Periodic Reporting for period 1 - IS-ENES3 (Infrastructure for the European Network for Earth System modelling - Phase 3)
Reporting period: 2019-01-01 to 2020-06-30
IS-ENES3 is organised around three main objectives with the overall goal of providing the infrastructure to better understand and project climate variability and change through technical excellence:
1. Pursue the integration of the Earth’s climate system modelling community and prepare the sustainability of its infrastructure;
2. Foster the common development of models and tools, and the efficient use of HPC;
3. Support the exploitation of model data by the Earth system science community, the climate change impact community and the climate service community.
IS-ENES3 delivers access services, networking and joint research activities. The key access services are the European part of the Earth System Grid Federation (ESGF), and a central point of entry to information about new data, software, models and tools. Networking aims at growing the user base, increasing the cohesion of the climate modelling community, promoting innovation and preparing for a long-term sustainable infrastructure. Research supports a new community sea ice model, promotes efficient use of high-performance computing, improves the European common model evaluation framework, and develops and enhances data services.
IS-ENES3 plays a key role in supporting the ESGF database of climate simulations, not only by distributing European data, but also by contributions to a global development community (both to the code and to the leadership). During the first period, a roadmap for the future architecture of ESGF has been developed and aspects of the necessary transition begun. IS-ENES3 also contributes to the development and use of international standards such as the metadata common information model and the climate forecast conventions used for climate data.
With over 15 000 users worldwide, ESGF plays a key role in supporting climate science and IPCC Assessment Reports. At of the end of the first period, 8 Petabytes (PB) of climate model data have been downloaded from European ESGF data notes and 3PB accessed by European users. 5PB of CMIP6 data were also replicated and made accessible at large European replica pools, supporting European climate modelling groups. New services were also established to make CMIP6 data compliant with FAIR data principles, including both a citation service providing data citation and license information for over 2000 CMIP6 data collections and the issuing of persistent identifiers for all CMIP6 files.
The evaluation of climate models is important to assess confidence in model results. IS-ENES3 has started to support a common European tool for model evaluation, and is working on coupling it with ESGF data to provide metrics on CMIP6 model data to be used for the preparation of the IPCC AR6 report.
IS-ENES3 also supports community approaches in climate modelling, such as the newly developed European platform for sea ice modelling, gathering expertise from previous three different models, and coupled to the European platform for ocean modelling. A high-profile international workshop emphasized the different needs for sea ice development and the importance of engaging a wider community.
Now in its third phase, IS-ENES3 also prepares for sustainability. Climate modelling, together with the necessary models, tools and data distribution are long term activities, extending well beyond the project horizon. A technical scoping for sustainability has been elaborated together with the IS-ENES3 service providers and a first legal screening has been finalised. The project is in the middle of the scientific scoping stage.
The ESGF future architecture will lead to improvements in both the user experience in accessing data and the host experience in deploying and maintaining the software.
Reducing redundancy and duplication of activities within the European climate modelling community via sharing of software and expertise is an important IS-ENES3 objective. This target is hard to measure (sharing expertise) and slow to deliver (timescales for making and validating changes to models and tools can be long) - but the paybacks in computational and human efficiency will be large.
The use of model data is increasing with the development of climate services, and in Europe the Copernicus Climate Change service is relying on the expertise of IS-ENES3. IS-ENES3 expects to further increase the user base, in particular in Eastern and Central Europe. One route to this increase is via training sessions which have begun during the first period.
At the end of the project, it is expected to have a strong and sustained IS-ENES infrastructure, supported by an updated ten-year strategy.