A CCN Water Scarcity Focus Area Workshop Wednesday 10th February 2010
Aims and Objectives (Enda O’Connell – Newcastle University)
- To determine how climate projections can be used in water resource assessments by water companies in terms of what we know currently, what we need to know and what research needs can be identified
- Key issues around incorporating UKCP09 within revision of PR09 business plans (by Nov 2010)
- Lack of spatial correlation and inter-annual variability in UKCP09 weather generator (especially long term characteristics)
- Demonstrated need for a move towards risk-based decision making
Chris Kilsby – Towards probabilistic water resources assessment. (Newcastle University)
- Overview of UKCP09 Weather Generator
- daily and hourly time resolution
- generates rainfall series with other weather variables derived from regression relations
- reproduces extremes but not directly fitted to extremes
- climate model change factors used so implicit bias corrections
- Climate Model and Weather Generator Limitations
- mismatch between RCM models and observations – anomalies in rainfalls and extremes – climate models underestimate frequency of the blocking systems which cause droughts
- seasonal and inter-annual variability under-estimated- problem of “5 season” variability
- currently works for individual sites – for larger scales need to use RCM outputs or Newcastle spatial rainfall generator (FRACAS project )
- See for example using Thames Catchment (Using probabilistic climate change information from a multimodel ensemble for water resources assessment)
- more needs to be done for realistic long term variability
Steven Wade -UKCP09: An update from UKWIR- Impacts of climate change on river flows (HR Wallingford)
- Issue of climate change and headroom to better understand risks and probability of failure – A 6 week rapid assessment for UKWIR
- Wish list includes historical drought data (1901-1902- three driest years since 1815)
- Need to understand and factor in land use change (current assumption of no change)
- Need to understand vulnerability of areas to long-term drought
- Need for consistent methods across planning and water supply
Glenn Watts – UKCP09: Update from the EA (Climate Change Science Manager, EA)
- Key issue is one of using UKCP09 consistently across the EA
- JBA leading on a project to introduce UKCP09 in EA
- Percentage changes in average monthly flows up to 2100
- RCMs in context of 10000 change factors
- How to use in next planning round – does it matter? Droughts, heatwaves, other extremes, surface/groundwater deployable outputs, headroom, outage? Demand?
- Working with CEH on catalogue of historic droughts- improving resilience
Jim Hall -Principles for risk-based decision-making (Newcastle University)
- Overview of well developed principles of risk-based water management and sources of uncertainty – aleatory (random variability – which may not be fully represented in the observed record) and epistemic uncertainties including residual ones
- Environment Agency (2008) Towards a clearer understanding of the relationship between the treatment of uncertainty the water resources planning process and the reliability (or level of security) of the supply-demand balance.
- What is tolerable probability of violating a level of service (including service failures that are supposed to “never” occur)?
- How much is it worth spending to reduce these probabilities? How much are we prepared to pay for reduced risk of failure?
- Need to evaluate robustness to residual sources of uncertainty (known limitations of CP09, probabilities of scenarios, correlations & inter-annual variability…) issues of sensitivity (what are the most important factors) and robustness
- UKCP09 a useful stimulus to promote a probabilistic approach around water resource planning
- Are some management policies more robust than others? How much is it worth paying for robustness?
Alison Browne – – Adaption and resilience in a Changing Climate (Lancaster Environment Centre)
- Overview of a new EPSRC-funded Research Project assessing risks of climate change on water infrastructure and understanding water demand in S and SE England
- Defining what is a drought? Who is it a drought to?
- Scenarios of demand and technology
- Understanding water use under changing circumstances
- Re-conceptualisation of domestic water demand – individual practices – variability around averages – culture as much as price.
- Developing a practice-based model of water demand
Breakout discussion session I
Q: What are the major practical issues associated with implementing UKCP09 projections?
Q: How are these issues currently being addressed?
- UKCP09 – record of extremes now available + observational data readily available
- Metrics for precipitation available (but not daily values)
- Access to data – could be usefully linked to UKCP09 portal
- Not able to provide spatial variability or coherence
- Some Water Companies have purchased daily (1950) monthly (1914) gridded data from the Met office
- project has historical data available for Europe
- Met Office Extremes Group – Simon Brown- drought severity and extreme rainfall work
- Fitness for purpose of these combined datasets?
- Not all companies have models to create pdf of impact (hence flow factors)
- Need to bring in UKCP09 storylines for specific purposes, particularly for adaptation strategies at local levels where no water resource models exist
- Consistency of methodologies needed for Ofwat reviews
- Multiple projects generating data for impact assessments and adaptation- to what extent could these be meshed together?
- Need for guidelines to outline how UKCP09 information can and should be used -inclusion of explicit assumptions and the need to remember learning/practice takes time
- Case study approach may be helpful
Breakout discussion session II
Q: How are various sources of uncertainty dealt with in current water resource assessment methodology?
Q: How can probabilistic climate projections be incorporated into the current water resource assessment framework?
- Headroom – incorporates multiple sources of possible uncertainty? Seems to have been sufficient to date but does it provide the best approach for testing adaptation strategies?
- Difficulty replacing tried and tested working methods across the industry- long take-up times problematic
- Difficulty in identifying how companies deal with uncertainties – different companies using different methods
- Might be beneficial to separate climate change from headroom- to understand what is the sensitivity of the system?
- EA planning guidance needs to be scalable and include probabilistic uncertainty in drought predictions
- Companies risk averse – robust / worse case approaches
- Climate change bill – associated impacts of carbon accounting?
- Barriers to change? – acceptability of greater uncertainty in UKCP09
- Water company focus…..how to improve quantification of uncertainty – what is the best method for the future – some uncertainties very difficult to quantify. Companies looking for a consensus view so working with each other and with Ofwat to move forward
- Need to understand system sensitivities and pinch points and start with sensitivity to climate
- Need for new tools and a more proactive rather than reactive approach
- Incorporating UKCP09 will hopefully unlock demand uncertainty – a key area with a history of neglect
- Importance of barriers – what is the capacity and willingness for people to include climate information as we move towards a more uncertain future? How do we communicate ideas around probability and willingness to pay?
Synthesis – Enda O’Connell (Newcastle University)
- Current ways of using deterministic information is open to failure
- CP09 is a driver for change -and a new approach for PR14
- Need guidance on how to move forward – adaptive measures.
- Approach of using historical data for design events is simplistic and limited: need longer records and better drought frequency analysis (nationally coordinated and using regional frequency analysis analogous to flood risk case)
- How to build CP09 into current methods? – unless it is regulated, WCs won’t do it.
- Methods should be proportionate to level of risk so more detailed analysis needed when investment is large. What is a proportionate amount of analysis?
- Government requirement for CC adaptation across all sectors based on CP09 may force WCs to do it
- How do we begin defining an acceptable level of lack of water?
- Issue of level of investment in R&D – Thames2100 considered investment of £4billion and had R&D budget of £4M. Ofwat’s Notified Items total is £1.4billion so proportionate R&D investment would be £6million.
- Need to communicate to the Water Industry that benefits would result in better decisions- how? In terms of probability and willingness to pay
- Water Innovation Platform for TSB – opportunity to make case for R&D on CC/CP09 as part of £100M case to TSB for Water Innovation Strategy
- Need to pilot new approaches on longer time scale than AMP cycle. Need to make a case to Ofwat – demonstrate what is fit-for-purpose approach through pilots. Analogy with other industries.
- Use of CC snapshots no longer valid – dynamic information should be used – integrated with information from other sources eg historic information etc. Need framework to do this/incorporate into decision-making
- EPSRC has invested in research on capacity building/adaptation. Barriers in WI to introducing new information: long-term problem – need to look beyond short-term.
- Need case studies to demonstrate benefits (eg as in flooding)
- Epistemic uncertainties (e.g. from demand or emission scenarios) could be treated separately/differently from probabilistic CC information: perhaps allows a compromise approach retaining headroom method, splitting out CC impacts on resource alone to be treated with probabilistic method
- No possibility exists in the current framework for inter-regional analysis and transfers to be considered
