European Centre for Medium-Range Weather Forecasts ECMWF STRATEGY 2016-2025: THE STRENGTH OF A COMMON GOAL European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL ECMWF’s purpose is to develop a capability for medium-range weather forecasting and to provide such weather forecasts to the Member and Co-operating States ECMWF is complementary to the National Meteorological Services and works with them in research, numerical weather predictions, supercomputing and training. European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Evolution of ECMWF medium-range skill over the past 35 years European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Evolution medium-range skill (NH Z500>0.8) European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL New emphasis: Percentage of large 2m temperature errors in the ensemble European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL The size of the challenge # 1 The difficulty: Sharp ensembles two weeks ahead Severe weather: Hurricane Bertha 6-9 days 2-5 days European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts The size of the challenge # 2 THE STRENGTH OF A COMMON GOAL Springtime cold spell over Europe 2m Temperature, 17-23 April 2017 European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Forecast targets by 2025 Ensemble predictions of high impact weather up to two weeks ahead Seamless approach, aiming towards predictions of large scale patterns and regime transitions up to four weeks ahead and global-scale anomalies up to a year ahead European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts Goals THE STRENGTH OF A COMMON GOAL Research at frontiers of knowledge Ensemble-based analyses and predictions that raise the international bar for quality and operational reliability reaching a 5 km horizontal resolution European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL How do we achieve these goals? Observations High resolution ensemble Earth-system Scalability Funding People European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Use of satellite data at ECMWF ECMWF processes an average of 40 million observations every day, from over 70 instruments. Collaboration with sister organisation EUMETSAT, and also ESA, CMA, JMA, NASA, NOAA among others ensures that ECMWF has access to the observations meteorology requires. European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Contributions to WIGOS data quality monitoring system In situ observation monitoring with data from ECMWF, NCEP, DWD and JMA Data availability and quality European Centre for Medium-Range Weather Forecasts

IRMA operational v. 5km As IRMA had such large consequences in the Caribbean including for some of our MSs, we used this case study to have a first view of what a 5 lm ENS could look like. Still too expensive to run a large set of experiments on current computer, and no DA, but one single forecast with 51 members. Trajectories not too dissimilar although ens mean better. Fit to observed surface pressure remarkably improved.

European Centre for Medium-Range Weather Forecasts Pillar 1– we started a few years back to experiment with an Earth system approach, and the new strategy places a strong emphasis on the positive results gained from such approach, and this a direction we are continuing to develop and implement. European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Availability and use of GTS snow depth information Impact of revised IMS snow cover assimilation on rms 1000hPa temperature and humidity errors   European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Impact of coupled assimilation during a cyclone The dotted line shows the time series of ocean temperature observations at a depth of 40 metres by an Argo float located on the track of the cyclone Phailin. European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts Seamless modelling systems THE STRENGTH OF A COMMON GOAL Scientific and infrastructure advantages of convergence of approaches across timescales Seasonal SEAS5 only differs from the 43r1 ENS extended (monthly) system when testing demonstrated clear improvement in forecast skill or mean state Horizontal (Tco319/ORCA25) and vertical resolution (L91/L75) identical Improvements found on one timescale applicable for others Decreasing non-orographic gravity wave drag ameliorates the effect of stratospheric temperature and winds biases on the QBO Preferred seasonal setting now found suitable for adoption for medium-range to monthly SEAS4 43r1 SEAS5 Forecast lead (months) Layout out the targets for our forecasts European Centre for Medium-Range Weather Forecasts

ECMWF Scalability Programme Governance: ECMWF, Member states, Regional consortia Projects: Observation processing: Lean workflow in critical path Object based data store Screening/bias correction Data assimilation: Flexible algorithms (C++) IFS integration Coupling with ocean and sea-ice Parallel minimization Numerical methods: Numerical methods h/v/t-discretization, multiple grids Prognostic variables, coupling Model output processing: Broker-worker workflow Near-memory processing Data compression Code adaptation: · Benchmarking · Vectorization · Programming models · Precision · DSL/libraries · Transition to operations Computer architecture support: · Benchmarking · Novel architectures · Compilers · Vendor support · Transition to operations European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts Scalability across the NWP chain: Single precision to deliver efficiency gains Moving to another prospect, the possibility to run on single precision. How much faster can our weather forecasts run with single precision? The magnitude of savings with single precision depends on various factors. When computing speed is limited by memory bandwidth, single precision can typically increase performance by a factor of two since data volume is halved. If floating point operations are the performance bottleneck, the speed-up factor will depend on the level of code vectorization. Single precision is twice as efficient when the code is fully vectorised. For forecast simulations with the ECMWF IFS on our Cray supercomputer, single precision runs are typically ~40% faster compared to double precision references. However, those numbers depend on the model resolution and the number and configuration of processors that are used in parallel. The use of single precision will potentially allow significant savings of computing cost in the near-term future and an increased throughput on supercomputers. Single precision will also help us to perform tests with the IFS at very high resolution. For these simulations, a very large number of computing nodes is needed to fit the model into the memory that is available. Single precision is reducing memory requirements significantly and less nodes can be used. Tests with single precision in simulations at 2.5 and 1 km global resolution are being performed at the moment. Another interesting factor of this project is that it started as part of a type of collaboration that we have created a few years back, the ECMWF Fellowship, whereas we invite an academic to work with our teams on a specific area of work. Single precision was started as part of Tim Palmer and his team at Oxford University’s Fellowship. Surface temperature in degree Celsius for five day forecasts for 8th January 2017 0:00 UTC. This date is during the European cold wave that caused very low temperature in Eastern and Central Europe. Results are shown for single precision and double precision simulations at 9km (TCo1279) resolution (left and middle) and the analysis as a reference (right). Differences between single and double precision are very small. European Centre for Medium-Range Weather Forecasts

ExtremeEarth co-design approach How What

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL Collaboration Partnering with universities, research institutes and NHMSs Strategic partnerships in Earth observation Providing collaborative data bases (e.g. TIGGE, S2S) European Centre for Medium-Range Weather Forecasts

European Centre for Medium-Range Weather Forecasts THE STRENGTH OF A COMMON GOAL In summary Operational forecasts AND Research High-impact weather, regime transitions and global-scale anomalies Integrated ensemble at 5km resolution by 2025 Earth-System model and analysis Scalable computation Environmental information services: Copernicus Collaboration European Centre for Medium-Range Weather Forecasts