The Strength of a Common Goal
ECMWF is an intergovernmental organisation supported by 34 European States. It provides forecasts of global weather to 15 days ahead as well as monthly and seasonal forecasts. The National Meteorological Services of Member and Co-operating States use ECMWF’s products for their own national duties, in particular to give early warning of potentially damaging severe weather.
It operates a sophisticated prediction model of the global atmosphere and oceans known as the Integrated Forecasting System (IFS), running on some two and a half million lines of codes. Operational since 1994, the IFS is constantly updated to add new features to adapt to the latest HPC advances.
ECMWF has been using supercomputers since 1977. It operates one of the largest supercomputer facilities of its type in Europe for meteorology worldwide and holds the world’s largest archive of numerical weather prediction and observational data. Member and Co-operating States can access ECMWF’s basic computing facilities, the meteorological archive, and temporary tape storage. Member States also have access to the supercomputers and permanent tape storage.
Progress in numerical weather prediction is intimately connected with progress in supercomputing. Over the years, more computing power has enabled us to increase the skill and detail of our forecasts. This has brought huge value to society, not least through early warnings of severe weather. But as the forecasting system becomes more complex, with current computing architectures, it will soon be impossible to issue forecasts within schedule and at a reasonable cost. Supercomputer energy consumption at ECMWF would have to increase unviably, from about 4 megawatts today to perhaps 50 megawatts or more in ten years’ time, if the more complex forecasting systems of the future were to be run on the current architecture.
A new generation of computing systems with exascale capabilities promise much greater energy efficiency – but they will rely on parallel processing at levels to which current NWP codes are not adapted. Changes are needed throughout the entire NWP processing chain if we are to exploit these new opportunities for energy efficiency.
ECMWF’s Scalability Programme, launched in 2013, brings together meteorological modellers, computer scientists and hardware providers from around the world for a coordinated approach to hardware and software development.
This ten-year programme encompasses the entire NWP processing chain, from processing and assimilating observational data to delivering forecasts to Member States. Projects will cover six main areas:
• Observational data processing
• Data assimilation
• Numerical methods
• Numerical data processing
• IFS code adaptation
• Computer architecture support
Florence Rabier first joined the European Centre for Medium-Range Weather Forecasts (ECMWF) as a consultant during her PhD in 1991, then as a scientist in data assimilation for 6 years in the 1990s. She came back from Météo-France in October 2013 to take up the position of Director of the newly formed Forecast Department. She is the Director General of ECMWF since January 2016.
Dr Rabier is also an internationally recognised expert in Numerical Weather Prediction, whose leadership has greatly contributed to delivering major operational changes at both ECMWF and Météo France. Dr Rabier is especially well known within the meteorological community for her key role in implementing a new data assimilation method (4D-Var) in 1997, which was a first worldwide. Her career so far has taken her back and forth between Météo France and ECMWF.
Dr Rabier has experience in both externally-funded and cross-departmental project management. She led an international experiment involving a major field campaign over Antarctica, in the context of the International Polar Year and THORPEX. She has been awarded the title of “Chevalier de la Légion d’Honneur”, one of the highest decorations awarded by the French honours system.
Weather and Climate Services and need of High Performance Computing (HPC) Resources
We are entering a new era in technological innovation and in use and integration of different sources of information for the well-being of society and their ability to cope with multi-hazards through weather and climate services. New predictive tools that will detail weather conditions down to neighbourhood and street level, and provide early warnings a month ahead, and forecasts from rainfall to energy consumption will be some of the main outcome of the research activities in weather science over the next decade.
As weather and climate science advances, critical questions are arising such as about the possible sources of predictability on weekly, monthly and longer time-scales; seamless prediction; the development and application of new observing systems; the effective utilization of massively-parallel supercomputers; the communication, interpretation, and application of weather-related informatio
n; and the quantification of the societal impacts. The science is primed for a step forward informed by the realization that there can be predictive power on all space and time-scales arising from currently poorly-understood sources of potential predictability.
Globally the tendency of weather forecasting is moving towards impact-based direction. Besides forecasting the physical parameters, like temperature, wind and precipitation customers and general public are becoming more interested in the impacts of weather and climate phenomena. These are for example traffic disturbances, impacts on energy availability and demand, impacts on agriculture and tourism. In the case of climate services the customers are e.g. investors, industry, insurance sector, construction companies, consumer businesses and agriculture. Besides the impact-based forecasting there is a tendency to move towards multi-hazard forecasting. As an example the Japanese earthquake led to a severe tsunami, which caused the Fukushima nuclear accident with dispersion of radioactive release to both atmosphere and ocean.
The further evolution of HPCs is having an impact on weather and climate service providers globally. Most of the national services cannot afford the largest available computers for running high-resolution
NWP and climate models with advanced physics. A key question is the possibility to get access to products provided by using the best models run at the largest HPCs. At the moment several countries and the European Commission are moving towards open data policies with free access to NWP model data. This allows also countries without own HPC resources and modelling skills to get access to state of the art products. On the other hand the growing interest of private sector in NWP modelling may have an opposite impact. The global strength of meteorological community so far has been relatively free exchange of know-how and the ability to collaborate across national borders, which is a key factor behind the success of the ECMWF.
At the moment the national weather services are considering how to ensure access to better HPC resources. For example the five Nordic and three Baltic countries have agreed to seek for a joint HPC solution to enhance the value for money of national HPC investment resources. It is also discussed whether the countries should in the future buy own HPCs or rather get access to resource owned by e.g. private sector.
Born 1961 in Helsinki, Finland, married, 2 sons (-90, -96), 3 daughters (-93, -93 & 2000)
PhD in Meteorology, Helsinki Univ., physics department 1993
Management training: Helsinki Univ. Economics 1998 & 2004
Secretary-General of the World Meteorological Organization. Elected for a mandate from 2016 to 2019
Director-General of Finnish Meteorological Institute 2002-2005, 2007-2015
• Government weather, climate and marine institute, ~680 employees, annual budget ~78 M€
o Doubling of external funding level 2002-2015 (now 40 % of budget)
o Best government institute in Finland 2005-2015, High customer and staff satisfaction
o One of the leading atmospheric research institutes in Europe
o Emphasis on supporting sister organizations world wide, 2015 in 58 countries
Director of Development and Regional Activities department of WMO, 2005-2007
• Staff in Geneva, Paraguay, Costa Rica, Nigeria, Kenya, Bahrain and Samoa
o Reorganization of the department with more emphasis on serving the needs of Members
o Establishment of Resource Mobilizing and Least Developed Country Offices,
Research professor on Remote Sensing/FMI 2000-2002, Head of research, Scientist 1986-1999
• Coordination of Earth Observation research and development of FMI with ~30 employees
• Satellite, global change and atmospheric pollution research and development projects funded by ESA, EUMETSAT, NASA, European Commission and national agencies
• Vice-chairman of European Commission Panel on Atmospheric Science 1995-2002
• Author of ~50 peer reviewed scientific publications on satellites, climate and pollution
• Member of WMO Executive Council 2008-
• Chairman of EUMETSAT Council 2010-2014 (European Meteorological Satellite Org.)
• Chairman of EUMETNET Council 2003-2005 (European Meteorological Infrastructure)
• Member of ECMWF Council 2002-5, 2007-15, (European Weather Centre)
• Principal Delegate of Finland to IPCC 2007-15 (Intergovernmental Panel on Climate Change)
• Chairman of the board of Univ. of Eastern Finland 2010-
• Alumni of the year of University of Helsinki 2012
• Member of the Finnish Science Academy 2010-
• Member of Finnish Technology Academy 2015-
• Member of the advisory and board of Directors of Fortum Energy Company (turnover 6.2b€) 2010-15
• Member of the Finnish Space and Arctic boards 2003-5, 2009-15
Languages: Finnish (mother tongue), English (Excellent), German (moderate), Swedish (moderate), French (basics), Russian (basics)