Marine and Environmental Research (MER) Lab’s Post

**Fascinating Research!** New Hampshire researcher, Dr. Elizabeth Clare, is using machine learning to analyze audio recordings of bats' echolocation calls. This innovative approach aims to better understand the complex behavior and social dynamics of these fascinating creatures. By leveraging machine learning algorithms, Dr. Clare and her team can identify patterns and characteristics in the audio recordings that would be difficult or impossible to discern manually. This research has significant implications for conservation efforts, as bats play a crucial role in ecosystems worldwide. By gaining a deeper understanding of their behavior, scientists can develop more effective strategies to protect these vital species. Source: https://lnkd.in/gWK3qUyr Update Date: 21 hours ago

🌊 New Research Alert! 🌊 MGEL's Tina Yack, PhD assisted in the development of a transformative method for fusing visual and acoustic data to estimate North Atlantic Right Whale abundance. By fusing aerial distance sampling (visual observations) with passive acoustic monitoring (hydrophone-recorded whale calls), the research addresses challenges in accurately assessing these critically endangered species. Published in The Annals of Applied Statistics, this study introduces a Bayesian hierarchical framework to combine data sources with differing spatial and temporal resolutions. 🔑 Key Insights: -> Integrated Modeling Approach: Combines aerial observations of surface-active whales and acoustic detections of vocalizing individuals into a unified spatial model. -> Enhanced Spatial Accuracy: Reduces biases from limited visual coverage and non-detectable acoustics, providing a more complete picture of whale distribution and abundance. -> Conservation Applications: Generates models with refined density estimates, allowing adjustments for environmental and behavioural variables critical in policymaking and species conservation. This study provides a replicable framework for blending multi-modal data, with implications for advancing near real-time monitoring and adaptive conservation efforts for marine species. Authored by Erin Schliep, Alan E. Gelfand, Christopher W. Clark, Charles A. Mayo, Brigid McKenna, Susan E. Parks, Tina Yack, PhD, and Robert S. Schick. https://lnkd.in/gZWuUiy8

📢 New paper out today in Trends in Ecology and Evolution: 'The potential for AI to revolutionize conservation: a horizon scan'. 🤖 🌱 For this study I led a horizon scan exercise, bringing together conservation scientists and AI experts to explore how AI may revolutionise conservation 💡 We identified 21 key ideas, ranging from novel interpretation of image and audio data, digital twins for ecosystems, to rapidly assessing scientific literature. A huge thank you to all my collaborators! ➡️ You can read all about the ideas here: https://lnkd.in/e7f3KDzx Full paper: https://lnkd.in/e7f3KDzx ai@cam writeup: https://lnkd.in/e6H-UEaJ

Monitoring for conservation has the potential for some big changes in the not too distant future. There are some interesting ideas and points for further discussion in this paper.

🚀 Exciting News! 🚀 I'm thrilled to share that my first-ever research paper, ‘AquaVision: AI-Powered Marine Species Identification’, has just been published in MDPI! 🎉 🔬 This study addresses the challenge of accurately identifying invasive alien fish species by using machine learning and image classification techniques. The primary aim was to develop an innovative algorithm that can identify the most common invasive Mediterranean fish species (within Maltese coastal waters) based on available images. In particular, these include Fistularia commersonii, Lobotes surinamensis, Pomadasys incisus, Siganus luridus, and Stephanolepis diaspros, which have been adopted as this study’s target species. Our findings highlight the potential of the developed algorithm to support marine biology research, enhance citizen science initiatives, and promote environmental management efforts through accurate fish species identification. I want to extend my gratitude to Dr Adam Gauci, & Prof. ALAN DEIDUN, for their invaluable guidance and contributions throughout this study. For those interested, you can find it in the link below. 👇 📄 https://lnkd.in/dieYhA22 #Research #Science #Publication #MarineBiology #MachineLearning #EnvironmentalManagement #CitizenScience #InvasiveSpecies #Mediterranean #ImageClassification #Oceanography #ArtificialIntelligence

#publication_alert I am pleased to share the addition of my publication in RS to my list of scholarly works. This work "Spatiotemporal dynamics of vegetation cover: integrative machine learning analysis of multispectral imagery and environmental predictors" has been successfully published in the Q2 Journal (Earth Science Informatics). Would love to hear your thoughts and engage in future discussion on this topic. Looking forward to more impactful research together in the future. #EarthSience #Fractionalvegetationcover #XGBoost #RandomForest #Machinelearning #GIS #Landsat #NDVI #vaporpressure #soilmoisture #Publication #Q2 #Research  #GeospatialAnalysis #Remotesensing #Collaboration #AcademicAchievements

EVENT: Join us for a public lecture by Distinguished Professor of Statistics, Kerrie Mengersen (QUT (Queensland University of Technology) on dealing with sensitive data. Attend in-person at The University of Queensland or online at 6pm AEST on Monday 24 June. Many datasets of interest to statisticians are subject to privacy conditions.  This can constrain access, analysis, sharing and release of results. In this presentation, we will consider two ways in which this issue might be addressed. The first is through federated learning, in which the analysis is undertaken in such a way that the data remain in situ and private. The second is synthetic generation of the data, such that the simulated data retains salient characteristics but retains the required privacy. We provide some extensions to the class of models that can be considered in federated learning, and an overview of synthetic generation of tabular data. The exposition of these ideas will be motivated by the creation of an Australian Cancer Atlas. This research is in collaboration with QUT colleagues Conor Hassan and Dr Robert Salomone, and is funded by the Australian Research Council and Cancer Council Queensland. Register here: https://lnkd.in/dUeMtdq #event #mathematicsresearch #STEMevent

Excited to Celebrate a Major Milestone! 📚 I'm thrilled to share the publication of our article, "Improving FluoroProbe Sensor Performance Through Machine Learning," a project that has been a true team effort alongside my brilliant co-authors: Dr. Ofir Tal, Dr. @Assaf Sukenik, and Dr. Tamar Zohary. 🌟 Highlights of our work: 🌊 Addressing challenges in FluoroProbe data caused by diverse phytoplankton community structures in lakes. 🤖 Leveraging machine learning to enhance and extend sensor capabilities effectively. 🔍 Using sensitivity analysis to reveal ecological and physiological insights through high-performing models. Together, we focused on improving how fluorescence sensors assess phytoplankton abundance and community structure, using data from the well-studied Lake Kinneret (Sea of Galilee). Our findings demonstrate how machine learning, particularly Support Vector Regression, can significantly improve sensor performance, enabling accurate identification of key taxa such as dinoflagellates. This accomplishment reflects the power of interdisciplinary collaboration and the shared commitment to advancing environmental monitoring and ecosystem understanding. I am deeply grateful for the opportunity to work with such inspiring colleagues! A link to the paper is in the first comment. 🌱 Here's to the exciting opportunities ahead as we continue exploring and innovating in aquatic science together! #MachineLearning #AquaticEcosystems #FluoroProbe #DataScience #Collaboration #FreshWater #EcoSysem

Here's another great example of an important study--predicting diversity of tree species in forests--using open data from the NEON Program that we proudly operate for the National Science Foundation NCRN, https://okt.to/JTGzLv

EC-ESA #Funding opportunity for #MarineBiodiversity projects MARCO-BOLO , DiverSea , CO-OP4CBD , B-USEFUL @MARBEFES, EcoScope, BioAgora Project , OceanICU project and Biodiversa+Portfolio ... (courtesy of European Space Agency - ESA & European CommissionDG RTD) : #Tender on #Clustering ... promote collaborative research & networking across EC or nationally funded science projects, with the aim of bringing together results, data, knowledge and value-added , to achieve new innovative results beyond the single objectives of the individual projects and beyond the state of the art. 💰 Funding: >500.000 EUR 📅 Deadline: 13 Jan 2025 Expected Outcomes: 🤝 collaboration & #clustering of ESA activities with EC-funded projects and relevant Ocean Health and Biodiversity activities funded by Member States; 🔢 Novel ocean health and biodiversity datasets from the synergistic use of EO, in-situ data and model outcome, for enhanced observation and understanding of open ocean and coastal biodiversity; 🌐 Enhanced scientific understanding of the status and change of the health of ocean ecosystems (marine, coastal) including on-set and drivers of change, vulnerability and responses to natural and anthropogenic pressures, specially under extreme conditions. Specific Objectives: 🗺️ Map major ESA/EC projects #OceanHealth & #Biodiversity, the available datasets, the potential synergies; 🔢 collate #Biodiversity dataset that will serve as a basis for future projects; ❓ identify 3 major scientific challenges; ❓ identify #monitoring and knowledge gaps; ❓ identify 3 data-rich supersites in Polar, open ocean, coastal areas; ✉️ inform policy; 🗺️ Develop a Science Agenda and Roadmap for 2028-2032; 📰 Promote peer-reviewed publications in high profile journals; 🌐 Contribute to the establishment of a coordinated European research area on #OceanHealth & #Biodiversity; 🔓 Foster #OpenScience & #FAIR Principles