Ongoing research project | 2019 to 2021
The LEXIS project is focused on developing a large-scale HPC & Cloud BigData analytics platform to demonstrate high Technical-Readiness-Level software and infrastructure for three pilot use cases, which each have very high industrial or socio-economical impact: (1) Aeronautics (2) Weather & Climate (3) Catastrophe Alert systems. The focus will primarily be on how HPC and BigData/Cloud systems interact; how they can share data; and methods to compose workflows of tasks running on both cloud and HPC systems.
For the weather & climate pilot, there are several complex workflows each consisting of various meteorological components:
- Conventional observation data sources (satellite, radar, etc.)
- Unconventional observation data sources (personal weather stations and other IoT devices)
- Global weather models running on HPC in the UK/Italy (by ECMWF)
- Regional weather models running on HPC in Italy, Germany or Czechia (by CIMA)
- Cloud-based applications running in Germany or Czechia including:
- Continuum hydrology simulations (by CIMA)
- RISICO fire-risk simulations (by CIMA)
- Limagrain agricultural impact models (by Numtech)
- Extreme Rainfall Detection System (ERDS, by ITHACA)
Leveraging ECMWF's expertise in data handling and data modelling, our role in the project is to design and develop a Weather & Climate Data API (WCDA), which will facilitate the interchange of data between the aforementioned components.
The WCDA, serving data through MARS and FDB, will be a comprehensive solution to the problem of sharing time-critical meteorological data between HPC and cloud models. Beyond LEXIS, it will lay the foundations for the convergence of HPC and Cloud at ECMWF, and enable data transfer between ECMWF's operational forecasts and the European Weather Cloud.