Periodic Reporting for period 3 - ESiWACE2 (Excellence in Simulation of Weather and Climate in Europe, Phase 2)
Reporting period: 2022-01-01 to 2023-03-31
The central goal of ESiWACE is to enable global storm- and eddy-resolving weather and climate models on next-generation supercomputers. This new type of models allows simulations that explicitly resolve small-scale features, like clouds and storms, and will vastly improve our ability to predict weather patterns in a changing climate. Numerical weather prediction and climate modelling always have been highly dependent on the available computing power and the ability to produce, store and analyse large amounts of simulated data. Increasing computational power is necessary for an increase in the achievable spatial resolution and the completeness and accuracy of physical processes that can be calculated and predicted by the models. Due to the enormous economic importance of weather and climate predictions, these simulations for decades have been routinely run on some of the most powerful supercomputers worldwide. With the transition to exascale computing, operational use of global storm-resolving models, i.e. models based on a very fine underlying mesh, spanning the globe with grid spaces of only a few km, becomes possible, finally allowing to explicitly resolve vertical energy transfers in the atmosphere. This marks a step change in the quality of weather and climate forecasting but also posed challenges that could only be tackled by a coordinated European effort. ESiWACE bundled and strengthened European activities to (1) enable leading European weather and climate models to leverage the performance of pre-exascale systems with regard to both compute and data capacity as soon as possible, and (2) prepare the weather and climate community to be able to make use of exascale systems when they become available.
The ESiWACE project is committed to contributing to the upscaling of leading European weather and climate models to a new class, so-called storm-resolving models, which allow for a new level of quality in assessing and predicting how the weather will change in a changing climate. Targeting Europe’s largest supercomputers, we aim at achieving the maximum possible spatial resolution of the simulations that still yields a throughput of one simulated year per day (SYPD). The threshold of one SYPD marks the line where models become fast enough for useful applications in weather and climate modelling. Since the beginning of the project significant progress could be demonstrated for four model configurations: The pan-european model of the EC-EARTH consortium, the ICON model used by the German Meteorological Service and several other weather and climate centres, the IFS model of the European Centre for Medium-Range Weather Forecasts and a novel French model developed at the Institut Pierre Simon Laplace. Some of these models are now being used in other large climate research projects funded by the EU and national Governments. In particular the Digital Climate twin to be developed in the European Destination Earth initiative is based on the coupled ICON model of the global atmosphere and ocean and on the IFS model. The latter is being coupled to the NEMO ocean model, which also received support via ESiWACE2, and alternatively to the FESOM2 ocean model, which benefited from dedicated ESiWACE2 services. To modernise the underlying software technology of the models and to facilitate their deployment on a range of existing and upcoming computer hardware architectures, ESiWACE2 invested in the development and evaluation of domain-specific programming languages. This emerging technology sacrifices the universality of a programming language for more legible code that can automatically be optimised for different hardware architectures. These could be applied to the widely used European ocean model NEMO, to LFRic, a model deployed in the UK, and also to parts of the ICON and IFS models. To extend our services to models beyond our flagship codes, we reached out to the wider scientific community by offering software support services. Together with our partners Atos and NLeSC, these software support services were offered in annual calls for applications. Over the project duration we supported more than ten different codes. The majority of the most powerful supercomputers cannot easily be employed by weather and climate researchers with their codes because the systems deploy GPUs to reach the required performance. Concrete coding support by high performance computing specialists, as offered by ESiWACE, is therefore extremely valuable in leveraging the potential of modern supercomputing. In addition, ESiWACE2 offered workshops and trainings on the topics relevant to weather and climate scientists for using these upcoming computer systems. Unfortunately, COVID-19 required some events to be held as virtual meetings. On the positive side, this increased the participants’ capacity by avoiding travel costs. Cutting-edge applications of weather and climate models on exascale computers imply the production of extremely large data sets, i.e. exabytes of data. ESiWACE developed Earth System Data Middleware (ESDM) in a collaboration of universities, compute centres and storage vendors to efficiently store huge amounts of data without bothering the user with technical details. This is complemented by extensions for post-processing, analysis and visualisation (PAV) and by interfaces and tools to facilitate the use of modern mainstream data formats on the one hand and of tape archives, a technology being used for decades now but still the most efficient way to store the data volumes we are dealing with, on the other hand.
The European centre of excellence ESiWACE has contributed significantly to a step change in the capability of European climate and weather models by scaling them up to unprecedented spatial resolution and making them exascale-ready. In an international intercomparison of extremely high-resolution atmosphere models initiated and supported by ESiWACE, the breakthrough potential and high socio-economic value of this type of models in general, and the high quality and competitiveness of the European models in particular, has been established. A particular impact and success of ESiWACE is that it was able to gather and heavily foster synergies throughout Europe on the joint way towards exascale. We significantly contributed to the provision of scientifically useful and world leading applications to be used on planned EuroHPC JU supercomputers. This is underlined by the fact that results of our work are now being used in new large scale European and national projects like the European Horizon 2020 project NextGEMS, the German WarmWorld project or the Swiss EXCLAIM initiative. Furthermore our results have a direct impact on the activities of the European Destination Earth initiative, which started to implement a so-called digital twin of the earth based on the models that were made exascale-ready with ESiWACE support.