Use case 1: Reproducible ecosystem models

Goal: Better take into account typical ecosystems only found at high latitudes (and currently poorly represented) by integrating actual data and empirical knowledge in the model development and operation.

Target audience: Experts of ecosystems, biologists and environmental scientists with no background in climate modeling

Use case owner: Hui Tang, University of Oslo, Norway

Partners:

• Lund University (Sweden)
• University of Oslo (Norway)
• Natural History Museum (Norway)

What?

Provide:

• a virtual laboratory for developing and testing parameterizations of ecosystems and related processes to scientists whose expertise is not numerical modeling, and thereby increase the pace of model improvement from the bottom up;
• workflow management tools for operating these ecosystem models in a reproducible manner to realize their full potential without need for constant technical assistance;
• operational tools and infrastructure to facilitate the exploitation of ecosystem modeling for a wider audience and by non-specialists.

In concrete terms, we will aim at working on:

• single-point/region simulations (FATES and/or LPJ-GUESS) at a number of specified locations in particular close to monitored sites;the number of sites and location should be flexible enough so that new sites can be easily added.
• allowing researchers to access and share data collected either during field campaigns or on a regular basis at selected measurement stations;
• allowing researchers to control data access: researchers need to be able to choose who they want to share with and what they want to share; however, metadata should always be available to everyone.
• offering the necessary tools to analyze and visualize the resulting data and in particular comparison between observations and model outputs.

How?

• Integration of FATES (Functionally Assembled Terrestrial Ecosystem Simulator) and/or LPJ-GUESS Ecosystem Model as new tools in Galaxy;
• Integration in Galaxy of tools for analyzing and visualizing the model outputs; ideally, we would develop the visualization tool using Galaxy Climate JupyterLab and we would “turn” jupyter notebooks into Galaxy tools.
• Develop target online tutorials to advertise Galaxy FATES tools.
• Publication of in-situ measurements (with comprehensive meta-data, citation, access control, etc.).

Reference: PeCAN project

EOSC-Nordic tasks

Needs to be covered by the following EOSC-Nordic tasks:

• T5.2.1: Cross-border data processing workflows
• T5.2.2: Code Repositories, Containerization and “virtual laboratories” (M1-36)
• T5.3.1: Integrated Data Management Workflows (M1-36)

Use case 2: FAIR Climate data for the Nordics

Goal: Provide FAIR climate data to all the communities interested in climate mitigation and climate change impact assessment. This use case aims at serving NeIC NICEST-2.

Target audience: Scientists, local authorities, policy makers and general public.

Use case owner: Hamish Struthers, National Supercomputing Centre, Sweden

Partners:

• DMI, Denmark
• CSC, Finland
• NORCE, Norway
• MetNo, Norway
• UoI, Iceland

What?

Facilitate publication and access to climate data (including but not restricted to CMIP) from Galaxy.

How?

• Integration of CMIP data in Galaxy (stored on Nordic ESGF nodes);
• Integration of other climate data in Galaxy (could also include model outputs from research work not directly related to CMIP, and any other relevant observations);
• Deployment of simple tools for visualizing climate data over the Nordics from Galaxy;
• Deployment of statistical tools (global mean, regional mean, time series, climate indices) for characterizing climate change in the nordic countries;
• Deployment of Machine Learning and Deep Learning tools in Galaxy (for example to identify local trends, develop classifications, extrapolate mitigation impacts, have an insight into downscaling, etc.).

EOSC-Nordic tasks

Needs to be covered by the following EOSC-Nordic tasks:

• T5.2.1: Cross-border data processing workflows
• T5.2.2: Code Repositories, Containerization and “virtual laboratories” (M1-36)
• T5.3.1: Integrated Data Management Workflows (M1-36)
• T5.3.2: Machine actionable DMPs (M1-36)

Use case 3: A community Virtual Laboratory for developing Climate diagnostics for the Nordics

Goal: Share and develop Earth System Model EValuation diagnostics and analysis along with the related data. This use case aims at serving NeIC NICEST-2.

Target audience: Climate experts and data scientists in the field.

Use case owner: Øyvind Seland, Norwegian Meteorological Institute, Norway

Partners:

• MetNo, Norway
• NORCE, Norway
• University of Oslo, Norway
• FMI, Finland
• DMI, Denmark
• NERSC, Norway
• University of Helsinki/INAR

What?

Provide a platform for facilitating the development of:

• new Earth System Model Evaluation Tool (ESMValTool) diagnostics for the Nordic regions (e.g., polar lows, European and Greenland blocking events) with strong emphasis on (high-latitude) observations;
• assemble reference datasets to validate models;
• ad-hoc analysis and visualization.

With these new tools, researchers will be able to compare outputs from different climate models in the Nordic countries (where numerical issues are often exacerbated and because the tuning was made difficult due to the relative lack of reference data) in order to identify deficiencies and develop strategies to improve the models e.g. EC-EARTH and NorESM (eventually combining the best parameterizations/modules).

How?

• Deployment of an e-infrastructure similar to Pangeo e.g. binderhub and jupyterhub;
• Customized software stack e.g., pangeo software stack and additional packages such as ESMValTool for developing ad-hoc diagnostics for the Nordics;
• Access to HPC, including GPUs to be able to handle, analyze and visualize large amount of climate data;
• Access to CMIP data, re-analysis, model outputs produced by researchers (see use case 2);
• Access to observations (see use case 2).
• Develop target online training material to advertise the tools.

EOSC-Nordic tasks

Needs to be covered by the following EOSC-Nordic tasks:

• T5.2.1: Cross-border data processing workflows
• T5.2.2: Code Repositories, Containerization and “virtual laboratories” (M1-36)
• T5.3.1: Integrated Data Management Workflows (M1-36)
• T5.3.2: Machine actionable DMPs (M1-36)

Use case 4: A community Virtual Laboratory for the development of machine-learning-based parameterizations and sub-models for clouds and land-surface processes in Earth System Models.

Goal: Improve existing parameterizations and prepare Earth System Models for future architectures by using machine learning.

Target audience: Climate experts and data scientists in the field.

Use case owner: Risto Makkonen, Finnish Meteorological Institute, Finland

Partners:

• University of Oslo, Norway
• University of Helsinki, Finland
• University of Iceland, Iceland

What?

Provide a platform for facilitating the development of:

• new parameterizations for Earth System models using machine learning;
• sub-models for clouds and land-surface processes in ESM using machine learning;

The framework should also facilitate the introduction of these new parameterizations in existing ESMs.

How?

• Interactive Climate JupyterLab tool in Galaxy with CESM, CTSM environment for developing, testing and running Earth System Modelling experiments.
• Publication of ESM workflows for creating and publishing Research Objects.
• Development of online tutorial to advertise Interactive Climate JupyterLab for running Earth System Models.

EOSC-Nordic tasks

Needs to be covered by the following EOSC-Nordic tasks:

• T5.2.2: Code Repositories, Containerization and “virtual laboratories” (M1-36)