Research Highlights for Predictability of Weather and Climate

Members of the predictability group organised a workshop to discuss the value of machine learning to improve weather and climate forecasting. The four day workshop, held at Corpus Christi College (2-5 September), was organised by Tim Palmer, Hannah Christensen, Matthew Chantry and Philippa Towler from the predictability group alongside Peter Duben from ECMWF. The workshop welcomed attendees from all around the world with representatives from academia, operational forecasting centres and industry.

A new paper that is now accepted in the Monthly Weather Review shows that single precision (32 bits) is sufficient for almost all parts of a weather forecast model to generate accurate simulations. In close collaboration with ECMWF we could show that single and double precision produce virtually the same quality in results for both short and long term simulations with the Integrated Forecast System. Savings due to the use of single precision are significant (~40%).

Sam Hatfield has won the Runner Up prize at the Famelab competition UK final, held at the Cheltenham Science Festival on June 3rd. This competition challenges contestants to communicate a science topic to a public audience and a panel of judges in under 3 minutes. The judges choose the best entries based on three criteria: content, clarity and charisma. Prizes for first place, runner up and the audience favourite are awarded. Regional heats are held across the country to decide which 8 contestants will take part in the UK final.

Sam Hatfield has won the prize for the best paper submitted to the Platform for Advanced Supercomputing (PASC) conference 2019, a leading high-performance computing conference in Europe. The paper, co-authored with Matthew Chantry, Peter Dueben and Tim Palmer, investigated the potential for accelerating Legendre transforms within high resolution spectral weather models, such as the ECMWF model IFS, using reduced-precision arithmetic. The prize recognizes cutting edge work within the broad discipline of applied high-performance computing.

Sam Hatfield has been awarded a £1000 travel grant from Elsevier for his poster and presentation at the Environmental Research Doctoral Training Partnership Student Conference in November 2016. This was awarded to the top three presentations from second year students. Sam presented the research carried out in the first year of his DPhil on inexact computing and data assimilation.

A 7-day weather forecasts is now as accurate as a 5-day forecast 20
years ago. However, there is still a widespread public misbelief that
forecasts are not reliable. Traditional meteograms, summarizing the
weather for a certain location throughout the next week, rarely contain
information about the uncertainty of a weather forecast, which may
support the little public confidence in weather forecasting.

In February 2018, Hannah Christensen was selected for the American Meteorological Society’s Early Career Leadership Academy. The Academy is intended for high-level performers and aims to build and sustain a diverse network of early career leaders in weather, water, and climate science.

In April 2018, Hannah Christensen took up a NERC Independent Research Fellowship in Oxford's AOPP on 'Reliable Climate Projections: The Final Frontier for Stochastic Parametrisation'. The NERC Independent Research Fellowship (IRF) scheme is designed to develop scientific leadership among the most promising early-career environmental scientists, by giving all fellows five years' support, which will allow them sufficient time to develop their research programmes and to gain international recognition.

A new paper from our group which has recently been published in the Quarterly Journal of the Royal Meteorological Society as part of the Special Collection "25 Years of Ensemble Forecasting". It studies the so-called "signal-to-noise paradox" in the atmosphere over the North Atlantic region on seasonal time scales in several seasonal forecasting data sets.

In April 2018, the London investment fund Winton launched an academic contest for UK universities, aimed at predicting UK weather over a 6 month period. The purpose of the game was to predict, for each month from April to September, what the average UK surface temperature and total UK rainfall over the month would be. Predictions were made by buying shares in a `weather market’ platform, where a share could be bought representing arbitrary possible outcomes for each month.

Sam Hatfield has won the Oxford final of the Famelab science communication competition. The competition challenges participants to deliver a complex scientific subject or idea in only three minutes to a live audience and a panel of three judges. Participants are judged based on three criteria: content, clarity and charisma. Sam delivered a talk about chaos theory and the difficulties of hurricane prediction, focusing on the case of Hurricane Sandy. He will go on to compete at the national finals at the Cheltenham Science festival in June.

Tim Palmer has written a Sci-Fi story, inspired by Isaac Asimov’s Nightfall, which warns of the dangers of inadequate models for predicting climate change. The story is called Sunrise and is available to download (please see the link under "Attachments").

Dr. Mio Matsueda has received the Syono award (Meteorological society of Japan) for his research achievements in the field of predictability of atmospheric phenomena and, for his excellent work in creating the TIGGE museum using operational medium-range ensemble forecasts.

The Syono medal is awarded to a scientist below 40 who makes a great contribution to meteorology and meteorological technique.

Congratulations, Mio!

Tim Palmer was awarded an Honorary Doctorate Degree by the University of Bristol at the University Congregation on 21st July 2016. In his acceptance speech, Tim noted the important role that collaboration with scientists around Europe had played in his career. He warned of the damage that Brexit may cause in the future, in particular for UK scientists’ ability to play leadership roles in collaborative EU projects.

Tim Palmer has written a personal perspective on Climate Modelling, published in Proceedings of the Royal Society A. Please click here to read the article.

We welcome two new people to the Predictability of Weather and Climate Group! Kristian Strømmen is a new postdoctoral researcher, who will be looking at the impact of stochastic parametrisation schemes in climate models as part of the European Commission project PRIMAVERA.

On Friday 13th November 2015 Hannah Christensen will give a prize lecture (for the RAS Keith Runcorn Thesis Prize) at The Geological Society titled “Can randomness reduce uncertainty? The use of stochastic physics in weather and climate prediction”.

For further details visit the RAS events web page.

Tim Palmer has been awarded the 2014 Institute of Physics's Dirac Gold Medal for theoretical physics, for his work on probabilistic weather and climate prediction, which, the judges said, "led a revolution in the fields of weather and climate and benefitted all sections of society". This is the first time this award has been given in the geosciences, previous winners being Peter Higgs, Stephen Hawking and Roger Penrose. It is clearly a great boost for meteorology that it has been recognised in this way by the theoretical physics community.

We are happy to announce that a special issue, which was edited by Tim Palmer, Peter Düben, and Hugh McNamara, on "Stochastic modelling and energy-efficient computing for weather and climate prediction" has been published in the Philosophical Transactions of the Royal Society A.

Tim Palmer has been elected a Fellow of the American Geophysical Union in the "Class of 2013". His citation reads: "For fundamental understanding of the predictability of weather and climate, and for pioneering the tools to estimate such predictability." Only one in a thousand members of the AGU is elected to Fellowship each year.

Tim Palmer gave the 9th Dennis Sciama Memorial Lecture "Lorenz, Gödel and Penrose: New perspectives on determinism and unpredictability - from climate prediction to fundamental physics".

Tim Palmer gave a short presentation on a "CERN for Climate Change" as part of an Oxford Martin Ideaslab, at the World Economic Forum in Davos in January 2013. The web videolink to the Ideaslab talks is here.

18 - 21 March 2013, Oriel College Oxford

Predicting future weather and climate as precisely and reliably as possible is an important and urgent scientific challenge with numerous societal applications. Probabilistic forecasts, which produce estimates of the likelihood of future events rather than a single prediction, better reflect uncertainties in the core processes and dynamics. Certain important questions must be adressed in order to build and refine such a forecast system:

Tim Palmer FRS, the President of the Royal Meteorological Society, has been awarded the European Meteorological Society (EMS) Silver Medal for 2012 for his pioneering work on predictability of weather and climate and for his outstanding skills and commitment in communicating the underlying concepts to non-specialists and the general public in Europe and beyond. The award ceremony will take place on Wednesday 12 September 2012 during the EMS Annual Meeting organised in Łódź in Poland from 10 to 14 September and Tim will be invited to give the Silver Medal Lecture.