LUMI will play a central role in the BioDT project

The LUMI supercomputer will be providing computational and data storage services for the biodiversity digital twins to be developed as a part of the Biodiversity Digital Twin for Advanced Modelling, Simulation and Prediction Capabilities (BioDT) project.

The BioDT project brings together a dynamic team of experts in biodiversity, high-performance computing, artificial intelligence, digital twinning and FAIR data to develop a Biodiversity Digital Twin¹ (BioDT) prototype platform.

This is going to be achieved through a cooperation between 22 European organisations that aim to push the current boundaries of predictive understanding of biodiversity dynamics by providing advanced modelling, simulation and prediction capabilities.

At the core of the Biodiversity Digital Twin prototype platform are use cases which are clustered in four groups. These use cases focus on the species and ecosystems of highest conservation and policy concern, and are vital to human well-being and biodiversity conservation efforts.

BioDT Use Case Groups

Multiple digital twins

Since different use cases and their scopes require different behaviour and different level of precision in comparison with its physical counterpart, there cannot be a single twin answering all possible questions. The BioDT project will thus develop multiple digital twins designed to meet the requirements of the different BioDT use cases.

In BioDT, the digital twins are used to mimic behaviour observed in nature, with the purpose of developing an improved understanding of biodiversity dynamics in response to diverse human pressures, including climate change. By exploiting in new ways existing technologies and data available across relevant research infrastructures, including GBIF, eLTER, DiSSCo and LifeWatch ERIC, the digital twins developed as part of the project will be able to more accurately model interactions between species and their environment.

Harnessing the advanced capabilities of LUMI

As detailed and realistic simulations require large amounts of computing resources, the BioDT project leverages the high-performance computing, artificial intelligence and data analytics capabilities of the LUMI supercomputer for developing and operating the BioDT prototype platform.

Resources have been allocated for the BioDT project via a LUMI development project from the Finnish national quota. The allocated high-performance computing resources via the Finnish quota are 1 000 000 CPU core hours and 100 000 GPU core hours, while allocated storage resource is 90 000 TiB hours. Similar resources have been requested via the EuroHPC JU resource quota for Development Access, with CPU hours and storage already having been granted.

Using the LUMI infrastructure as the basis for building and operating the digital twins will allow to vastly speed up calculations, generate new scientific insights and improve prediction capabilities related to biodiversity dynamics. For example, LUMI’s advanced computing capabilities will allow to widely increase the land area covered by the digital twin models. It will be possible to model in a single simulation area representing a full country instead of localised land area as it is mostly done today.

Platform for a diverse user base

Overall, the BioDT platform is intended to be a clean and lightweight system connecting existing and newly developed shared services for digital twins with the planned users of the platform which include biodiversity research infrastructures and nodes, data providers and researchers, policymakers, industrial actors including small and medium-sized enterprises (SMEs), civil society and citizen scientists.

The BioDT project with the help of the LUMI supercomputer will provide a crucial infrastructure to drive long-term biodiversity research and facilitate science-driven policy and rapid-response actions to enforce current commitments to protecting biodiversity in the long term. To maximise the usability and contribute toward the European Commission’s goal of devising a full digital twin of the Earth, the created twins can then be linked with twins developed as part of other digital twin initiatives.

¹ A digital twin is a virtual representation of real-world entities and processes, synchronized at a specified frequency and fidelity. Here, fidelity refers to the level of precision captured by the digital twin in comparison with its physical counterpart. https://www.digitaltwinconsortium.org/initiatives/the-definition-of-a-digital-twin/

Author: Anna-Liisa Allas, CSC – IT Center for Science, Finland

Follow BioDT:
LinkedIn: https://www.linkedin.com/company/biodt/
Twitter: https://twitter.com/BiodiversityDT
YouTube: https://www.youtube.com/channel/UCqKjtPwy2Tu9lg2sAx1piIA