Climate Analytics Service
The ENES Climate Analytics Service (ECAS) provides free access to server-side near-data processing capabilities through four different installations hosted at DKRZ, CMCC, IPSL and UKRI/CEDA premises, which offer a virtual climate research environment based on a JupyterHub and parallel back-end compute capabilities.
News!
Online training on 8-9 March
registration open here!
The ECAS integrates data and tools for scientific data analysis, manipulation and visualization, offering access to CMIP5, CMIP6 as well as CORDEX datasets from ESGF. Look at the benefits:
1. Compute-to-data: no more heavy data transfer
2. Universal availability: log in wherever you are and no need to back up
3. Parallel processing: to compute accessing more than one file at once or a big file that does not fit one core memory
The ECAS is a service with institutional-based deployments addressing national requirements and needs with four instances: DKRZ, CMCC, IPSL and UKRI/CEDA, which represent different technical implementations (or flavours). The road map of the ENES compute service architecture towards a sustainable data analytics and processing layer for climate research in Europe is being defined within the IS-ENES3 project. Find below a short description about the four ENES climate analytics service facilities at DKRZ and CMCC (IPSL and UKRI/CEDA coming soon).
Important notes: unfortunately we cannot offer unlimited resources. The ECAS only provides computer resources for pre- and post-processing (not suitable for simulation runs) in shared computing nodes. If you would like to use exclusive extensive resources, please apply to our Analysis Platforms service. Besides, this is a European service mainly addressed to the European earth systems community, the non-European communities are welcome if resources are available.
ENES Climate Analytics Service at DKRZ
The ECAS at DKRZ is a free of charge server-side data-near processing service based on Jupyter notebooks (where Python is promoted, but you can also use R and Julia) to directly load and process CMIP and other climate data.
The new ECAS instance is hosted at the German Climate Computing Center (DKRZ) Jupyterhub https://jupyterhub.dkrz.de which offers:
- support for fast computations provided via the Xarray Python package for labelled multi-dimensional arrays, which is particularly tailored to working with NetCDF files, and integrates tightly with Dask for parallel computing, and
- direct access to the CMIP data pool, that we update and maintain, because DKRZ is an ESGF data node, and it is easily accessible via the Intake Python package.
Visit our use cases repository where we demonstrate how to use ECAS resources. There is for instance the Jupyter notebook of the Summer Days Climate Index calculation with CMIP6 using Intake and Xarray that we introduce in the following 4 steps Quick Start (click on the images to enlarge them):
Quick Start
|
1. Get a DKRZ account here: once you submit the form in less than one day you will get an email from us with your user name |
2. Log in here and request to join the 1088 project:
please, add in the description the mandatory items we indicate below |
|
3. Log in to the Jupyterhub here and open your notebook:
choose a job profile depending on how much computing power and memory you need. Choose a "prepost" one if your notebook requires internet connection. "Account" stands for the project you joined, i.e. "bk1088" |
4. Directly load and use the data on your notebook:
|
Mandatory items to add in the description (step 2 in the Quick Start above): when you request to join project 1088, please specify the following:
- "I apply to be a user of the ENES Climate Analysis Service (ECAS)".
-
Summarize your activity including your affiliation, the data you need and what for, and the impact of the results (no more than 3 sentences): "I work at (your institution) in (the country your work) for (your project). I would like to use (the data you need) to calculate (your pre- or post-processing analysis). My results will be part of (publications, reports,...) and be presented at (conferences, seminars,...)" .
-
Tell us the duration of your access. By default you can use the service during 1 month and it is extendible on demand, up to 3 months. Reapplication is possible.
It will take us about one day to evaluate this info. If everything is ok, we will send you a confirmation email and you can start using the service (see step 3 above)!
 |
For any question or comment, please reach us at [Email protection active, please enable JavaScript.].
|
ENES Climate Analytics Service at CMCC
The CMCC climate analytics service instance is accessible at https://ecaslab.cmcc.it/ and provides:
-
support for fast computations through the Ophidia data analytics framework, which enables data-intensive analysis exploiting advanced parallel computing techniques and smart data distribution methods;
-
a large set of pre-installed Python libraries for data manipulation, processing and visualization (Xarray, Pandas, Matplotlib, Cartopy, etc.);
-
a data publication service to provide catalog, metadata, and access services for scientific data;
-
variable-centric CMIP data archive synchronised with ESGF catalog.
Free registration is required. Get your account here.
The Ophidia framework documentation, including starter tutorials, descriptions of all processing operators and example workflows, is available here.
How do you feel? Ready to jump to the server-side near-data computing? Just choose one of our providers and start!
By applying to use the ECAS:
- you agree to include the following acknowledgement when presenting your results in scientific articles, conferences,...: "This [insert type of result] is part of the IS-ENES3 project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 824084".
- you agree that your personal data will be processed according to the IS-ENES privacy policy.
ECAS DKRZ and CMCC instances are also EOSC-enabled thanks to its interoperability with EGI and EUDAT services, find ECAS in the EOSC service portfolio.
