Facing the Challenges in simulation-based Earth System Sciences and the Role of FAIR Digital Objects

Motivation Results of simulations with climate models form the most important basis for research and statements about possible changes in future global, regional local climate. These output volumes are increasing at an exponential rate (Balaji et al. 2018, Stevens 2019). Efficiently handling these amounts data is a challenge researchers, mainly because development novel workflow approaches have not proceeded same as volume has been increasing. This problem will only become more pronounced ever performance High Performance Computing (HPC) - systems used to perform weather (Lawrence 2018). For example, framework European Commission's Destination Earth program Digital Twins (Bauer 2021) expected produce hundreds terabytes model every day EuroHPC computing sites. The described can be dissected into several aspects, two which we focus on this contribution. Available System Sciences (ESS) increasingly made openly accessible by various institutions, such universities, centres government agencies, addition subject-specific repositories. Further, exploitability simulation beyond expert community humans automated agents (as FAIR principles (F-Findable, A-Accessable, I-Interoperable, R-Reusable), Wilkinson 2016) currently very limited if impossible due disorganized metadata or incomplete provenance information. Additionally, developments regarding globally available workflows spirit Object (FDO) (Schultes Wittenburg 2019, Schwardmann 2020) just beginning. Cultural Change In order address challenges respect mentioned above, current efforts DKRZ (German Climate Center) aimed complete restructuring way performed simulation-based (Anders 2022, Mozaffari Weigel 2020). perfectly suited endeavor, researchers resources services conduct entire suite their data-intensive ranging from planning setting up simulations, analyzing output, reusing existing large-volume datasets publication long-term archival. At moment, DKRZ-users do possibility orchestrate via central service, but rather use plethora different tools piece them together. Framework Environment Freva element new environment shall represented (Free Evaluation Framework) software infrastructure, offers standardized tool solutions ESS optimized high-performance computer (Kadow 2021). designed well FDO framework. crucial aspects here are: standardisation objects input analysis processing, already implemented remote access Persisitent Identifier (PID), still system-internal capture sharing results also groups large communities. It planned extend functionality so that system automatically determines required specific researcher’s question (provided some interface), enquires databases (local disk tape, cloud federated resources) retrieves possible. If (yet), able configure, set submit HPC-System, created becomes (cf. Fig. 1). turn ingested Freva’s catalog reuse. Next, document performed. provided either numerical fields, images animations depending need. As final step, applied and/or underlying published accordance guiding principles. FDOs towards global integrated Data Space To make process sketched out above reality, application concept essential (Schwardmann 2020, Schultes There long tradition dissemination reuse sets. Community standards like those developed internationally coordinated intercomparison studies (CMIP) allow low-barrier Globally resolvable PIDs regular basis. Current close being compatible implementing FDOs, however, now work open points concept, clear definition community-specific requirements including PID Kernel Types specifications, operation type registries technical implementation FDOs. With place following recommendations, queries across spatially distributed types We introduce implementations it highlight face. Using show vision scientist earth science .