Lorentz Principles

The Lorentz Principles for the Communication of Model-based Information 

Leonard A Smith, Arthur C Petersen and Erica L Thompson

#LorentzPrinciples                               Statement of Principles

Aims of the Lorentz Principles

How should the reliability of scientific findings be assessed and communicated to policy-makers? Different organizations, ranging from the IPCC to the CIA, have developed Uncertainty Guidances to assist experts in communicating uncertainties more clearly. Within many domains, however, such Uncertainty Guidance does not yet inform many numerate practitioners. This has restricted the development of a deeper understanding and characterization of what today’s science says, and hampered the communication of uncertainty in model-based findings to decision-makers in both the private and public sectors. The Lorentz Principles have been developed to meet that need, and the aim of this website is to facilitate further discussion, criticism, and refinement of the Lorentz Principles and their application.

The Lorentz Principles 

Principle 1
Good decision-making is best supported by scientific information deemed adequate for purpose. Where today’s best available information is not adequate, this must be made obvious.
Read more on Principle 1           Comment on Principle 1

Principle 2:
Good decision-making is enhanced when both our ignorance of the future and the limits of today’s science are communicated clearly.
Read more on Principle 2           Comment on Principle 2

Principle 3:
Scientific information to inform decisions regarding the future must always be accompanied by a quantitative statement regarding its expected robustness and potential irrelevance.
Read more on Principle 3           Comment on Principle 3

Principle 4:
Information and insight regarding (a) the behaviours of computational models, (b) the properties of theoretical mathematical constructs and (c) observations of the world itself, must always be distinguished clearly, especially when these three distinct entities share the same name.
Read more on Principle 4           Comment on Principle 4

Principle 5:
“Traceable accounts of uncertainty” must be provided, covering all known significant sources of uncertainty including, but not limited to, those of simulation (imprecision, ambiguity, model inadequacy…) and those identified via expert judgement.
Read more on Principle 5           Comment on Principle 5

Principle 6:
Uncertainty Guidance varies with the context, origins and consumer of the information. Effective Uncertainty Guidance is tailored to the aims, understanding, and risk-appetite of the consumer.
Read more on Principle 6           Comment on Principle 6

Principle 7:
Communication of scientific information for decision support outside the scientific community is more effective when professional means of communication are employed.
Read more on Principle 7          Comment on Principle 7

Principle 8:
Provision of over-precise analytic or model-based “information” (oversell) damages the credibility of all science, and can result in very poor (over-confident) adaptation decisions by practitioners. 
Read more on Principle 8           Comment on Principle 8

Principle 9:
Future-tuned simulations designed intentionally to illustrate some selected model property must be clearly distinguished from forecasts and projections (predictions) which, while conditioned on future forcing, are tuned using only the past.
Read more on Principle 9           Comment on Principle 9

Principle 10:
Basic “good practice” for extrapolation tasks (science in the dark) differs from that of more straightforward science in the light tasks where the system is thought stable and a large archive of forecast-outcome pairs are available. Nevertheless, violations of good practice remain well defined and, if allowed, the impacts of knowingly bad-practice elements of an analysis must be clearly identified along with their implications for the relevance of those results for decision support. 
Read more on Principle 10           Comment on Principle 10 

Feeding back

We welcome constructive comments on the draft text of each Lorentz Principle and look forward to a discussion about the Principles, their application in practice, and their consequences. Please use the comments box to leave your thoughts. Leave a comment.

Or find us on social media for a chat #LorentzPrinciples: 
Leonard Smith: @lynyrdsmyth 
Arthur Petersen: @ArthurCPetersen
Erica Thompson: @h4wkm0th

Where did the Lorentz Principles come from?

A Lorentz Center WorkshopUncertainty Guidances in Science and Public Policy” was convened in Leiden, the Netherlands, from 13-17 November 2017, with the aim to review existing Uncertainty Guidances and to propose ways forward for improved communication of uncertain climate information. The workshop brought together 20 natural scientists, social scientists and philosophers – as well as practitioners who use scientific information to tackle real-world problems.The workshop developed draft principles for the responsible use, provision and design of scientific information on climate change for policy use and decision-making. Basically, the principles are to provide the background for discussing “good practice”, and bad, for questions of nontrivial extrapolation. These draft principles were subsequently refined and reviewed, resulting in ten Lorentz Principles for the Communication of Climate Information on 5 April 2019. We welcome your further input, discussion and refinement.

Where can I read more?

A list of relevant further reading is on our page of references. If you would like to suggest additions to this list, please leave a comment.

The Lorentz Principles are supported by: 

Leonard Smith, LSE
Arthur Petersen, UCL
Erica Thompson, LSE
Luke Bevan, UCL
Seamus Bradley, University of Leeds
Mandeep Dhami, Middlesex University London
Nigel Harvey, UCL
Mike Hulme, University of Cambridge
Andrew Kruczkiewicz, IRI, Columbia University
Dewi le Bars, Royal Netherlands Meteorological Institute (KNMI)
Reason Machete, Botswana Institute for Technology Research and Innovation
Wilfran Moufouma-Okia, WMO
Magda Osman, QMUL