Lancaster Environment Centre

NERC in collaboration with Defra have commissioned Lancaster Environment Centre to undertake a scoping study on the subject of Reducing Uncertainty in Models for Environmental Decision-Making. The purpose of the study is to identify priority research areas with particular relevance to NERC’s Environment, Pollution and Human Health Theme, requiring investment that will take this agenda forward.

LEC, in coordination with Know Innovation Ltd will host a set of four workshops to address the topic. Each of the workshops will address a different theme of uncertainty in environmental models, including atmospheric processes, terrestrial (including hydrological) processes, chemicals in the environment and methodological approaches to modelling uncertainty. Each workshop will address a number of key questions/challenges:

1. What are the most significant sources of uncertainty in the environmental modelling area being considered?
2. Which of these are of most importance to decision makers?
3. Which of these can be most reduced?
4. What research programmes are already addressing this task?
5. Where and how should further research be focussed? Is the reduction of uncertainty necessarily always the most appropriate action?

Further information on the remit of the study can be found in the original Invitation to tender document.

Workshop 1 - Atmosphere

Convenor: Dr Rob MacKenzie
Date: Monday 14th June 2010
Location: Lancaster Environment Centre (Room LG505)

The relevant LWEC objective, "to protect human, plant and animal health from diseases, pests and environmental hazards", frames our perspective. Atmospheric hazards can be split into five broad classes: (i) extreme weather (particularly heat-wave, drought, and storm), (ii) changes to atmospheric opacity (particularly UV transmission to the surface as a result of stratospheric ozone depletion), (iii) secondary pollutants and long-range transport (e.g. ozone), (iv) acute, near-source pollution (especially particulate matter and nitrogen dioxide in urban areas, and fumigation by accidental or controlled release from industrial plants), and (v) airborne transmission of disease and pests. For all these classes of atmospheric hazard, the models needed to aid decision-making share many core requirements - a treatment of the atmospheric flow, interaction with the underlying surface, etc. - as well as each having specialised components (e.g. atmospheric chemistry, virus biochemistry). This scoping study will identify to what extent uncertainty in atmospheric hazards are dominated by common underlying components of the problem - the flow and/or land/atmosphere interaction - as opposed to specifics for each hazard.

The atmospheric workshop called to address the scoping study will seek to have invited speakers who are active in research across the five classes of hazard described above, and speakers from the policy-making and regulator communities. For each hazard class we will, where appropriate, seek speakers who take different approaches to the issue. So, for instance, to discuss uncertainties in urban air pollution modelling, we would look to invite developers of mesoscale fluid dynamics modelling (e.g. from the UK WRF users group) as well as developers of the Gaussian plume dispersion approach (e.g. from the CERC consultancy, Cambridge). Rather than ask speakers to address the broad issue of reducing uncertainty for environmental decision-making from the outset, we will set speakers the task of describing a recent, policy-relevant piece of work, and to describe the uncertainties that limit the results of that piece of work. After hearing these specific narratives of uncertainty the facilitators will engage the audience to try and draw out the most important general lessons.

To keep the scope broad but the size of the workshop manageable, we will seek participation of key personnel with experience across several kinds of hazard modelling. We will make use of ongoing activities noted in the call and additionally the AirTrack and Opeanair KE projects. AirTrack is led by LEC. Within LEC we have expertise in many of the relevant issues connected to this workshop theme, such as biogenic VOC emissions modelling (Hewitt), urban/regional ozone and secondary PM modelling (MacKenzie, Whyatt, Hewitt), global air quality modelling (Wild), pollutant dispersion modelling (Whyatt, Sweetman), upper-troposphere-lower stratosphere modelling (MacKenzie) and climate/uncertainty modelling (Jarvis).

Workshop 2 - Land & Water

Convenors: Prof. Phil Haygarth & Dr Trevor Page
Date: Tuesday 29th June 2010
Location: Lancaster Environment Centre (Room LG505)

Workshop 2 will focus on reducing uncertainty in land water models and can use the land water continuum source-mobilisation-delivery structure (Haygarth et al., 2005) (similar to the source-pathway-receptor structure specified in the bid) as a means of focussing discussions. One recent argument made in the cost curve modelling (Haygarth et al., 2009) is that uncertainties in land water models can be amplified along the continuum, with greatest challenges towards the delivery (Beven et al., 2005) end of the continuum, with least uncertainty around where sources of pollutants lie. There may also be a need to explore linkages and weaknesses in available data and consider to what extent field variability in data contribute to poor understandings in models (Page et al., 2005). Many of the relevant models in land water science have been nutrient orientated and these will be studied and gaps in process knowledge will be explored. However, it is speculated that the largest knowledge gap is in relation to potential pathogen transfers, such as E. Coli or Cryptosporidium (Oliver et al., 2005a; Oliver et al., 2005b; Oliver et al., 2007). Within LEC we have expertise in all relevant issues connected to this workshop theme, such as nutrients (Haygarth, Heathwaite, Quinton), pathogens (Oliver, Haygarth, Heathwaite) and modelling (Page, Beven), amongst others. A list of potential models that may be evaluated in the workshop are PSYCHIC, PIT/Indicators, PEDAL 1&2, INCA, PSALM, PEASE (Cost Curve), Export Coefficient, MACRO, NCYCLE, Diffuse Pollution User Manual, but there are many other we may consider and as a start to the process we shall work with the community to produce a more fulsome list of the models that are currently available for the policy makers, and should be evaluated. Haygarth and the team have a strong track record of working with Defra ‘model user’ policy staff and has contacts to involve, but we shall work with Defra and NERC to select best policy uses that wish to be represented in the discussions. Workshops can focus on the environmental processes within the models and where future investments should be made, we can attempt to provide an inventory assessment on capacity and capability UK wide and determine where and to what scale future investments could be made.

Workshop 3 - Chemicals

Convenors: Dr Andy Sweetman & Prof. Kevin Jones
Date: Monday 5th July 2010
Location: Lancaster Environment Centre (Gordon Manley Building Training Room 1)

The protection of humans and the environment from exposure to potentially harmful chemical substances has received considerable political attention over the last 40 years. This has resulted in many pieces of chemicals regulation, the most recent of which in the EU is REACH (Registration, Evaluation and Authorisation of Chemicals) which came into force in 2007. The aim of REACH, and similar regulations, is to improve the protection of human health and the environment through the better and earlier identification of the intrinsic properties of chemical substances through a transparent and defendable chemical risk assessment process. At the heart of the risk assessment process, as with many pieces of national and international regulation, are mathematical models designed to calculate the risk from exposure to chemical substances based on a knowledge of their intrinsic properties, their predicted fate and behaviour in the environment and their potential toxicity to humans and wildlife. The types of models used in the risk assessment process are many and varied, and can be used both as research tools and to inform policy makers as to the risks involved with the manufacture and use of chemicals. The model structures are also varied in nature and complexity ranging from those that predict environmental concentrations (sharing many of the components of atmospheric and land-water pollution transport models discussed above) and which proceed from a presumed-known release of chemical into a generic environment, to those that calculate risk by predicting exposure from occupational or consumer activities or from dietary exposure.

Chemical risk assessment generally involves a number of steps; hazard identification, exposure assessment and risk characterisation. Models can either be used to provide data for individual parts of the risk assessment process (e.g. hazard property QSARs ), or to carry out the complete risk assessment process (e.g. EUSES ) from source to receptor. By necessity such models vary greatly in their complexity, but broadly they range from generic models that provide a transparent assessment of chemical fate and/or exposure, based on a limited amount of input data in a generic environment, to site specific models that provide a high degree of specificity but require considerably more input data. The former generally use a conservative ‘worst case’ approach with generous safety factors but can over estimate risk, while the latter often require a high level of parameterisation for which data are often not available.

Understanding variability and uncertainty are key elements in the advancement of such models in the development and implementation in chemicals regulation policy. Generally the main areas of uncertainty are (1) scenario uncertainty (2) model uncertainty (3) parameter variability (4) parameter uncertainty. It is the aim of this workshop to discuss each of these areas, and with the use of facilitators provide a consensus of opinion as to where the areas of greatest uncertainty lie and how they maybe reduced. Further, and importantly, the workshop participants will also be asked to discuss how current risk assessment and exposure models, which cover a disparate range of complexities and spatial scales, could be successfully integrated into a framework to ensure there is a coherent and consistent approach to the risk assessment process and any subsequent policy developments in the future.

Workshop 4 - Synthesis: Modelling uncertainty for decision making

Convenor: Prof. Keith Beven
Date: Wednesday 7th July
Location: Defra Innovation Centre, Reading

There is a fascinating tension between formal risk-based decision making methodologies and real practice in environmental decision making. This tension arises because of the probabilistic foundations of (most) risk-based decision and the epistemic nature of many of the sources of uncertainty that arise in making predictions of environmental systems. Statisticians have a range of well developed techniques for treating uncertainties as if they were aleatory, and thus amenable to quantification and management through methods of probabilistic risk assessment. Increasingly, environmental modellers now recognize that many of the uncertainties they deal with are epistemological, stemming from imperfect model structure, data incompleteness, or both, so that model residuals may have complex structure and the real information content of calibration data significantly less than that assumed under aleatory approaches. In such circumstances, expert judgment, often backed by Bayesian methods for dealing with subjective probability estimates, comes to the fore. Insofar as such judgments are subjective and necessarily contestable, social scientists have emphasized the importance of closer engagement between scientists, stakeholders and the public in understanding and managing risk and uncertainty. Engagement is important both to ensure effective knowledge exchange and because competing problem framings and values can give rise to ‘ontological’ or “deep” uncertainties relevant to decision and policy making. Such uncertainties mean that stakeholder consultation and negotiation become crucial in managing them in an adaptive way.

Treating uncertainties as if they are simply aleatory allows the full power of probabilistic methods to be invoked (as noted in the recent NERC review of uncertainty in natural hazards) but can lead to overconfidence in model predictions and solutions (see, for example, Beven, 2006; Beven et al., 2008, Beven, 2009). Less formal methods (such as Info-Gap decision making, Ben-Haim, 2006) also have their limitations and do not necessarily address the epistemic uncertainty issue directly. A critical issue therefore in modelling and reducing uncertainty for decision making is how to progress beyond treating epistemic uncertainties as probabilistic at one extreme or simple scenarios at the other, within an adaptive learning and decision making process including stakeholder consultation and negotiation (Beven, 2007; Faulkner et al., 2007). In developing a framework for uncertainty estimation for flood risk mapping under FRMRC2, for example, epistemic uncertainties arise in the choice of flood frequency distribution, the stationarity of flood frequency statistics, the choice of effective roughness coefficients, the representation of flood plain topography and infrastructure, the choice of model numerics and implementation, and the representation of potential damages. This workshop will be designed to address this issue, with participants invited to provide examples of both formal and real world decision making as a means of introduction to a discussion of how to model and reduce the effects of epistemic uncertainties in defining a future research programme.

References

• Ben-Haim, E, 2006, Info-Gap Theory: Decisions Under Severe Uncertainty, 2nd edition, Academic Press: London
• Beven, K J, 2006, A manifesto for the equifinality thesis, J. Hydrology, 320, 18-36.
• Beven, K J, 2007, Working towards integrated environmental models of everywhere: uncertainty, data, and modelling as a learning process. Hydrology and Earth System Science, 11(1), 460-467.
• Beven, K J, 2009, Environmental Modelling – An Uncertain Future? Routledge: London
• Beven, K., A.L. Heathwaite, P.M. Haygarth, D.E. Walling, R. Brazier, and P.J.A. Withers. 2005. On the concept of delivery of sediment and nutrients to stream channels. Hydrological Processes 19:551-556.
• Beven, K J, Smith, P J, and Freer, J, 2008, So just why would a modeller choose to be incoherent?. J. Hydrology, 354,15-32.
• Faulkner, H, Parker, D, Green, C, & Beven, KJ, 2007, Developing a translational discourse to communicate uncertainty in flood risk between science and the practitioner, Ambio, 16:692-703.
• Haygarth, P.M., L.M. Condron, A.L. Heathwaite, B.L. Turner, and G.P. Harris. 2005. The phosphorus transfer continuum: Linking source to impact with an interdisciplinary and multi-scaled approach. Science of the Total Environment 344:5-14.
• Haygarth, P.M., H. ApSimon, M. Betson, D. Harris, R. Hodgkinson, and P.J.A. Withers. 2009. Mitigating diffuse phosphorus transfer from agriculture according to cost and efficiency. Journal of Environmental Quality.
• Hewitt, C.N., A.R. MacKenzie, P. Di Carlo, J.R. Dorsey, M. Evans, D. Fowler, M.W. Gallagher, C. Helfter, J. Hopkins, H. Jones, B. Langford, J.D. Lee, A.C. Lewis, S.F. Lim, C. di Marco, P. Misztal, S. Moller, P.S. Monks, E. Nemitz, D.E. Oram, S.M. Owen, G. Phillips, T. Pugh, J.A. Pyle, C.E. Reeves, J. Ryder, J. Siong, U. Skiba, D.J. Stewart, R. Thomas, 2009, Nitrogen management is essential to prevent tropical oil palm plantations from causing ozone pollution, Proc. Natl. Acad. Sci., doi: 10.1073/pnas.0907541106.
• Horseman, A.M. , A. R. MacKenzie, and M. P. Chipperfield, 2009, Tracers and traceability: implementing cirrus in a chemistry-transport model as an example of the application of quality assurance to legacy models, Geosci. Model Dev. Discuss., 2, 1299-1333.
• Lowe, D., A. R. MacKenzie, N. Nikiforakis, and J. Kettleborough, 2003, A condensed-mass advection based model of liquid polar stratospheric clouds, Atmos. Chem. Phys., 3, 29-38.
• Lowe, D., and A. R. MacKenzie, Polar Stratospheric Cloud Microphysics and Chemistry, 2007, J. Atmos. Solar-Terrestr. Phys., doi:10.1016/j.jastp.2007.09.011.
• MacKenzie, A. R., R. M. Harrison, I. Colbeck, and C. N. Hewitt, 1991, The role of biogenic hydrocarbons in the production of ozone in urban plumes in southeast England, Atmos. Environ., 25A, 351-359.
• Oliver, D.M., C.D. Clegg, P.M. Haygarth, and A.L. Heathwaite. 2005a. Assessing the potential for pathogen transfer from grassland soils to surface waters, p. 125-180 Advances In Agronomy, Vol. 85.
• Oliver, D.M., L. Heathwaite, P.M. Haygarth, and C.D. Clegg. 2005b. Transfer of Escherichia coli to water from drained and undrained grassland after grazing. Journal of Environmental Quality 34:918-925.
• Oliver, D.M., C.D. Clegg, A.L. Heathwaite, and P.M. Haygarth. 2007. Preferential attachment of Escherichia coli to different particle size fractions of an agricultural grassland soil. Water Air and Soil Pollution 185:369-375.
• Page, T., P.M. Haygarth, K. Beven, A. Joynes, P. Butler, C. Keeler, J. Freer, P. Owens, and G. Woods. 2005. Spatial variability of soil phosphorus in relation to the topographic Index and critical cource areas: sampling for assessing risk to water quality. Journal of Environmental Quality 34:2263-2277.
• Pugh, TAM, AR MacKenzie, CN Hewitt, B Langford, PM Edwards, KL Furneaux, DE Heard, JR Hopkins, CE Jones, A Karunaharan, J Lee, G Mills, P Misztal, S Moller, PS Monks, and LK Whalley, 2009, Simulating atmospheric composition over a South-East Asian tropical rainforest: Performance of a chemistry box model, Atmos. Chem. Phys. Discuss., 9, 19243-19278.
• Ren, C., A. R. MacKenzie, C. Schiller, G. Shur, and V. Yushkov, 2007, Diagnosis of processes controlling water vapour in the tropical tropopause layer by a Lagrangian cirrus model, Atmos. Chem. Phys, 7, 5401-5413.

Contact

For further information please contact the project coordinator Dr Paul McKenna.

Telephone: +44 (0)1524 510301
Email: p.mckenna@lancaster.ac.uk