Search

The central objectives of Work Package 1 of ESiWACE2 is to develop and run coupled model simulations at a resolution as high as possible and with a throughput rate of at least one simulated year per day (1 SYPD) producing full model output. This throughput would be sufficient to use these simulations for operational weather predictions and potentially also to run the models for a couple of decades, to address scientific questions that are related to climate change. Four model configurations are considered: EC-Earth with OpenIFS coupled to NEMO, the Integrated Forecasting System (IFS) coupled to NEMO, the ICON model with both atmosphere and ocean, the DYNAMICO model coupled to NEMO, covering both weather and climate configurations as well as well-established models and models that are currently under development. Furthermore, ocean-only simulations with the NEMO models are also pushed to very high spatial resolution (1/36 ◦ ; ∼3 km). According to the initial plan, the main aim of this deliverable was to perform scalability tests on the pre-exascale EuroHPC systems. However, this was impossible as these machines are only just becoming available. The deliverable is therefore focusing on the ongoing work to make the models more efficient and to enable the porting of the codes to heterogeneous hardware. As the European pre-exascale machines were not available, some tests of the models on other European HPC systems and some of the world’s fastest supercomputers (Summit and Fugaku) are also documented. Furthermore, the work on the treatment of output from ensemble simulations using XIOS that was moved from Deliverable D1.1 is also documented here, showing the successful use of XIOS to generate ensemble model output straight from ensemble simulations., ESiWACE2 stands for Centre of Excellence in Simulation of Weather and Climate in Europe Phase 2. ESiWACE2 is funded by the European Union's Horizon 2020 research and innovation programme (H2020-INFRAEDI-2018-1 call) under grant agreement 823988.

The computational performance of supercomputers is usually expressed in terms of their theoretical peak performance, or measured with rather simple and easy to handle benchmark codes, like the High-Performance Linpack Benchmark, a measure of the floating-point computing power. These metrics allow to easily rank the performance of computers but have only limited value when it comes to the question of how well a computer system is suited for specific complex applications like weather and climate models. The EU-funded project ESCAPE-2, which ran from 2019 to 2021, therefore developed a benchmark suite called High Performance Weather and Climate Benchmark, in short HPCW, which consists of widely used European weather and climate models and standalone model components (also called dwarfs in the ESCAPE-2 context) that are less complex than the models themselves but represent selected characteristics of the models. After the end of ESCAPE-2, the centre of excellence ESiWACE2 took over the responsibility to establish HPCW as a useful tool for the communities of weather and climate modellers on the one hand, and for computer manufacturers on the other hand. At the end of ESiWACE2 we provide a consolidated and extended version of HPCW that has been installed and tested on some of the largest high performance computing systems in Europe. The work will be continued in the third phase of the centre of excellence, ESiWACE3., ESiWACE2 stands for Centre of Excellence in Simulation of Weather and Climate in Europe Phase 2. ESiWACE2 is funded by the European Union's Horizon 2020 research and innovation programme (H2020-INFRAEDI-2018-1 call) under grant agreement 823988.

This deliverable summarises the beneficiaries’ strategy and concrete actions related to the protection and the exploitation of the project results after the end of the project., ESiWACE2 stands for Centre of Excellence in Simulation of Weather and Climate in Europe Phase 2. ESiWACE2 is funded by the European Union's Horizon 2020 research and innovation programme (H2020-INFRAEDI-2018-1 call) under grant agreement 823988.

The models used in weather and climate simulation are large, complex and long-lived. It is then very difficult to adapt them to run well on supercomputers which are quickly evolving and tend to have a lifespan of only five years or so. Domain-Specific Languages (DSLs) are one solution to this problem: they enable the model developers to focus on the scientific challenges while the underlying tool chain takes responsibility for achieving good performance on a given supercomputer. This report describes the computational performance obtained through the application of the PSyclone and dusk/dawn DSL toolchains to full atmosphere and ocean models as used for weather and climate studies at various centres. These are: the UK Met Office’s LFRic atmosphere model, the NEMO ocean model, the IFS-FVM atmosphere model and the ICON atmosphere model., ESiWACE2 stands for Centre of Excellence in Simulation of Weather and Climate in Europe Phase 2. ESiWACE2 is funded by the European Union's Horizon 2020 research and innovation programme (H2020-INFRAEDI-2018-1 call) under grant agreement 823988.

Managing data from modern high resolution weather and climate simulations is becoming problematic. Here we present a summary of some of the middleware software developed within ESiWACE2 to help manage the deluge of data during and after simulations. Middleware is software that lies between the computer, network, storage fabric and the scientific code that exploits that fabric. The objective of the middleware developed here is to exploit some of the scientific knowledge we have about our data to more efficiently use the underlying fabric. To that end, we present a description of two classes of software which we have developed and/or improved during ESiWACE2: (1) Tools for in-flight analysis of ensembles of simulation which ideally can support cross-ensemble calculations, avoiding the need for writing all variables all the time or at analysis frequency for all ensemble members to disk; and (2) Middleware for managing writing data to disk and managing the migration of data through storage tiers. The particular products we discuss are XIOS and it’s ensemble analysis tools, the Earth System Data Middleware (ESDM), the Near Line Data System (NLDS) and tools which exploit the Climate-Forecast Aggregation conventions (CFA-C, CFA-Python, CFA-CPP and CFStore). Not all of these pieces of software are feature complete, but for each we discuss their current status, where to find code and documentation, and their roadmap to the future., ESiWACE2 stands for Centre of Excellence in Simulation of Weather and Climate in Europe Phase 2. ESiWACE2 is funded by the European Union's Horizon 2020 research and innovation programme (H2020-INFRAEDI-2018-1 call) under grant agreement 823988.

This presentation describes the progress in the development of two domain specific languages for weather and climate applications, PSyclone and gt4py (formerly dusk/dawn). Several models ofESiWACE-2 have been ported and adapted to use the DSLs: NEMO, ICON, LFric and IFS.&nbsp

This deliverable reports on the two storage technologies from two vendors, DDN and Seagate, which were designed to work with the Earth System Data Middleware (ESDM) - details of which were reported in Deliverable D4.1 “Advanced software stack for Earth system data”[ESiWACE2 D4.1]. The expectation was that various storage appliances would have support for ESDM and could be deployed with ESDM at the weather and climate sites as ESDM would become available. Seagate demonstrated the CORTX Motr (formerly Mero) software installation being capable of working with commodity hardware, with the ability to be installed at the German Climate Computing Center (DKRZ) and other centres in the future as per their requirements. This is a change in strategy from Seagate for CORTX Motr. At the beginning of the project, the plan was to have an appliance based on Mero – as Mero was closed source software. However, during the time frame of ESiWACE2, Mero was open sourced as CORTX Motr – making it feasible to be deployed very widely on commodity hardware. Showcasing CORTX Motr’s capability to work on commodity hardware working with ESDM, we believe is a powerful message for the community rather than providing just a “closed” appliance. This deliverable hence takes that approach. DDN demonstrated the relevance and the efficiency of the Earth Systems Data Middleware (ESDM) developed during the ESiWACE project on state-of-the-art industrial storage technologies. DDN implemented a sophisticated integration of ESDM with the flash native, burst buffer solution product of DDN’s portfolio, Infinite Memory Engine (IME). As part of the integration ESDM now supports the IME native interface that bypasses the FUSE, allowing to obtain the maximum performance of IME. The code of the integration is available and accessible to everyone through GitHub. Using IME, DDN was able to produce performance results significantly beyond the state of the art in term of hardware / software cooperation. Furthermore, ESDM provides performance in the range of the low-level API while still offering a higher level of semantic to the end-user. The results were obtained within one of DDN’s labs using 5 IME appliances and 10 clients. With an aggregated peak I/O bandwidth of 100 GB/sec, this setting is representative of the core of the HPC / technical computing market. In addition to the lab set-up and the integration effort, DDN has led more theoretical and research-oriented work. In respect to the Active Storage task, DDN provided a complete design architecture and implemented a proof of concept (PoC). The prototype is expected to be ready within the Q1 of 2023. The PoC code of the active storage is available on GitHub.

Presentation given by Joachim Biercamp at iCAS 2022 in Stresa (IT), 11-15 September, 2022

Recent studies have established that the typical atmospheric and oceanic resolutions used for the CMIP5 coordinated exercise, i.e., around 40km-150km globally, are limiting factors to correctly reproduce the climate mean state and variability. In the framework of the ESiWACE project, the Barcelona Supercomputing Center (BSC) developed a coupled version of the EC-Earth 3 climate model at a groundbreaking horizontal resolution of about 15km in each climate system component. In the atmosphere, the horizontal domain was based on a spectral truncation of the atmospheric model (IFS) at T1279 (15 km) together with 91 vertical levels. The ocean component (NEMO) ran on the ORCA12 tripolar (cartesian) grid at a horizontal resolution of about 1/12° (16 km), with 75 vertical levels. This very-high-resolution (VHR) configuration was used in the Glob15km project to run a 50-year spinup from which one historical and one control simulation of 50 years each were started, following the HighResMIP protocol from CMIP6. These experiments are currently being used to identify the improvements in process representation with respect to coarser resolution and to pin down physical and dynamical reasons behind these differences induced by resolution change. The VHR coupled configuration was a great benchmark to reveal the most critical scalability problems of the EC-Earth 3 model. Within the ESiWACE2 project, those issues have been tackled to allow operational climate predictions at more than 1 SYPD (Simulated Years Per Day) with production-mode configurations. The new Tco639-ORCA12 configuration is based on the community EC-Earth 4 model, made up of OpenIFS cycle 43r3 and NEMO 4, and uses a cubic octahedral grid in the atmosphere. In this version of EC-Earth, both the atmospheric and the oceanic component output diagnostics through the asynchronous XIOS servers, contributing to reducing the I/O overhead and improving scalability, which will be evaluated at the end on one of the forthcoming pre-exascale EuroHPC systems. This new configuration has already been scaled in MareNostrum 4 with a peak performance of 2 SYPD. It will be subject to the last phase of development and optimization until it is finally deployed in one of the forthcoming Pre-Exascale platforms to facilitate the execution of experiments with an unprecedented time-to-solution and allowing additional output capabilities to help understand the climate mean state, variability, and extremes.

Recent studies have established that the typical atmospheric and oceanic resolutions used for the CMIP5 coordinated exercise (Coupled Model Intercomparison Project, phase 5), i.e., around 40km-150km globally, are limiting factors to correctly reproduce the climate mean state and variability. In the context of ocean forecasting like the Copernicus Marine Environment Monitoring Service (CMEMS, https://marine.copernicus.eu/), increasing resolution appears necessary to improve the quality of service and to satisfy the users’ needs in the operational application. Resolving scales below 100 kilometers, particularly submesoscale processes (1-50 km), is necessary to better represent the circulation in the open ocean and improve the large-scale representations. Therefore, Ocean General Circulation Models (OGCM) have to evolve to resolve the global ocean flows at kilometric scale by exploiting massively parallel architectures and reducing I/O bottlenecks. In the framework of the ESiWACE project, unprecedented configurations with resolutions in the range of 1km-10km were developed for some of the more supported European climate models, like NEMO or EC-Earth. In the EC-Earth case, the Barcelona Supercomputing Center (BSC) developed a coupled version of the EC-Earth 3 climate model at a groundbreaking horizontal resolution of about 15km in each climate system component. This very-high-resolution (VHR) configuration was used in the Glob15km project to run multi-decadal experiments that are currently being used in the STREAM project to identify the improvements in process representation with respect to coarser resolution model versions. Within the ESiWACE2 project, new versions of the so-called demonstrators are being developed to allow operational climate predictions at more than one simulated year per day (SYPD) with production-mode configurations. In the case of EC-Earth, the most critical scalability problems have been tackled in the new EC-Earth 4, based on OpenIFS cycle 43r3 and NEMO 4, and the new Tco639-ORCA12 configuration uses a cubic octahedral grid in the atmosphere. In EC-Earth 4, both the atmospheric and the oceanic component output diagnostics through asynchronous XIOS servers, contributing to reduce the I/O overhead and improving scalability, which will be evaluated at one of the forthcoming pre-exascale EuroHPC systems. In the NEMO context, a new global configuration (called ORCA36) has been developed with a 2-3 km resolution (1/36°). It is run with NEMO 4 OGCM and asynchronous XIOS servers. ORCA36 is forced by the high spatial resolution ECMWF IFS dataset at one-hour time frequency, which will improve ocean prediction, but will affect the time-to-solution. The use of a finer spatial resolution also implies that a smaller time step has to be used. Running such a global high-resolution configuration will increase the computational needs, the size of model inputs and outputs and requires writing model outputs at a higher frequency. In order to reach a good time-to-solution, in the framework of ESIWACE 2 project and also the IMMERSE H2020 project, new HPC features have been developed in the NEMO OGCM to improve its scalability and will be tested on ORCA36 configuration.

We have documented the current progress during the implementation of the production mode simulations on supercomputers of the consortium in deliverable D1.1 (https://zenodo.org/record/5795991) which is submitted together with this Milestone document. The model development teams are making good progress when preparing the model configurations for the pre-exacscale EuroHPC systems. We now hope to be able to get access to those machines soon, to make the final scalability tests in 2022.

The milestone target was to report on the delivery of software to deliver tight integration between the scheduler (CYLC) and the I/O middleware (ESDM) so as to support maximally efficient use of storage resources during pre-exascale workflows. However, as the work was being done, analysis showed that the benefits available at this time of carrying out the planned tight integration would not warrant completing all the software development; we believe the existing level of integration with the scheduler running code which includes storage configuration delivers on the underlying requirements. Accordingly, we mark the milestone complete (the underlying requirement has been met), and have redeployed resource yielded by not completing the tight integration software on other aspects of WP4.

Recent studies have established that the typical atmospheric and oceanic resolutions used for the CMIP5 coordinated exercise, i.e., around 40km-150km globally, are limiting factors to correctly reproduce the climate mean state and variability. In the framework of the ESiWACE project, the Barcelona Supercomputing Center (BSC) developed a coupled version of the EC-Earth 3 climate model at a groundbreaking horizontal resolution of about 15km in each climate system component. In the atmosphere, the horizontal domain was based on a spectral truncation of the atmospheric model (IFS) at T1279 (15 km) together with 91 vertical levels. The ocean component (NEMO) ran on the ORCA12 tripolar (cartesian) grid at a horizontal resolution of about 1/12° (16 km), with 75 vertical levels. This very-high-resolution (VHR) configuration was used in the Glob15km project to run a 50-year spinup from which one historical and one control simulation of 50 years each were started, following the HighResMIP protocol from CMIP6. These experiments are currently being used to identify the improvements in process representation with respect to coarser resolution and to pin down physical and dynamical reasons behind these differences induced by resolution change. The VHR coupled configuration was a great benchmark to reveal the most critical scalability problems of the EC-Earth 3 model. Within the ESiWACE2 project, those issues have been tackled to allow operational climate predictions at more than 1 SYPD with production-mode configurations. The new Tco639-ORCA12 configuration is based on the EC-Earth 4 model, made up of OpenIFS cycle 43r3 and NEMO 4, and uses a cubic octahedral grid in the atmosphere. In this version of EC-Earth, both the atmospheric and the oceanic component output diagnostics through the asynchronous XIOS servers, contributing to reduce the I/O overhead and improving scalability, which will be evaluated at one of the forthcoming pre-exascale EuroHPC systems. Recent studies have established that the typical atmospheric and oceanic resolutions used for the CMIP5 coordinated exercise, i.e., around 40km-150km globally, are limiting factors to correctly reproduce the climate mean state and variability. In the framework of the ESiWACE project, the Barcelona Supercomputing Center (BSC) developed a coupled version of the EC-Earth 3 climate model at a groundbreaking horizontal resolution of about 15km in each climate system component. In the atmosphere, the horizontal domain was based on a spectral truncation of the atmospheric model (IFS) at T1279 (15 km) together with 91 vertical levels. The ocean component (NEMO) ran on the ORCA12 tripolar (cartesian) grid at a horizontal resolution of about 1/12° (16 km), with 75 vertical levels. This very-high-resolution (VHR) configuration was used in the Glob15km project to run a 50-year spinup from which one historical and one control simulation of 50 years each were started, following the HighResMIP protocol from CMIP6. These experiments are currently being used to identify the improvements in process representation with respect to coarser resolution and to pin down physical and dynamical reasons behind these differences induced by resolution change. The VHR coupled configuration was a great benchmark to reveal the most critical scalability problems of the EC-Earth 3 model. Within the ESiWACE2 project, those issues have been tackled to allow operational climate predictions at more than 1 SYPD with production-mode configurations. The new Tco639-ORCA12 configuration is based on the EC-Earth 4 model, made up of OpenIFS cycle 43r3 and NEMO 4, and uses a cubic octahedral grid in the atmosphere. In this version of EC-Earth, both the atmospheric and the oceanic component output diagnostics through the asynchronous XIOS servers, contributing to reduce the I/O overhead and improving scalability, which will be evaluated at one of the forthcoming pre-exascale EuroHPC systems.

Presentation on ESiWACE given by Joachim Biercamp on 7 July 2021 at the minisymposium "Towards the New Generation of Earth System Models: Challenges on the Road to Exascale" at the virtual PASC21 conference.

Presentation by Daniel Klocke and the DYAMOND team given on 6 July 2021 at the minisymposium "Towards Kilometer-Scale Global Storm-Resolving Weather and Climate Simulations, Part II" at the virtual PASC21 conference.

Presentation given in the session "Scientific Visualization & Data Analytics Showcase" at the virtual Supercomputing Conference 2020 (SC20).

This milestone has the form of a report that provides design considerations and presents design choices. It includes an architectural view of the data management components in ESiWACE2 WP4 (Task 1 to Task 5) and expands upon some designs in WP5. The purpose for the documentation of the design and interaction of the components is to provide a common ground in the ESiWACE2 consortium and to support the development of components that are utilized in different use cases.

ESiWACE2 will push resolution towards unprecedented levels for coupled simulations in production mode that show a performance that would allow operational weather and climate predictions (at least 1 SYPD) with realistic model output with a resolution that is as high as possible. Four coupled models have been chosen for this exercise. The model configurations that are developed within WP1 of ESiWACE2 will be ported to the European pre-exascale EuroHPC systems planned for 2021. The scaling behaviour of the models will be diagnosed and documented. To achieve this goal towards Deliverable D1.1 and D1.3, this Milestone is specifying the model configurations that will be used for the four models that will be run in production mode. Three of the models are aiming for the model output requirements of the 2nd phase of the DYAMOND intercomparison to define a realistic model output for production simulations. https://www.esiwace.eu/services/dyamond/dyamond-2nd-phase-the-winter-experimental-protocol/view

Presentation on the links between the ESiWACE COE and the ESCAPE-2 FET project given at the 1st dissemination workshop of ESCAPE-2 in Reading.

Kilometer-scale ensemble simulations are expected to significantly boost and impact weather and climate predictions in the future. However, these simulations will only be enabled by exascale compute power and corresponding data capacity. In the following, we discuss a European effort in terms of the e-infrastructure Centre of Excellence in Simulation of Weather and Climate in Europe (ESiWACE). ESiWACE provides infrastructural means to prepare the weather and climate communities for simulations at the exascale. We give an overview of several ESiWACE infrastructure components and discuss their role in reaching the goal of kilometer-scale ensemble predictions. We particularly review the outcomes of the ESiWACE demonstrators, that is community-driven kilometer-scale models that have been developed throughout the last years.

This deliverable provides the final implementation of the Earth System Data Middleware (ESDM) within the ESiWACE project. The developed middleware aims at deployment in both simulation and analysis workflows where the volume and rate of data lead to performance and data management issues with traditional approaches. A prototype has been coded, reviewed and agreed between the project partners. The source code for the middleware is tagged in Git as ESDM‐V0.5, and it is available under LGPL license in a repository on Github: https://github.com/ESiWACE/esdm. The Earth System Data Middleware (ESDM) has been designed to deal with the fact that existing data libraries for standardised data description and optimised I/O, such as NetCDF, HDF5 and GRIB, do not have suitable performance portable optimisations for large‐scale simulation workflows in climate and weather. There are three broad data‐related challenges that weather and climate workflows need to deal with, which can be summarised as requiring to handle: the velocity of high volume data being produced in simulations; the economic and performant persistence of high volume data; and high volume data analysis workflows with satisfactory time‐to‐solution. ESDM provides a middleware architecture that delivers economical performance portability across different environments addressing these three challenges at once, which usually are independently considered by all major centres. The final product also aids the interests of stakeholders: developers have less burden to provide system‐specific optimisations and can access their data in various ways. Data centres can then utilise the storage of different characteristics. The prototype was documented and tested with I/O benchmarks on large scale systems to prove the feasibility of the design., ESiWACE has received funding from the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No 675191

The new ESiWACE2 project presented in short. Its objectives, codes of interest, services, training and the interaction with the industry.

Summary This deliverable reports about the HPC (High-Performance Computing) workshop, which was held in Toulouse on April 6 and 7 in 2016. The report explains how this workshop relates, and builds upon, preceding workshops organized under the infrastructure project called IS-ENES2 (ENES stands for "European Network for Earth-System modelling"). The agenda of the workshop, as well as the list of attendees, are parts of this deliverable. Indications are also given about the way it is presently envisaged to exploit and further disseminate the results of the workshop. &nbsp

31 August 2018: ATTENTION, this deliverable has been updated with Deliverable D3.6 Zenodo link: https://zenodo.org/record/1453864#.W7248faYTwk Summary Nowadays, to build, run and configure up-to-date weather and climate models it takes more components than solely the source codes. Models are subject to a long list of dependencies in the form of libraries, open source and/or commercial software packages or many other pieces of software such as system utilities. These requirements make the life of system administrators in HPC centres hard. When a new user gets resources on a specific platform, it will take time before the actual simulations can start, because platform related problems with models and software need to be solved first. The aim of this deliverable is to explain the methodology to select, configure and install a complete ESM software stack in a handbook. This handbook is intended to ease the deployment of models and to reduce the time to start simulations at an HPC facility. The time to solution from idea to published result should thus be substantially reduced. Ideally, following the guidelines in this document, the centres willing to run weather and climate models will be prepared to offer all the tools required by these models, and their users, the scientists. As a matter of precision, in this document we do not define the software stack needed for a centre to run weather and climate models. This definition comes from the deliverable D3.1 of ESiWACE.

Summary This document describes experiences from three summer schools, especially from the most recent one in 2016, on modelling the Earth System with a multi-model ensemble. These models have been run by the student participants at different HPC sites solving scientific assignments. The experiences of the technical teams preparing the models for these experiments shed light on the usability of the models from different perspectives. One of the aspects of particular importance for ESiWACE is performance portability, which will be a focus for the WP in the next reporting period.

About this document Work package in charge: WP3 Usability Actual delivery date for this deliverable: 28 February 2017 Dissemination level: PU (for public use) Lead authors: Met Office (MetO): Mick Carter, Dave Matthews, Matthew Shin Other contributing authors: Max-Planck-Institut für Meteorologie (MPI-M), Reinhard Budich (Reviewer) Summary This document describes the development work done to the Cylc meta-scheduler as well as the engagement and support activities done within the project up to the date of the report. This is put in the context of the wider effort of Cylc development. The bulk of the effort in this period has been development and the priorities of that development were agreed at a workshop on workflows organised by the IS-ENES2 project.

About this document Work package in charge: WP3 Usability Actual delivery date for this deliverable: February 2017 Dissemination level: PU (for public use) Lead author: Barcelona Supercomputing Center (BSC), Kim Serradell Other contributing authors: Max-Planck-Institut für Meteorologie (MPI-M), Reinhard Budich (Reviewer), Sergey Kosukhin The University of Reading (UREAD), Grenville Lister Summary Based on the work done in the scope of T3.1.3 in WP3 “Usability” and reported in D3.5 “How to Select, Configure and Install ESM Software Stacks”, in this document we will narrow the software stack needed to deploy and run the demonstrators defined in Work Package 2 “Scalability”. The goal is to build an exhaustive list of all the components needed by these ESMs to make the installation and execution as easy as possible.

Version 1, Specification of a Standard Recommendation for an ESM End-­‐to-­‐end Workflows and Application Software Environments Framework About this document Work package in charge: WP3 Usability Actual delivery date for this deliverable: March 2016 Dissemination level: PU Lead author: Reinhard Budich (MPI-­‐M) Other contributing authors: Max-Planck-­Institut für Meteorologie (MPI-­‐M), Sergey Kosukhin Barcelona Supercomputing Center (BSC), Oriol Mula-­Valls, Kim Serradell The University of Reading (UREAD), Grenville Lister Summary Collected, analyzed and systematized hardware and software requirements for Modelling the Earth System. Initiated the specification of a standard recommendation for an ESM End-­‐to-­‐end Workflows and Application Software Environments framework The last version of the White Paper is saved here: https://www.esiwace.eu/results/misc/the-application-software-framework/view

About this document Work package in charge: WP3 Usability Actual delivery date for this deliverable: March 2016 Dissemination level: PU Lead author: Reinhard Budich (MPI-­‐M) Other contributing authors: Max-­‐Planck-­‐Institut für Meteorologie (MPI-­M), Sergey Kosukhin Barcelona Supercomputing Center (BSC), Oriol Mula-­‐Valls, Kim Serradell The University of Reading (UREAD), Grenville Lister Summary Narrowed down the “Application Software Framework: A White Paper” to the software specification for the 3rd Third European Earth System and Climate Modelling School (3rd E2SCMS) - Helsinki, 9-21 June 2016. Initiated by the European Network for Earth System modelling (ENES) with support of the FP7 project IS-ENES2 (http://is.enes.org), the E2SCMS Earth System and Climate Modeling Summer School is open to early career scientists (advanced PhD candidates, postdoctoral scientists, scientific programmers) who are affiliated with European research institutions. The course is supported by international experts in Earth system and climate modeling ensuring an advanced and stimulating learning environment. The course begins with a series of lectures, which are followed by practical sessions where participants apply what they have learnt by analyzing the results of their own Earth system simulations. In this part of the course three well established Earth System Models will be used: HAD-GEM (FAMOUS), MPI-ESM, and EC-Earth, which will be run at international supercomputing centers remotely from the course.

About this document Work package in charge: WP5 Management & Dissemination Actual delivery for this deliverable: 31 October 2015 Dissemination level: PU (for public use) Lead author: German Climate Computing Center (DKRZ), Project office, Kerstin Fieg Other contributing partners: European Centre for Medium‐Range Weather Forecasts (ECMWF), Hilda Carr CNRS‐ Institut Pierre‐Simon Laplace (CNRS‐IPSL) Francesca Guglielmo Barcelona Supercomputing Center (BSC), Isadora Jimenez Max Planck Institute for Meteorology (MPI‐M), , Chiara Bearzotti Summary The Media and Communication Plan (MCP) describes the communication structures implemented to steer the communication between the project and all potential stakeholders and multiple audiences providing targeted information for each identified audience. The background for this document is to be found in the Art. 38.1 of the Grant Agreement Communication activities by beneficiaries: „The beneficiaries must promote the action and its results, by providing targeted information to multiple audiences (including the media and the public) in a strategic and effective manner“. The Media and Communication plan represents the project strategy for: Communicating the results of our research to a number of audiences we have identified in the document and with an adequate number and choice of communication tools Promoting the project and its results beyond the project‘s own community Promoting interaction with selected audiences The Plan also indicates 1) which audiences we target and 2) which communication measures are adopted to open a dialogue with these audiences. A special section of the MCP is also dedicated to communication tools adopted within the consortium to ensure internal information flow and exchange and thus foster a smooth and integrated implementation of the activities described in the Description of the Action.

This report provides an architectural view of all the data management components (“data theme”) set up in the ESiWACE2 project. It provides insights into the design of the ESDM interface for supporting data-intensive post-processing, analytics and visualization (PAV) and case studies from the weather and climate change domain. The document addresses the milestone MS5.1 of the ESiWACE2 project. More in detail, starting from background work and the description of the main use cases targeted by the ESDM, a set of technical requirements for the ESDM unified API has been identified, with a strong focus on the new PAV layer for in-flight analytics and visualization. The general ESDM architecture is introduced, highlighting the main components for I/O and processing. The proposed architecture tries to address the identified requirements and provides the guidelines to steer the development stage through the definition of the main architectural components, the relationship among these and some preliminary technical details.

This deliverable provides the final implementation of the Earth System Data Middleware (ESDM) within the ESiWACE project. The developed middleware aims at deployment in both simulation and analysis workflows where the volume and rate of data lead to performance and data management issues with traditional approaches. A prototype has been coded, reviewed and agreed between the project partners. The source code for the middleware is tagged in Git as ESDM‐V0.5, and it is available under LGPL license in a repository on Github: https://github.com/ESiWACE/esdm. The Earth System Data Middleware (ESDM) has been designed to deal with the fact that existing data libraries for standardised data description and optimised I/O, such as NetCDF, HDF5 and GRIB, do not have suitable performance portable optimisations for large‐scale simulation workflows in climate and weather. There are three broad data‐related challenges that weather and climate workflows need to deal with, which can be summarised as requiring to handle: the velocity of high volume data being produced in simulations; the economic and performant persistence of high volume data; and high volume data analysis workflows with satisfactory time‐to‐solution. ESDM provides a middleware architecture that delivers economical performance portability across different environments addressing these three challenges at once, which usually are independently considered by all major centres. The final product also aids the interests of stakeholders: developers have less burden to provide system‐specific optimisations and can access their data in various ways. Data centres can then utilise the storage of different characteristics. The prototype was documented and tested with I/O benchmarks on large scale systems to prove the feasibility of the design.

The central objective of work package 1 of ESiWACE2 is to develop and run coupled model simulations at a resolution as high as possible and with a throughput rate of at least one simulated year per day (1 SYPD) producing full model output. This throughput would be sufficient to use these simulations for operational weather predictions and potentially also to run the models for a couple of decades, to address scientific questions that are related to climate change. Four model configurations are considered (EC-Earth with OpenIFS coupled to NEMO, the Integrated Forecasting System (IFS) coupled to NEMO, the ICON model with both atmosphere and ocean, the DYNAMICO model coupled to NEMO) covering both weather and climate configurations as well as well-established models and models that are currently under development. Furthermore, Ocean-only simulations with the NEMO models are also pushed to high spatial resolution (1/36 degree; ~3 km). This deliverable will give an overview of the main components and algorithms that are used by the different models and outline challenges for the use of pre-exascale machines for weather and climate simulations at very high resolution. The document also includes scalability plots when running the various models on several supercomputers. Furthermore, the efforts in work package 1 regarding the coupling infrastructure and model in- and output (I/O) will be described.

Presentation held by Sylvie Joussaume at the European Big Data Value Forum, 22-November-2017, Versailles (FR)

Presentation held by Sylvie Joussaume at the European Big Data Value Forum, 22-November-2017, Versailles (FR)

This report provides an architectural view of all the data management components (“data theme”) set up in the ESiWACE2 project. It provides insights into the design of the ESDM interface for supporting data-intensive post-processing, analytics and visualization (PAV) and case studies from the weather and climate change domain. The document addresses the milestone MS5.1 of the ESiWACE2 project. More in detail, starting from background work and the description of the main use cases targeted by the ESDM, a set of technical requirements for the ESDM unified API has been identified, with a strong focus on the new PAV layer for in-flight analytics and visualization. The general ESDM architecture is introduced, highlighting the main components for I/O and processing. The proposed architecture tries to address the identified requirements and provides the guidelines to steer the development stage through the definition of the main architectural components, the relationship among these and some preliminary technical details.

The central objective of work package 1 of ESiWACE2 is to develop and run coupled model simulations at a resolution as high as possible and with a throughput rate of at least one simulated year per day (1 SYPD) producing full model output. This throughput would be sufficient to use these simulations for operational weather predictions and potentially also to run the models for a couple of decades, to address scientific questions that are related to climate change. Four model configurations are considered (EC-Earth with OpenIFS coupled to NEMO, the Integrated Forecasting System (IFS) coupled to NEMO, the ICON model with both atmosphere and ocean, the DYNAMICO model coupled to NEMO) covering both weather and climate configurations as well as well-established models and models that are currently under development. Furthermore, Ocean-only simulations with the NEMO models are also pushed to high spatial resolution (1/36 degree; ~3 km). This deliverable will give an overview of the main components and algorithms that are used by the different models and outline challenges for the use of pre-exascale machines for weather and climate simulations at very high resolution. The document also includes scalability plots when running the various models on several supercomputers. Furthermore, the efforts in work package 1 regarding the coupling infrastructure and model in- and output (I/O) will be described.