Title: GIS for the Ocean; Transforming Ocean Data into Actionable Knowledge
Abstract: Introducing GIS for the Ocean, an innovative NOAA/Esri collaboration designed to transform complex ocean data into intuitive insights that support a thriving blue economy. Harnessing the combined power of NOAA data and Esri technology, the GIS for the Ocean (GIS4O) initiative strives to democratize access to actionable information, empowering users—from researchers and policymakers to coastal communities and businesses—to more easily understand ocean patterns and trends. The foundation for the initiative is the Ocean and Coastal Information System (OCIS), a curated suite of ocean and coastal datasets from NOAA and beyond that inform a wide range of critical marine issues. Leveraging Esri’s ArcHub platform, these data are presented through user-friendly visualizations and case studies that range from conservation and sustainable fisheries, to marine planning and future climate scenarios. Promoting a replicable workflow, the project underscores the value of interoperable datasets that can be utilized across various applications to inform policy and planning.
Title: An introduction to Ecological Benthic Units (EBU)
Abstract: Ecological Benthic Units (EBU) are zones within benthic environments with unique biological, physical, and chemical properties. They are key to understanding biodiversity, habitat structure, and ecological functions of benthic ecosystems. The EBU project aims to provide a deeper understanding of the GLOBAL seafloor using the best available data, science, and technology. This presentation briefly outlines the EBU project methods, goals, and highlights the available geospatial resources provided by the project.
Title: Smart Siting for Sustainable Aquaculture: Leveraging Technology and Local Partnerships for Environmental Outcomes
Abstract: Sustainable aquaculture growth hinges on identifying ideal locations that offer social, economic, and environmental benefits. Smart siting, utilizing advanced remote sensing, predictive models, and accurate spatial data, can responsibly expand aquaculture to provide a healthy food supply while minimizing negative impacts on aquatic species and habitats and ensuring economic viability. The Nature Conservancy's (TNC) Global Aquaculture team works with governments, communities, and other key stakeholders to actively address potential threats through science-based spatial planning. This smart siting strategy's success relies on strong local relationships that support sustainable aquaculture development. TNC's approach, first piloted in Palau, has expanded to Mexico, the African Great Lakes, Angola, and French Polynesia, with ongoing efforts to conduct spatial analyses and provide decision support tools. Collaboration with government and key stakeholder partners is key to embedding siting analyses into decision-making processes for a sustainable aquaculture future.
Title: CoastSeg and SDStools: An open-source “ML-powered” software ecosystem for automated coastal shoreline mapping from decades of satellite imagery
Abstract: CoastSeg and SDStools are an ecosystem of Python software, data, and Machine Learning (ML) models for the purposes of mapping time-series of coastal shorelines from satellite imagery. The CoastSeg software supports the workflows governing image acquisition and resampling from public (Landsat, Sentinel) and commercial (PlanetScope) satellite imagery, and all the processing steps required to extract and build databases of shoreline positions from decades of available data. Such information is useful for understanding the dynamics of coasts, identifying chronic erosion and flooding hazards, disaster planning and mitigation, and help study shoreline ecosystems, among many potential uses. The SDSTools software facilitates data QA/QC, other post-processing and mapping applications, as well as data visualization, and shoreline forecasting. Collectively, CoastSeg and SDSTools offer an accessible way to obtain rich time-histories of shoreline position for generic use among coastal scientists, practitioners, and policymakers.
Title: Using species distribution modelling to predict the distribution of biodiversity in the California Current
Abstract: The California Current Ecosystem (CCE) is an eastern boundary upwelling region characterized by high primary productivity that supports rich assemblages of forage and top predator taxa. Species distribution models (SDMs) are highly useful tools to predict the distribution of this biodiversity and project potential future changes that may be relevant to marine spatial planning and ecosystem-based management. Here, we present the results from two studies using SDMs to predict the distribution of (i) five ecologically important seabird species and (ii) an assemblage of 80+ species of epipelagic micronekton. The seabird models are used to elucidate potential changes in the efficacy of marine spatial planning in the CCE under climate change and offshore wind energy development; the forage models are used in a comparative assessment of SDM methodologies. Together, these models quantify the environmental drivers of biodiversity in the CCE across taxa, and can be applied to identify hotspots of ecosystem productivity.
Title: Movegroup: an R package to calculate and visualise space use and home range for groups of animals
Abstract: Calculating home ranges produces 50 & 95% space use contours which reveal habitat preferences and assist MPA selection. The Brownian bridge movement model (BBMM) approach models movements between subsequent relocations, translating animal movement speed into between-point variance. Dynamic BBMMs allow variance (spatial error) to change along a movement trajectory, improving performance for irregular tracks like satellite tag data. dBBMM space use can be calculated for individuals with the move package, but not for groups: firstly it requires a continuous chronological record, but multiple individuals likely overlap in time. Secondly, some individuals’ tags stay on longer, some individuals travel further, invalidating group-level calculations if not addressed with group-level scaling. Movegroup scales multiple individuals to ensure biologically accurate results, accommodates unbalanced receiver arrays, scales and reprojects multiple tracks to the same projection and extent, plots maps, and saves space use and variance calculations. It’s well documented, and on GitHub and CRAN.
Title: Species distribution models for silky shark in the eastern tropical Atlantic Ocean
Abstract: Given the statistical challenges involved in modeling the distribution of wide-ranging and highly mobile species, and the fact that silky sharks are the most frequently caught elasmobranch bycatch species in the tropical Atlantic across multiple fisheries, this study uses the EU tropical tuna purse seine fishery observer data collected from 2010 to 2023, to build a SDM for characterizing the spatiotemporal distribution and environmental preferences of silky shark in the eastern tropical Atlantic Ocean. To address the complexities of species distribution modelling and better capture uncertainty and predictive capacity, we employed three complementary methodologies: Generalized Additive Mixed Models (GAMMs), Boosted Regression Trees (BRTs) and Bayesian Additive Regression Trees (BARTs). Results were consistent across modelling platforms. This type of studies is essential for better understanding fishery-related risks and informing more effective bycatch mitigation strategies across multiple fisheries that catch silky shark incidentally.
Title: Predicting fisheries-bycatch hotspots across the North Pacific
Abstract: The capture of non-target species (bycatch) remains a major barrier to fisheries sustainability and continues to threaten marine megafauna populations. Albatrosses are particularly vulnerable to bycatch in longline and trawl fisheries due to their low reproductive rates, wide-ranging movements and attraction to fishing vessels. Identifying when, where and with which fishing fleets bycatch risk is greatest is crucial for targeted management interventions. We collated extensive biologging datasets (>20 years, >1,500 tracks) from three albatross species and built species distribution models (SDMs) to predict bird densities across the North Pacific. We overlaid these with fishing effort data from Automatic Identification Systems to identify bycatch-risk hotspots. SDMs performed well and captured latitudinal shifts in albatross distributions due to ocean warming events. Our models highlight the regions and time periods with elevated risk for albatrosses, as well as the fishing fleets and management bodies responsible for reducing bycatch.
Title: Satellite-Derived Bathymetry: An Assessment of Empirical Models
Abstract: Satellite-derived bathymetry (SDB) makes possible the extraction of bathymetric data from multispectral satellite imagery. Empirical SDB models have evolved through the years, and one of the goals of these models is to reduce the need for field data for calibration purposes. This presentation will go over some of the limitations and strengths of SDB and its methods. The potential of SDB to act as a bridge for multidisciplinary research between marine science, physics, and GIScience, will also be discussed.
Title: Estimating salt marsh extent in the Dutch Caribbean
Abstract: Monitoring the climate benefits of salt marshes over time requires accurate area estimates, yet the small ecosystem patches common in island territories are often left out of global salt marsh maps, which target larger patches typically found on mainland coasts. Small marsh patches may be difficult to map using remote sensing and land cover classification models, yet their inclusion is critical for these territories to quantify salt marsh climate benefits. We used the Global Wetland Layer map product to quantify salt marsh extent for three islands in the Dutch Caribbean: Aruba, Curaçao, and Bonaire. The salt marsh extent estimates can support the Kingdom of the Netherlands in meeting its Nationally Determined Contributions under the Paris Agreement. Our approach can be replicated in other Small Island Developing States as they pursue inclusion of coastal and marine ecosystems in their climate mitigation plans.
Title: Building Large-Scale Satellite-Derived Shoreline Change Records in Alaska
Abstract: This talk will give an overview of the ongoing efforts to develop historical satellite-derived shoreline change records along the western and northern coasts of Alaska, from the eastern end of the Yukon Delta to the Canadian border. This work utilizes state-of-the-art machine learning models, available through the software CoastSeg, to automatically extract and filter notoriously noisy shoreline data from dense records of satellite imagery. The system developed is being applied to over 4,000 km of Alaskan coastline, processing imagery from Landsat 5, 7, 8, and 9; Sentinel-2; and PlanetScope satellite missions that span from the 1980s to present. Our system is heavily reliant on a new transect and shoreline section organizational scheme, which consists of newly generated index files (50-m spaced transects, reference shorelines, reference polygons, and regions of interest) for each individual shoreline section. Additionally, new techniques for extracting beach slopes and other features of interest from digital elevation data will be presented, as well as some discussion on shore type classification in Alaska. Together, these records will provide richer historical data for scientists studying temporal and spatial patterns of beach and bluff dynamics along the Alaskan coast, as well as further inform the public on coastal hazards affecting Alaska's coastal communities. These remotely sensed historical records are imperative for understanding past as well as projecting future coastal hazards, particularly in remote, relatively data-poor, and rapidly changing Alaskan environments.
Title: Mapping Monterey Bay in Exhibition Design
Abstract: Monterey Bay is a special place with unique geographical features, remarkable biodiversity, and dynamic conditions and is thriving thanks to protections and subsequent recovery. In our upcoming exhibit at The Monterey Bay Aquarium, we will be interpreting the Monterey Bay via a 3D tactile map for visitors to interact with made in collaboration with MBARI. This study highlights the collaboration of design + science in mapmaking and exhibition design.
Title: Modeling the habitat use of the southernmost deep-diver species, the Emperor Penguin in the Ross Sea
Abstract: Emperor penguins undergo an energetically demanding annual molt, requiring them to remain on stable ice where they must fast until their feathers regrow. The foraging strategies following the post-molt period are critical, as individuals must replenish lost energy stores before the next reproductive fast. However, understanding habitat use during this period remains challenging due to the remoteness of the Eastern Ross Sea and the scarcity of in-situ environmental data. To address this, we modeled the habitat use of post-molt emperor penguins using environmental variables from a mix of data sources, either remote sensed or modeled. Our analysis aims to assess habitat selection and investigate if the preferred habitat differs between breeding and non-breeding penguins. This study highlights how a mix of data sources can improve ecological modeling in data-scarce polar regions and help us understand the behavioral adaptations of emperor penguins in response to environmental constraints.