Published on 23.10.2024

The Aquatic Life Foundation Project (AqQua): AI-Enabled Global Monitoring of Plankton, a Crucial Carbon Sink in Peril

Decorative image
Image: Klas Ove Möller, HEREON

The Helmholtz Foundation Model Initiative (HFMI) is a lighthouse program of the Helmholtz Association, coordinated by Helmholtz Imaging and Helmholtz AI. HFMI is dedicated to harnessing recent advancements in AI and the development of ‘Foundation Models’.

HFMI funding was now awarded to the Aquatic Life Foundation Project (AqQua), a project designed to transform our understanding of aquatic ecosystems and their role in climate regulation. Spearheaded by a distinguished team of interdisciplinary researchers and with Helmholtz Imaging serving as AI Lead, AqQua leverages cutting-edge AI and high-performance computing (HPC) to analyze and interpret a global collection of plankton image data. Aimed at mapping and monitoring the diverse life forms in our planet’s waters, this project stands out as a prime example of innovative research.

Dagmar Kainmüller, co-spokesperson for AqQua remarks, “We know that plankton plays a major role in sequestering carbon from the atmosphere. We also know that plankton biomass is decaying due to climate change. With that said, it is horrifying to me how little is known about the abundance of most plankton species, and how uncertain estimates of respective carbon uptake are. AqQua is a truly unique opportunity to push for operational global monitoring of plankton biodiversity and carbon export, and I very much look forward to the concerted effort we will pursue with our fantastic interdisciplinary consortium.”

Project Overview

AqQua seeks to create the first foundational pelagic imaging model by consolidating billions of images from diverse aquatic environments around the globe. These images capture a vast range of species—from tiny plankton to larger marine organisms—and hold key insights into biodiversity and ecosystem health. By developing an AI model capable of classifying species, extracting biological traits, and estimating organic carbon content, AqQua will provide critical data necessary for understanding the biological carbon pump — a natural process that sequesters vast amounts of carbon from the atmosphere, significantly affecting global climate.

Strategic Goals and Methodology

AqQua aims to integrate remote sensing and environmental data with direct observations to produce detailed global maps of species distribution and ecosystem status. This comprehensive approach promises to greatly enhance the accuracy and resolution of current models that estimate carbon fluxes in marine and freshwater systems, addressing critical uncertainties that match the scale of anthropogenic CO2 emissions.

Led by speakers Prof. Dr. Rainer Kiko and Prof. Dr. Dagmar Kainmueller, along with contributions from specialists across seven Helmholtz Centers and more than 40 partner institutions around the globe, AqQua plans to transform the scattered and diverse pelagic imaging data into a unified, accessible model. This initiative will not only advance our fundamental understanding of aquatic life, but also support decision-making in relation to emerging ocean-bound carbon dioxide removal technologies.

“It all started at the Helmholtz Imaging Conference in 2023, where Plankton Researchers reached out to us with the aim of getting the most out of the hundreds of millions of images they acquire. The scale to which this has grown is truly amazing. AqQua will be a large Team Science effort driven by dire need. I am thrilled to be on board”, Dagmar Kainmüller adds.

Challenges and Opportunities

Developing such a comprehensive model involves significant challenges, including the harmonization of data across different imaging systems and the creation of self-supervised learning frameworks tailored for unique pelagic image characteristics. Moreover, the project aims to overcome the scarcity of labeled images, which is a major hurdle in training effective AI models.
AqQua is set to make a substantial impact by providing tools that are easy to use and accessible to a broad community of stakeholders, including scientists, policymakers, and industry leaders. Through its open competition in the core computer vision community, it also seeks to foster continuous methodological advancements.

Conclusion

As AqQua moves forward, it holds the promise to significantly enhance our capability to monitor and understand the intricate dynamics of aquatic ecosystems. This project not only highlights the potential of AI in environmental science but also underlines the critical role of interdisciplinary collaboration in tackling some of the most pressing global challenges. The outcomes from AqQua are expected to catalyze further research and action, potentially altering how we manage and conserve our aquatic resources in the face of climate change.


About HFMI

The Helmholtz Association with its 18 centers and large-scale facilities is a world leader in the generation of cutting-edge research data. The challenge of leveraging these vast amounts of data for scientific progress requires synergistic data analysis solutions that can be universally applied across various analysis tasks and data domains. Recent advancements in AI research have given rise to a transformative paradigm designed precisely for this challenge: “Foundation models.” These models are trained on vast and diverse datasets at scale, making them highly adaptable to solve a wide range of tasks. With its unparalleled data repositories across six Research Fields and the robust infrastructure of its Information and Data Science Framework, the Helmholtz Association is uniquely positioned to forge cutting-edge foundation models. Thus, Helmholtz launched the Helmholtz Foundation Model Initiative (HFMI), a systematic approach to developing generalist AI models in a synergistic manner. Initially, HFMI funded four Pilot Projects and an overarching Synergy Unit. Now, a panel of experts has selected three further projects for funding. Resulting open-source models are anticipated to significantly expedite data analysis and are designed to be shared globally, ensuring that our association’s pioneering role extends its benefits and leadership to the broader research community around the world.