, notably Walleye Pollock ( Gadus chalcogrammus ), by managers using risk table approach. provides population estimation process quickly, thus helping fill gap EBFM.

Language: Английский

---

Comparative ecosystem modelling of dynamics and stability of subtropical estuaries under external perturbations in the Gulf of Mexico

ICES Journal of Marine Science, Journal Year: 2023, Volume and Issue: 80(5), P. 1303 - 1318

Published: April 6, 2023

Abstract Human intervention and climate change jointly influence the functions dynamics of marine ecosystems. Studying impacts human on ecosystem is challenging. Unlike experimental studies, research natural systems not amendable at scale time, space, biology. With confounding factors well balanced for two adjacent subtropical estuaries except urbanized disturbances, we conducted modelling using indirect reasoning by exclusion to quantify relative disruption estuarine ecosystems under variability. One major finding this study that tends magnify species fluctuations, complicate interaction network, enhance strength combined with disclosed downscaling effects (indexed as North Atlantic Oscillation Multi-decadal Oscillation) hydrology biological communities. In addition, functional groups appeared respond more diversely external forcing in company interventions. While perturbation was shown destabilize ecosystems, making them vulnerable environmental variability change, buffering diversity trophic tend underpin functions. The findings contribute holistic assessment strategic management subjected disturbances face change.

Language: Английский

---