Copy
Centre for the Analysis of Time Series
CATS eNews                                                              Issue 18 (October 2013)
 
This quarter, we bring you updates on the Centre's activities over the past few months, as well as highlighting some upcoming dates for your diary.

As ever, please do keep the CATS administrative team up to date with all of your CATS-related activities!
6 November - Professor Leonard Smith will give a talk at LSE entitled "Predictability, Probability(s) and Physical Insight."  This will be a preview of the Charney Lecture that Professor Smith has been invited to give at this year's AGU Conference in December. The talk will take place in Clement House room CLM2.02 at 6.30pm.  If you would like to attend please email Lyn Grove.

7-8 November - Following on from the success of our Workshop on "Uncertainty Quantification, Risk and Decision-making" last year, CATS is hosting a second event in this series, "Model (in)adequacy in policy making." This year's workshop will be held on 7 and 8 November, here at LSE.  Speakers include: Paul Embrechts (ETH), Joe Halpern (Cornell), Najar (Kellogg), Brian Hoskins (Imperial College), Jason Lowe (Met Office), Sujoy Mukerji (Oxford).  Places are limited but if you are interested in attending please email Lyn Grove.

11 November - CATS will be running a second 'Uncertainty Workshop' for researchers on the EPSRC-funded Blue-Green Cities project (Delivering and Evaluating Multiple Flood Risk Benefits in Blue-Green Cities).  The aim of this second workshop is to investigate uncertainty in greater detail - what types of uncertainties are inherent in each model, how can these uncertainties be quantified, how do unquantifiable uncertainties propagate throughout the model cascade?
Barrieu, P. "A note on pricing q-forward contracts" with Veraart, L. (submitted to Scandinavian Actuarial Journal)

Barrieu, P. "Assessing financial model risk" with Scandolo, G. (submitted to European Journal of Operational Research)

Barrieu, P. "Indifference pricing with uncertainty averse preferences" with Giammarino, F.  Journal of Mathematical Economics, Vol. 49, p.22-27 (2013).

Barrieu, P. "Monotone stability of quadratic semimartingales with applications to general unbounded quadratic BSDEs" with El Karoui, N.  The Annals of Probability, Vol.41, p.1831-1853 (2013).

Barrieu, P. "Reinsurance and securitisation in life insurance risk: the impact of regulatory constraints" with Loubergé, H.  Insurance: Mathematics and Economics, Vol.52, p.135-144 (2013).   

Barrieu, P. "Robust capital requirements with model risk" with Ravanelli, C. (submitted to Risk)

Caparelli, V., Franzke, C., Vecchio, A., Freeman, M.P., Watkins, N.W. and Carbone, V. "A spaciotemporal analysis of U.S. station temperature trends over the last century" Journal of Geophysical Research: Atmospheres, Vol.118, 1-8, doi:10.1002/jgrd.50551, 2013.

Daron, J.D. and Stainforth, D.A. (2013) "On predicting climate under climate change" Environmental Research Letters, 8 (034021). Video-abstract.

Frigg, R., Bradley, S., Du, H. and Smith, L.A. "Laplace's Demon and the Adventures of his Apprentices" forthcoming in Philosophy of Science.

Machete, R.L. (2013) "Contrasting Probabilistic Scoring Rules." Journal of Statistical Planning and Inference, 143 (10) pp.1781-1790.

Machete, R.L. (2013) "Model Imperfection and Predicting Predictability." International Journal of Bifurcation and Chaos, 23 (8) 1330027. 


Smith, L.A. and Petersen, A.C. "Variations on Reliability: Connecting Climate Predictions to Climate Policy" forthcoming in Boumans, M., Hon, G. and Petersen, A.C. (eds.), Error and Uncertainty in Scientific Practice, London: Pickering & Chatto, 2014.    

Stainforth, D.A., Chapman, S.C. and Watkins, N.W.
(2013) "Mapping Climate Change in European temperature distributions" Environmental Research Letters, 8 (034031). Video-abstract. Dave's 'Mapping' paper has been selected for ERL's monthly highlights (over 1400 downloads since published).

Suckling, E.B. and Smith, L.A. (2013) "An evaluation of decadal probability forecasts from state-of-the-art climate models" (accepted, Journal of Climate).
 Each issue we ask one researcher to contribute a short update on their recent research progress.

This issue Henry Wynn discusses the CELSIUS project.

Energy policy is a political hot potato in the UK with debates about the near monopoly position of powerful utility companies, tensions between price, profits and the international green agenda and local resistance to fracking and wind farms.  The energy market is complex with many stakeholders, competing objectives and different business models.  One way of thinking about the issues is as socio-technical systems in which new, and sometimes old, technologies impinge on ordinary people.  It is not clear that existing democratic institutions are able to handle the pace of development.  Real implementation at a local level involves elective democracy under planning laws, including statutory duties to consult, but opposition can develop informally, for example by the use of social media.

CELSIUS (Combined Efficient Large Scale Integrated Urban Systems) is a multi-partner EU project (the largest winning project in the EU Smart Cities & Communities 2011 call).  It is led by the City of Gothenburg and involves a number of leading utilities organizations as well as academic partners.  The project concerns district heating projects: hot water, or steam, piped into buildings from purpose-built generators or excess heat from power stations, waste incinerators, and other industrial units.  The objective is to build up a portfolio of expertise drawing on more than a dozen demonstrators projects in five cities in the EU, including London, with the aim to help roll out many more projects throughout Europe.  It is perhaps not surprising that LSE has been given the
role of developing the "acceptance" work package: to look for social and economic bottlenecks and to gather and develop best practice for gaining acceptance.

Although the technology is not new it presents a range of challenges for local communities.  For example, since district heating is a long-term investment it requires a major commitment.  In cities where there is a strong tradition of public ownership of property, and public investment in heating, acceptance is likely to be easier than where owner-occupiers are part of a more fluid energy market, as in the UK where pipes with hot water coming into blocks of flats have to compete with individual gas boilers in households and, more importantly, electricity from the national grid.  The project leaders are the City of Gothenburg, which has probably the world's strongest record in district heating, 80% coverage as opposed to 3% in London, and a very strong record of public investment.

The interface between technological progress and participatory democracy is seen by CATS as ripe for research and the CELSIUS project is being seen as an excellent platform to build upon, working with other research groups at LSE and internationally.  A welcome to Ofer Engel who has been appointed as the LSE Research Fellow on the project and has just been awarded his PhD in the highly relevant area of social networks.
3-4 December - Leonard Smith has been invited to participate in a limited two-day workshop on assessing uncertainties for extrapolations, taking place in Las Vegas.  The workshop goal is to flesh out principles and approaches for quantifying uncertainty in extrapolative, model-based predictions, and includes a field trip to the Nevada National Security Site.

9-13 December - Leonard Smith and Emma Suckling will be attending the AGU annual Fall Meeting in San Fransisco.  Leonard Smith will give the Charney lecture; his lecture is entitled "Predictability, Probability(s) and Physical Insight."

13-17 February 2014 - Leonard Smith will be speaking at the AAAS Annual Meeting 2014, Chicago.  His talk will be entitled: "Confounding Solid Science and Uncertain Modelling: Improving Climate Policy Modelling."
25 October - Jim Baker gave a presentation on the science of near-term climate extremes at a day-long symposium on "Rising Temperatures and Emerging Threats: the Intersection of Climate Change and National Security in the 21st Century" at the Vermont Law School in South Royalton, Vermont. The proceedings will be published in the Vermont Journal of Environmental Law later this year.

17 October - Keith Beven (Lancaster Environment Centre) gave the European Geophysical Society Leonardo Lecture in Greece.

6-11 October - Leonard Smith gave an invited talk at the workshop "The role of oceans in climate uncertainty" at the Banff International Research Station (BIRS). His talk was entitled "Types of Uncertainty, Kinds of Probability, and (Re)Designing Climate Simulation from Scratch."

4 October - Keith Beven gave a Benchmarks in Hydrology seminar and workshop on uncertainty estimation for environmental models at the University of Saskatoon.

29 September-3 October - Keith Beven gave a keynote address at GeoMontreal.

19-20 September - Leonard Smith gave an invited talk "On the Use and Abuse (and rational interpretation) of Probability Forecasts" at the "Forecasting, Monitoring, Controlling: Dealing with a dynamic world" workshop at UCL. Erica Thompson presented a talk on "The Hawkmoth Effect in forecasting, monitoring and controlling complex dynamic systems." Abstracts.

16-20 September - Nick Watkins spoke at an International Space Science Institute Workshop on "Self-organized criticality and turbulence," Bern, Switzerland.

8-20 September - Ewelina Sienkiewicz presented a poster on "How to quantify the predictability of a chaotic system" at the NCAS Climate Modelling Summer School in Oxford.  She also attended the summer school course which included: background to the theory of climate models, principles of the science behind them, how they operate, understanding of challenges in climate modelling, practical training on how to design, run and evaluate climate models.

10 September - Leonard Smith gave a plenary talk at the NERC PURE Associates workshop in Oxford. The event was attended by a mix of academic/ government/ industry/ NGOs interested in applying to this NERC call.

6 September - Alex Jarman attended a talk on "Statistical prediction of Atlantic basin hurricane activity" given by Phil Klotzbach at the European Centre for Medium-range Weather Forecasts (ECMWF).

2-7 September - Leonard Smith was an invited lecturer at the SUMO Summer School of the Macedonian Academy of Sciences and Arts in Ohrid, Macedonia, http://www.sumoproject.eu. His lecture was entitled "Aims and Means of Supermodelling by Cross-Pollination in Time."

18 July - CATS hosted an 'Uncertainty Workshop' for researchers involved in the Blue-Green Cities project.  This first workshop aimed to investigate the flows of data and information through the model cascade and highlight specific links between the project's work packages, e.g. sharing of data, passing data from one WP to another, using information from research in one WP to inform decisions in another WP.

8-12 July - Leonard Smith gave an invited talk at the Davos Atmosphere and Cryosphere Assembly in Switzerland.  His talk was on "Dynamic probabilities, mature probabilities, and the links between data assimilation and ensemble forecasting in actual decision support."

1-5 July - Nick Watkins spoke at the CliMathNet Conference.  His talk was on "Compound Extremes and Bunched Black (or Grouped Grey) Swans."

1-4 July - Reason Machete gave a talk entitled "Defective Probabilistic Forecasts: Comforting the perplexed" at the 26th European Conference on Operational Research: The Forecasting Stream, Rome.

1-3 July - Pauline Barrieu gave a plenary lecture at the 17th International Congress on Insurance: Mathematics and Economics, in Copenhagen, Denmark.

24-26 June - Pauline Barrieu gave a plenary lecture at AFIR-ERM Colloquium, in Lyon, France.

17-21 June - Leonard Smith gave an invited talk at the workshop "Methods of Chaos Detection and Predictability: Theory and Applications (MCDPTA13)" at the Max Planck Institute for the Physics of Complex Systems, Dresden. His talk was entitled "Predictability, uncertainty and Physical Insight." 

17-20 June - Emma Suckling and Hailiang Du presented at the 2013 SIAM Conference on Mathematical and Computational Issues in the Geosciences in Padova, Italy.  Du's talk was entitled "Pseudo-Orbit Data Assimilation for Atmospheric GCMs." Emma's talk was entitled "Pseudo-Orbit Gradient Descent for Lagrangian Data Assimilation."

 17-18 June - Leonard Smith, David Stainforth, Erica Thompson and Ewelina Sienkiewicz took part in a closed workshop at the Netherlands Environmental Assessment Agency led by Professor Arthur Petersen; they formed part of a panel of earth science experts who reflected on the way uncertainty is characterised in the "Final Government Distribution" of the Summary of Policymakers (SPM) for the Working Group I contribution to the IPCC assessment report.

14 June - Nick Watkins gave a poster at the 3rd annual conference of the UCL Institute for Risk and Disaster Reduction.

21-22 May - Ewelina Sienkiewicz presented a poster on "How to quantify the predictability of a chaotic system" during the Annual PhD Presentation Event in the Statistics Department at LSE.

12-15 May - CATS research away days in Lyme Regis.
Erica Thompson has been awarded NERC PPSDA Funding to run a workshop on "Understanding Uncertainty in Environmental Modelling" aimed at early career researchers. 

CATS submitted three applications to the NERC PURE Associates call and two of these have been successful. The call aims to fund scientists, in projects of between three to six months, to work with business, policy-makers and NGOs to apply their knowledge and skills to improve assessments of natural hazard risk and enhance decision-making.  Our successful projects will involve
i) working with RNLI on "Improving the safety of RNLI operations through better use of probabilistic weather information" (associate Ed Wheatcroft);
ii) working with DECC on "Visualisation of climate model output and uncertainties for the DECC 2050 Global Calculator"  (associate Erica Thompson).

Congratulations to Keith Beven, who has been awarded the 2013 President's Prize from the British Hydrological Society.  The prize is awarded every two years in recognition of work which has advanced hydrology significantly.

Shell Internship:
Process line-up optimization of a large contaminated gas field

The responsible development and monetisation of contaminated (high H2S, CO2) gas fields is high on the agenda of IOCs and governments, as increasingly, reserves are made up of such challenging Hydrocarbon discoveries.
A number of fundamental questions such as:  
  1. The choice of gas processing technologies,
  2. The distribution or centralization of the processing facilities including the pipeline network, remain unresolved.
The processing technologies (e.g. Cansolv, Clauss, SCOT) have different contaminant removal efficacies and associated costs.  Distributed versus centralized optimization involves (though not limited to) the maximum and minimum scale-up/down interval of a technology and related capital cost, the knowledge limits of design and operation etc.
The intern will work in an integrated informal team of Process Engineers, OR specialists, Economists and cost engineers.
The main objective of the study is to formulate as a network optimization problem:
  1. the chemical processes in a fit for purpose fashion (e.g. functional Process Unit descriptions or black box-yield based models) and
  2. the pipeline- network from wellhead to custody transfer point.
Attempts will be made to linearize the problem such that an MILP formulation can be derived. If that approximation is deemed too coarse, the intern is expected to be flexible and interested in looking at MINLP formulations including investigating commercially available solvers.

Please contact Jerome Ellepola  (jerome.ellepola@shell.com) or Nort Thijssen (nort.thijssen@shell.com) for further inquiries.
Closing date: 1 January 2014.
Copyright © 2013 Centre for the Analysis of Time Series, All rights reserved.
You are receiving this email because you are affiliated with CATS or LSE.

Our mailing address is:
Centre for the Analysis of Time Series
Tower 1, 11th Floor
LSE, Houghton Street
London, WC2A 2AE
United Kingdom
Email Marketing Powered by MailChimp