The Fondazione Centro Euro-Mediterraneo sui Cambiamenti Climatici (CMCC) is a non-profit research institution, established with the financial support of the Italian Ministry of Education, University and Research and the Ministry of the Environment Land and Sea, the Ministry for Agricultural and Forestry Policies and the Ministry of Finance.
CMCC involves and links private and public institutions jointly investigating multidisciplinary topics related to climate science research. CMCC research activities are distributed among eleven research divisions that share different knowledge and skills in the field of climate science: Advanced Scientific Computing (ASC) Division; Climate Simulation and Prediction (CSP) Division; Economic analysis of Climate Impacts and Policy (ECIP) Division; Impacts on Agriculture, Forests and Ecosystem Services (IAFES) Division; Ocean modelling and Data Assimilation (ODA) Division; Ocean Predictions and Applications (OPA) Division; Risk Assessment and Adaptation Strategies (RAAS) Division; Regional Models and geo-Hydrological Impacts (REHMI) Division; Sustainable Earth Modelling Economics (SEME); Information Systems for Climate science and Decision-Making (ISCD); Innovative Platforms for Science Outreach (IPSO).
CMCC’s mission is to investigate and model our climate system and its interactions with society to provide reliable, rigorous, and timely scientific results, which will in turn stimulate sustainable growth, protect the environment, and develop science driven adaptation and mitigation policies in a changing climate. CMCC collaborates with experienced scientists, economists, and technicians, which work together in order to provide full analyses of climate impacts on various systems such as agriculture, ecosystems, coasts, water resources, health, and economics. CMCC also supports policymakers in setting and assessing costs, mitigation, and adaptation policies.
CMCC operates its own Supercomputing Center (located in the “Ecotekne” Campus – University of Salento, Lecce). It is the only computational facility in Italy specializing in Climate Change research (https://www.cmcc.it/super-computing-center-scc). Zeus, the supercomputer currently in operation, is based on 348 Lenovo SD530 bi-processor nodes (for a total of 12.528 cores) all interconnected by means of an Infiniband EDR network. The HPC system has a computing power (theoretical peak performance) of 1.202 TFlops. CMCC plans to have its new HPC facility (named Juno) in production starting from the end of the first quarter of 2022. The new supercomputer Juno will have a computing power (theoretical peak performance) of about 1.134 TFlops and will be based on the new Intel processors generation just announced this month (3rd Generation Intel Xeon Scalable codenamed “Ice Lake”) and also on the latest generation of NVIDIA GPU (NVIDIA Ampere architecture).
Development and maintenance of NEMO configurations
The NEMO model is a key part of the CMCC modelling framework to operate ocean forecasting systems and produces ocean reanalysis products on global and regional scales.
CMCC develops global configurations from coarse (1º) to eddying (1/16º) resolution. Resolutions of 1º and 1/4º are used as part of the CMCC coupled climate and Earth-System models and form the ocean component of the CMCC contribution to the CMIP6 experiments.
CMCC develops and operationally maintains:
- Global Ocean Forecasting System GOFS16, based on the global ocean-sea ice NEMOv3.4-LIM2 model that has been set up following an eddy-resolving configuration (GLOB16) developed at CMCC. The data assimilation component, called OceanVar, a 3DVar scheme developed and maintained by CMCC, already employed at 1/4° resolution and recently adapted to bear large amount of data through a hybrid-parallelization scheme. The system assimilates near real time in-situ temperature and salinity profiles, along-track Sea Level Anomaly data from satellite altimeters and Sea Surface Temperature satellite retrievals;
- the CMCC Global Ocean Reanalyses System (CGLORS) at 1/4° resolution that consists of two different products v5 and v7, starting in 1980 and 1993 respectively, and extended twice a year. The ocean-ice general circulation model consist of NEMO-LIM2 model, coupled to OceanVar which assimilates Sea Level Anomaly from along track satellite data and vertical profiles of temperature and salinity from insitu observations; the satellite Sea Surface Temperature from satellite is used to correct the heat flux;
- the Mediterranean Forecasting System, based on NEMO v3.6, at 1/24° horizontal resolution (~4 km) and 141 vertical levels, 2-way coupled with WW3 v3.14, forced by operational ECMWF IFS analysis and forecast atmospheric fields. It implements open boundary conditions at the Atlantic Ocean and Dardanelles strait and explicitly resolves tides (8 tidal components). The system is coupled to OceanVar, which assimilates near real time Sea Level Anomaly from along track satellite data and vertical profiles of temperature and salinity from insitu observations; the satellite Sea Surface Temperature from satellite is used to correct the non-solar heat flux. The processing system produces every week 14-days analysis, and every day 1-day simulation and 10-days forecast;
- the Mediterranean Reanalysis System, initialized in 1985 and extended yearly, based on the operational Forecasting System, but with several differences: no coupling with waves, no explicit tidal representation, closed boundary at the Dardanelles strait, forced with ECMWF ERA5 atmospheric data and assimilating reprocessed observations;
- the Mediterranean Interim System, based on the Reanalysis System, which provides an operational monthly extension of the reanalysis time series assimilating near real time observations;
- the Black Sea Forecasting System, based on NEMO v4.0.2, at 1/40° horizontal resolution (~2.5 km) and 121 vertical levels, forced by operational ECMWF IFS analysis and forecast atmospheric fields. It implements open boundary conditions at the Marmara Sea to provide the best representation of the inflow/outflow through the Bosphorus Strait. The system is coupled to OceanVar for the assimilation of near real time satellite along track Sea Level Anomaly, Sea Surface Temperature and insitu temperature and salinity profiles. The processing system produces 3 (daily cycle) or 14 (weekly cycle) days analysis, 1-day simulation and 10-days forecast;
- the Black Sea Reanalysis System, based on NEMO v3.6, at 1/36° x 1/27° horizontal resolution and 31 vertical levels, forced by ECMWF ERA5 atmospheric reanalysis. It implements closed boundary condition at the Bosphorus Strait as surface boundary condition (“inverse river”). The system is coupled to OceanVar for the assimilation of reprocessed observations. The timeseries is initialized in 1988 and delivered to users starting from 1993;
- the Black Sea Interim System, based on the Reanalysis System, for the operational monthly extension of the timeseries, based on assimilation of available near real time observations.
The Mediterranean and Black Sea systems operate in the framework of Copernicus Marine Service (CMEMS) for the provisioning of the past reconstruction and forecasting of the ocean state in the regional basins.
Further information on CMCC models are available here.
CMCC supports and contributes to the development and maintenance of the different components of the NEMO framework and participates in the ongoing scientific discussions in the working groups on HPC, TOP and Sea Ice.
In particular, CMCC provides needed expertise to contribute to the NEMO computational performance optimization. Due to the experience on scalability improvement, optimization and parallelization of numerical models, and documented contributions to the international exascale roadmaps for next generation climate models, CMCC works on code re-design in order to better exploit the modern HPC systems capability. This includes the investigation of hybrid parallel approaches to improve current NEMO scalability on many-cores architectures, also equipped with accelerators/coprocessors, and single-core performance investigation to increase peak performance exploitation.
CMCC is also actively working on the development of the TOP component for tracers transport to improve the orthogonality between physical and biogeochemical processes and it maintains the external coupling interface (e.g., for the Biogeochemical Flux Model).