Download Development of a metamodel tool for regional integrated climate

General enquiries on this form should be made to:
Defra, Science Directorate, Management Support and Finance Team,
Telephone No. 020 7238 1612
E-mail:
[email protected]
SID 5

Research Project Final Report
Note
In line with the Freedom of Information
Act 2000, Defra aims to place the results
of its completed research projects in the
public domain wherever possible. The
SID 5 (Research Project Final Report) is
designed to capture the information on
the results and outputs of Defra-funded
research in a format that is easily
publishable through the Defra website. A
SID 5 must be completed for all projects.
A SID 5A form must be completed where
a project is paid on a monthly basis or
against quarterly invoices. No SID 5A is
required where payments are made at
milestone points. When a SID 5A is
required, no SID 5 form will be accepted
without the accompanying SID 5A.

This form is in Word format and the
boxes may be expanded or reduced, as
appropriate.

ACCESS TO INFORMATION
The information collected on this form will
be stored electronically and may be sent
to any part of Defra, or to individual
researchers or organisations outside
Defra for the purposes of reviewing the
project. Defra may also disclose the
information to any outside organisation
acting as an agent authorised by Defra to
process final research reports on its
behalf. Defra intends to publish this form
on its website, unless there are strong
reasons not to, which fully comply with
exemptions under the Environmental
Information Regulations or the Freedom
of Information Act 2000.
Defra may be required to release
information, including personal data and
commercial information, on request under
the Environmental Information
Regulations or the Freedom of
Information Act 2000. However, Defra will
not permit any unwarranted breach of
confidentiality or act in contravention of
its obligations under the Data Protection
Act 1998. Defra or its appointed agents
may use the name, address or other
details on your form to contact you in
connection with occasional customer
research aimed at improving the
processes through which Defra works
with its contractors.
SID 5 (2/05)
Project identification
1.
Defra Project code
2.
Project title
CC0362
Development of a metamodel tool for regional integrated
climate change management
3.
Contractor
organisation(s)
Institute of Water and Environment,
Cranfield University
54. Total Defra project costs
5. Project:
Page 1 of 23
£
316,416.36
start date ................
01 September 2003
end date .................
31 August 2005
6. It is Defra’s intention to publish this form.
Please confirm your agreement to do so. ................................................................................... YES
NO
(a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They
should be written in a clear and concise manner and represent a full account of the research project
which someone not closely associated with the project can follow.
Defra recognises that in a small minority of cases there may be information, such as intellectual property
or commercially confidential data, used in or generated by the research project, which should not be
disclosed. In these cases, such information should be detailed in a separate annex (not to be published)
so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report
without including references to any sensitive or confidential data, the information should be included and
section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No"
answer.
In all cases, reasons for withholding information must be fully in line with exemptions under the
Environmental Information Regulations or the Freedom of Information Act 2000.
(b) If you have answered NO, please explain why the Final report should not be released into public domain
Executive Summary
7.
The executive summary must not exceed 2 sides in total of A4 and should be understandable to the
intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together
with any other significant events and options for new work.
ABSTRACT
There is widespread acceptance in the UK that the climate is changing due to human emissions of
greenhouse gases. Such changes in climate will affect all sectors of society and the environment at all
scales, ranging from the national to the local. It is also recognised that, whilst the climate changes, not
everything else stands still, so that the socio-economic and political changes which influence the impact of
climate change also need to be considered in a full assessment of the effects of our changing climate.
The ‘Regional Climate Change Impact and Response Studies in East Anglia and North West England’
(RegIS) study developed and applied a methodology for stakeholder-led, regional Integrated Assessment
(IA). Key needs for future stakeholder-led Integrated Assessments identified at a UKCIP-hosted
stakeholder workshop at the end of RegIS were:
 Results from far greater numbers of scenario (climate and socio-economic) combinations;
 Better understanding of the effects of scenario uncertainty on the impacts;
 sensitivity analyses of the integrated models to be able to better understand the effects of multiple
interactions on impacts.
 Greater degree of model linkage and interaction and more feedbacks and adaptive responses.
However, most existing models for the sectors represented have unacceptably long runtimes for allowing
rapid simulation and interactive engagement with the IA. It was from this that the two-year ‘Development
of a metamodel tool for regional integrated climate change management’ (RegIS2) project was born.
Its aim being to investigate and develop the use of computationally simpler modelling techniques, so
called ‘metamodels’ or ‘reduced form models’, within a user friendly interface (the ‘Regional Impact
Simulator’) to make the RegIS methodology available to the wider stakeholder community.
The RegIS2 project has further built upon the ‘Drivers-Pressure-State-Impact-Response’ (DPSIR)
framework which was used successfully in RegIS. By incorporating the widely used DPSIR framework
within the RegIS2 Integrated Assessment Methodology, the methodology is transportable to other regions
of the UK, thereby providing a framework for regional versions of the Regional Impact Simulator or indeed
a national version. In addition, the two contrasting regions (East Anglia and North West England) provide
good ‘test-beds’ for the diversity of the problems and approaches needed for a tool to support regional
Integrated Assessment in England and the remainder of Great Britain.
SID 5 (2/05)
Page 2 of 23
The main output from the RegIS2 project is the Regional Impact Simulator, a user friendly tool which
allows the impacts and interactions of future climate or socio-economic change to be assessed by
stakeholders across the agricultural, hydrological, biodiversity and coastal sectors. The Regional Impact
Simulator contains the:
 Metamodels – a range of calibrated and validated metamodels simulating fluvial and coastal
flooding; rural landuse change and agricultural production; water resources, supply and demand;
habitat change and species suitability;
 Scenarios- baseline scenarios plus the UKCIP02 climate scenarios for 2020s Low, 2020s High,
2050s Low and 2050s High and the four RegIS regional socio-economic scenarios (Global
Sustainability, Global Markets, Regional Stewardship and Regional Enterprise), and
 Datasets necessary to run the metamodels.
The philosophy behind the Regional Impact Simulator was to make the interface as intuitive as possible,
such that an interested individual could use it with minimal recourse to help files and without need for
training. The design of the interface was therefore crucial to the success of the project, and the Regional
Impact Simulator has been through multiple iterations with stakeholders. The final version contains four
main screens of differing purpose, and increasing levels of interaction with the model variables:
 The first, About screen introduces the software (purpose of the screens, project team etc.);
 The second, Predefined Scenario Futures, screen allows the user to rapidly identify relative
sensitivity of an Indicator(s) to the different climate and/or socio-economic scenarios. The
Regional Impact Simulator can be run after only clicking on a climate scenario, a region and
(optionally) a socio-economic scenario;
 The third, Exploratory Analysis, screen lets the user explore a scenario in greater detail by
changing the scenario pressure variables which act as input parameter values in the models. This
screen can be used to carry out sensitivity analyses under the baseline conditions or guided
uncertainty analyses of futures;
 The fourth, Influencing the Impacts, screen allows the user to test regional adaptation responses
to the impacts previously identified.
The Regional Impact Simulator was built using a modular Object Oriented approach, in which the modules
(or meta-models) were written in various programming languages and compiled in libraries of executable
functions (DLLs) or executables (EXEs). The modular approach to model integration has the advantage of
allowing independent model development, so that the models can be developed and tested as stand-alone
versions before being integrated with the other models. The meta-models are run as serial processes
(sequentially in the order: High River Flow, Flood, Agriculture, Water Resources, Biodiversity) with the
inclusion of parallel processes where possible. The interface development played an important role in
minimising the run-times of the coupled meta-models. On a typical new PC (2.8 GHz processor with 1Gb
RAM), the Regional Impact Simulator will complete a Predefined Scenario Futures integrated assessment
for East Anglia in less than 15 seconds for one habitat, or 28 seconds for all habitats (though speed of
computer has a large impact).
Stakeholder evaluation of the Regional Impact Simulator suggests that it has attained an acceptable level
of intuitiveness and user-friendliness in design and run-time. The rapid run-time enables impacts to be
studied in far more detail than in a normal impact study using complex models, though this ironically leads
to more questions as to why a response happens. This should however be viewed in a positive light as,
where normally the answers were simply a black box to be taken on trust by Stakeholders, it is now
possible for a user to study what is happening in response to changes in some detail.
It is important that the fast run-times do not affect the perceived credibility of the Regional Impact
Simulator. The multiple facets that determine the credibility of an IAM in the eyes of stakeholders dictate
that it has been approached from a variety of perspectives:
 Basing the metamodels on widely accepted pre-existing models;
 The slider bars with which the User changes the Pressure or Responses show the numerical
values, so that the magnitude of changes being made are clear;
 The fast runtimes enable Users to carry out many simulations to rapidly investigate the sensitivity
of system under baseline conditions, and to compare this, where appropriate, with their own
knowledge;
 In the Influencing the Future screen, the user is automatically guided to the Responses that that
affect the chosen Indicator. It is therefore transparent to the user which choices will have greatest
effect;
 A hierarchy of User support - tooltips (for immediate provision of a short definition or explanation);
concise higher level Help files which provide basic background understanding for the vast
proportion of Users, and full documentation within the Help files for technical users;
 Explicitly communicating that the very model inputs themselves which derive from the scenarios
SID 5 (2/05)
Page 3 of 23
are uncertain. The colour coding approach used in Regional Impact Simulator, and the actual
breadth of the coloured uncertainty bands on the slider bars conveys clearly that different values
(and the assumptions behind them) are likely to give rise to differing outcomes which the User can
rapidly explore due to the rapid run times
Integrated, regional assessment represents a powerful tool in environmental change impact and
adaptation studies that has many advantages over ‘traditional’ sectoral approaches. IA models such as
RegIS allow the relationships between sectors to be evaluated and thus the consequences of changes in
one sector to be assessed with respect to another sector. Single sector studies are clearly limited by their
inability to treat all of the processes that may be important in affecting a sector. It becomes clear when
undertaking assessments in this way that the various regional-scale sectors are interdependent with
changes and potential adaptation options for one sector sometimes having profound effects on other
sectors.
During the evaluation process, it was asked whether it would be useful to develop the Regional Impact
Simulator further. There were no negative responses, with most respondents being highly positive. The
evaluation process showed that there is demand for such assessment tools as the Regional Impact
Simulator, but that there are always ways of improving on their development, not least with respect to the
interface design and functional capacity. A number of developments would enhance the capacity and
utility of the system:
a) application to new regions and/or coupling regional and national scale applications- this was
suggested by most respondents in the evaluation process,
b) refinements to the user interface,
c) refinement to existing models,
d) addition of new meta-models (e.g. urban development, forestry),
e) new assessment concepts (risk, valuation, visualisation).
The Regional Impact Simulator addresses both autonomous and societal adaptation strategies. However,
whilst autonomous adaptation is dealt with explicitly by the modelling approaches, the Regional Impact
Simulator does not indicate the policy mechanisms that would be needed to implement such societal
adaptation strategies in practice. The need remains for stakeholders to make decisions themselves about
what is or is not feasible, achievable and/or desirable in terms of adaptation responses, as they are best
placed to do this, informed by using the Regional Impact Simulator to explore the effect of acting on the
pressure variables through regional responses. In such a way, the Regional Impact Simulator should
promote better informed decision makers and decision-making.
Project Report to Defra
8.
As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with
details of the outputs of the research project for internal purposes; to meet the terms of the contract; and
to allow Defra to publish details of the outputs to meet Environmental Information Regulation or
Freedom of Information obligations. This short report to Defra does not preclude contractors from also
seeking to publish a full, formal scientific report/paper in an appropriate scientific or other
journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms.
The report to Defra should include:
 the scientific objectives as set out in the contract;
 the extent to which the objectives set out in the contract have been met;
 details of methods used and the results obtained, including statistical analysis (if appropriate);
 a discussion of the results and their reliability;
 the main implications of the findings;
 possible future work; and
 any action resulting from the research (e.g. IP, Knowledge Transfer).
SID 5 (2/05)
Page 4 of 23
Introduction
The ‘Regional Climate Change Impact and Response Studies in East Anglia and North West England’ (RegIS)
project (CC0337) (Holman et al., 2001; Holman and Loveland 2001) represented the first attempt at quantitatively
modelling the cross-sectoral impacts of climate change within an integrated framework at a regional scale within
the UK. The RegIS project demonstrated that an evaluation of the regional impacts of climate change, through an
integrated methodology, on the agriculture, hydrology, biodiversity, and coastal sectors is possible, but not
straightforward. Although the ‘Drivers-Pressure-State-Impacts-Responses’ modelling framework used was
successful, the project was not able to fully integrate the cross-sectoral climate change impacts because of runtime and data problems with some of the detailed mechanistic research models used. The complexity of these
models hindered the wider applicability of the RegIS methodology. Although the development of the methodology
was stakeholder-led, the application of the methodology was not.
At the end of the project, the UK Climate Impacts Programme (UKCIP) hosted a technical workshop for
stakeholders to discuss the project results and to reflect on the lessons learnt. Key issues for future stakeholderled Integrated Assessments came out of this workshop, as stakeholders wanted:
 Results from far greater numbers of scenario (climate and socio-economic) combinations - RegIS
simulated the effects of two climate-only and two climate-socio-economic scenario futures for the 2050s
timeslice only;
 Better understanding of the effects of uncertainty in the scenarios of future climate (e.g. Hulme et al.
2002) and socio-economic conditions on the impacts;
 The ability to conduct sensitivity analysis to be able to better understand the effects of multiple
interactions on impacts.
 Greater degree of model linkage and interaction and more feedbacks and adaptive repsonses.
This current RegIS2 project proposed to investigate and develop the use of computationally simpler modelling
techniques, so called ‘meta models’ within a user friendly interface (the ‘Regional Impact Simulator’) to make the
RegIS methodology available to the wider stakeholder community.
Objectives
The objectives of the ‘Development of a metamodel tool for regional integrated climate change
management’ (RegIS2) project were to:
1. To refine the RegIS risk assessment methodology for conducting climate change impact assessments on
agriculture, hydrology, biodiversity and coasts into a form usable by Stakeholders, that explicitly addresses local
and regional scale impacts, interactions between sectors, the use and linkage of metamodels and stakeholder
concerns and responses.
2. To develop regional scenarios of the pressure of climate change, and state-impact characteristics which
influence social and economic sensitivity to climatic change.
3. To identify, with stakeholders in the North West and East Anglian regions, critical impacts, interactions and
adaptive responses, and suitable interface design.
4. To develop a robust and user friendly software tool, containing an encrypted version of the RegIS
geographically-referenced database for the North West and East Anglian regions of relevant environmental data
and scenarios of climate and socio-economic change, the metamodels and the input data needed for
Stakeholders to analyse climate change impacts, their interactions and adaptive responses using the
metamodels.
5. To develop, calibrate and validate metamodels based upon simulations using existing impact models for the
North West and East Anglian regions for four sectors: agriculture, water, biodiversity, and coastal zones.
6. To investigate the effects of climate change on critical impacts and interactions using the linked metamodels
and scientific expertise. Specifically, to identify significant driving forces, impacts and interactions, analyse the
relative sensitivity of impacts and interactions, and evaluate changes in impacts and interactions over short,
medium and long-term time scales.
7. To investigate adaptive responses to climate change using linked metamodels and scientific and stakeholder
expertise. Specifically, to analyse adaptive mechanisms for minimising negative and maximising positive
impacts/interactions, define appropriate adaptive responses for implementation over short, medium and long-term
time scales, and simulate the effect of different policy scenarios on adaptive strategies.
SID 5 (2/05)
Page 5 of 23
Selection of the study areas
The Regional Impact Simulator aims to improve the understanding of local/regional impacts of global change in a
wide range of stakeholder audiences, and thereby to aid public discussion. In a climatically, physiographically
and economically diverse country such as Great Britain, this was felt to be best supported by allowing the User to
explore consequences in two contrasting regions. The Regional Impact Simulator was developed to be used in
two contrasting regions of England – the North West of England and East Anglia (Table 1)- these were the two
regions used in RegIS, thereby allowing this project to gain value-added benefit from the earlier research project,
socio-economic scenario development and modelling. By capturing a significant range of the heterogeneity of the
country, it was hoped that this would allow users to extrapolate the findings and understanding to other regions of
the UK.
Table 1 Characteristics of the study areas
Region
Boundaries
Development
North West
Cheshire, Greater Manchester,
Merseyside, Lancashire and
Cumbria
Extensive urban development
within Manchester/Merseyside
conurbations
Flooding
Flood-prone areas in narrow river
valleys and coastal fringes
Agriculture
Mostly livestock farming systems,
extensive in uplands and hills but
more intensive in lowlands
Mostly provided from surface
water resources, with significant
imports
Important upland habitats (e.g.
upland hay meadows, artic
alpines)
Water resources
Biodiversity
East Anglia
Norfolk, Suffolk and
Cambridgeshire
Low level of urbanization, with
the exception of key urban
centres such as Cambridge and
Norwich
Significant low lying or below
sea level areas (e.g. Fens)
prone to coastal flooding
Intensively cultivated, with an
emphasis on arable agriculture.
Irrigation important.
Greater dependence on
groundwater resources
Important lowland (arable field
margins, fens) and coastal (e.g.
coastal grazing marsh,
saltmarsh) habitats
The North West is the larger of the two regions and is dominated by the higher (up to about 900 masl) land in the
north and along its eastern boundary, although lower land is found in the west and south of the region. The
proximity of the North Atlantic provides a moderating influence on the climate, producing a generally milder and
wetter climate (average annual rainfall = 650 - 3200 mm), allowing for the effects of altitude. The climate of East
Anglia is influenced by its low relief (most land is below 60 m above sea level (masl)) and its proximity to the
continent. It is the region least affected by the moderating influence of the sea, so rainfall is lower and there is a
greater daily or monthly temperature range than elsewhere. Rainfall (average annual = 550 - 750 mm) is more or
less equally distributed throughout the year.
Methodological context
Fundamental to delivering a software interface (the ‘Regional Impact Simulator’) containing computationally
simpler ‘metamodels’ to the wider stakeholder community, was to build upon the successful stakeholder
integration within RegIS (described in Holman et al 2005a). In doing so, it was necessary to agree with
stakeholder representatives:
•
definitions of the components of the RegIS Integrated Assessment Methodology (IAM)
•
the fundamental guiding design principles for the Interface;
•
useful output indicators for each of the metamodels;
•
appropriate regional adaptive responses to incorporate within the Interface;
•
appropriate Interface functionality.
SID 5 (2/05)
Page 6 of 23
RegIS2 Integrated Assessment Methodology
The assessment framework was based on the Driver-Pressure-State-Impact-Response (or DPSIR) framework,
which was first proposed by the OECD in the early 1990s (OECD, 1993) and has been widely used in
environmental management for the last 10 years, (e.g., Turner et al., 1998; La Jeunesse et al., 2003; Holman et
al., 2005a; Holman et al., 2005b). As a conceptual model it has proved useful for promoting dialogue between the
different disciplines who must work together to solve complex environmental problems and it has been adopted
by relevant government agencies such as the England and Wales Environmental Agency and the European
Environment Agency (EEA, 2003). However, even when used in a conceptual role, the authors are aware of
important differences of view as to what is a driver, pressure, etc. This lack of clarity hinders more widespread
application of the DPSIR framework. There is also an increasing demand for more integrated assessment of
environmental change, which translates into a need to develop formal integrated assessment models. When the
DPSIR framework is being used as the basis for such models the differences of interpretation have to be resolved
and a consensus developed across the model developers and users, and ideally much more widely.
The foci and methodological approaches of the DPSIR domains used in RegIS2 for conducting an integrated
climate change impact assessment are given below.
Drivers:
Definition: the underlying exogenous (to the region) causes of environmental change, e.g. climate and socioeconomic change, national and international (EU) policy.
Method: identify and describe qualitative, narrative storylines, e.g. based on the IPCC-SRES framework.
Pressure:
Definition: the variables that quantify drivers within the region, e.g. regional population, regional GDP, crop prices
or gross margins, sea level, standard of protection, temperature, precipitation, water availability, etc.
Method: develop regional, quantitative scenarios.
State:
Definition: the variables that represent the sensitivity of the system/sector to the pressure variables, e.g. river
flows, land use areas, species distributions.
Method: use meta-models to derive sectorally-specific state variables that include autonomous (spontaneous)
adaptation
Impact:
Definition: a measure of whether the changes in the state variables have a negative or positive effect on
individuals, society and/or environmental resources
Method: derive impact indicators (summary statistics) and translate these into impact classes (good, bad, OK)
reflecting value judgements. Impact indicators are based on whether the state variables reach a certain threshold
determined by expert knowledge/ judgement, stakeholder views and historic analogues. Examples of impact
indicators include the number of flood days, species sensitivity index, etc.
Response:
Definition: planned (societal level) adaptation that aims to minimise negative impacts (or maximise positive
impacts / benefits) by acting on the socio-economic pressure variables – a response may include several policy
measures, e.g. changing water consumption, restricted housing development, conservation plans, etc.
Methods: identify adaptation options with stakeholders; use RegIS2 interface/models to evaluate ways of
minimising impacts by modifying the socio-economic pressure variables. Responses cannot act on the drivers
(e.g. climate change mitigation policy) nor physical pressure variables (e.g. climate variables and sea levels).
Figure 1 shows the linkages within the modelling within the DPSIR framework. The Drivers are in all cases
qualitative descriptions of future environmental issues. They bound the problem that subsequent analyses will
address. The pressures are the quantitative representation of these drivers that provide the inputs to the RegIS2
meta-models. Note that the pressures from one sector are sometimes derived from the states of another sector:
urbanisation and land cover being notable examples. The states and impacts are then the central outcomes of the
RegIS2 meta-models. Finally, the responses then act on the pressure variables. The pressures can be modified
in ways that reflect, for example, alternative policy strategies.
It is important to note that there is no feedback of the responses to the drivers, i.e. climate mitigation policy is not
considered to be an option at the scale of the study regions. Furthermore, the responses are based on generic
options, without detailed specification of how these options could be implemented in practice. Such
implementation strategies are based on socio-political and commercial processes which the stakeholders using
the Regional impact Simulator are better able to assess.
SID 5 (2/05)
Page 7 of 23
Figure 1 Schematic diagram of the model linkage within the DPSRI framework of the RegIS IAM
[Haxagons are RegIS outputs; parallelograms are internal model outputs]
RegIS2 ‘Regional Impact Simulator’ design
Initial concepts
The philosophy behind the Regional Impact Simulator was to make the interface as intuitive as possible, such that
an interested individual could use it with minimal recourse to help files and without need for training. The design
of the interface was therefore crucial to the success of the project, a process requiring a participatory design
approach (van Asselt and Rijkens-Klomp 2002) with the help of stakeholder who formed the project’s Steering
Committee. They represented key groups including national and regional Government, non-government
organizations and industry. There are few existing examples of participatory IA software, the principle exception
being GB-QUEST (Carmichael et al 2004) which has a different purpose - to engage stakeholders in a debate
about what would constitute a desirable future. The initial concepts agreed with the Steering Committee were
that the:
 Design
o layout should be informed by the DPSIR components on the RegIS2 IAM
o Slider bars were the chosen means of varying input parameters in order to increase the
transparency of the model/scenario assumptions;
o Guidance must be given as to ‘realistic’ ranges of values associated with the uncertainty attached
to that parameter within the scenario;
o The user should be able to concurrently view and compare output from more than one sector;
o since different users or audiences prioritise or value issues differently, they will wish to explore
topics in varying degrees of depth or detail, which the design should allow;
 Speed
o the user should be able to rapidly get outputs from the system, without having to go through an
extensive or prolonged model set-up;
o the run times should be as short as possible to prevent Users ‘switching off’;
SID 5 (2/05)
Page 8 of 23

Scale
o
The principle output should be regionally aggregated to prevent the system being used for
detailed site analysis, but that the user should be able to view more detailed mapped output;
The user should be able to vary certain input parameters to look at consequences of changes
(Response) but, because the tool is not intended for intervention specific policy analysis, the
responses available should be given as regional outcomes rather than the policies that would
accomplish those outcomes. For example, a possible Response to insufficient water supply
might be to reduce domestic per capita water consumption. Users may consider outside of the
Regional Impact Simulator which policies or strategies (such as water pricing, metering etc) might
be appropriate to accomplish this outcome.
o
Based upon the initial underlying concepts for the Regional Impact Simulator, the first agreed ‘mock-up’ design is
shown in Figure 2.
Scenario selection
Climate scenario
2020 Low
Socio-economic scenario
2020 High 2050 Low 2050 High
?
?
?
GS
?
WM
?
?
RS
RE
Brief
description
of scenario
?
?
Region
East Anglia
North West
?
?
Coasts
Aggregated indicators
Area flooded = 100km2
Loss of Coastal GM = 10km2
Gain of saltmarch = 20km2
User selected mapped indicator
River flows
Agriculture
GDP
Population
Subsidy
Input value selection
RUN
Scenario ‘best
guess’
uncertainty limit
Slider
Scenario
default value
Figure 2 Initial concept design of the RegIS2 interface
Final design
The Regional Impact Simulator went through multiple subsequent iterations within the project team and with the
Steering Committee as the design evolved. Important additions or evolutions to the initial design were:
 The Interface developed from the original concept of a single screen to the final version which contains
three screens of differing purpose, and increasing levels of interaction with the model variables. It was
felt that the single Interface screen was becoming too complex and would be off-putting to some nontechnical stakeholders because of the apparent multitude of choices. The three screens provide an
ordered hierarchy of interaction:
o The first, Predefined Scenario Futures, screen (Fig. 3)- allows the user to rapidly identify
relative sensitivity of an Indicator(s) to the different climate and/or socio-economic scenarios. For
the chosen climate scenario (baseline or future changed climate for the 2020s or 2050s) and
socio-economic scenarios (baseline or future), a default set of values are used, based upon preexisting data (e.g. from the climate scenarios- Hulme et al., 2002) or the expert judgement of the
RegIS2 team.
o The second, Exploratory Analysis, screen (Fig. 4)- it is widely acknowledged that there is
uncertainty in any scenario, which cannot be defined by a single set of unique values, so this
screen lets the user explore a scenario in greater detail by changing the scenario pressure
variables which act as input parameter values used in the models.
o The third, Influencing the Impacts, screen (Fig 5)- allows the user to test regional adaptation
responses to the impacts previously identified. After an impact indicator is selected, a list of
SID 5 (2/05)
Page 9 of 23





relevant response parameters is presented which are known to affect the selected indicator and
which one may be able to act upon. It does not include climate change variables (e.g.
temperature) as one cannot directly act upon them. Changing any of the sliders and re-running
the model changes the impact indicator and allows the user to identify regional adaptation
responses which reduce or minimise the impact.
Extensive use of tooltips were implemented so that by hovering the mouse over a component of the
screen (scenario name, Indicator etc) a short definition or explanation appears allowing the user to gain a
basic understanding of the concept without having to resort to the Help Files;
Explicit communication of scenario uncertainty was given by a traffic light-based system of colour coding
of the slider bars which differentiates between uncertainty which is “credible” within the context of a
scenario storyline (green for ‘go’) and more “extreme” uncertainty which may be possible but is outside of
the considered wisdom for the scenario (yellow for ‘caution’). For obvious reasons, the ability for the user
to go into the ‘red’ was not given;
The ability to compare spatial Indicator results with those from a previous run;
A library of plant species photographs was included, so that the user could see the species being
modelled;
It was decided to separate Pressures and Responses (which act upon Pressures) within the Interface,
because the effects of climate mitigation (e.g. Responses acting upon the climate Pressures) are not
appropriate at the regional scale.
Figure 3 The Predefined Scenario Futures screen, for investigating default scenario combinations
SID 5 (2/05)
Page 10 of 23
Figure 4 The Exploratory Analysis for conducting sensitivity analysis and exploring scenario uncertainty
and interactions
Figure 5 The Influencing the Impacts screen for testing regional adaptation responses to identified
impacts
SID 5 (2/05)
Page 11 of 23
Scenarios
Climate
Four climate change scenarios were included - the 2020s Low emissions, 2020s High emissions, 2050s Low
emissions and 2050s High emissions. They are all based on scenarios developed on behalf of the UK Climate
Impacts Programme (UKCIP), known as the UKCIP02 scenarios (Hulme et al., 2002). The Low and High
emissions scenarios were selected in order to capture the effects of uncertainties in future greenhouse gas
emissions on UK climate, while the two timeslices allow investigation of the evolution of change. In addition,
these four scenarios are all used within the MONARCH biodiversity project (Berry et al. 2005), allowing
comparison. All the UKCIP02 scenarios are based on the high-resolution regional climate model (HadRM3) from
the Hadley Centre for Climate Prediction and Research. As they are based on a single climate model, they do
not cover uncertainties in how the climate system will respond to these emissions (scientific or modelling
uncertainty). The credible uncertainty range on the climate pressure slider bars indicate the range of predictions
of regional climates over the UK simulated by other global climate models based on Hulme et al. (2002).
Increases in mean annual temperature, relative to 1961-90, range from 0.3 to 1.1oC by the 2020s and 0.5 to
2.6oC by the 2050s. In all seasons, and for all scenarios, there is a northwest to southeast gradient in the
magnitude of the climate warming over the UK, the southeast consistently warming by at least several tenths of a
degree Celsius more than the northwest. Changes in total annual precipitation, relative to 1961-90, range from –
4 to +3% by the 2020s and –9 to +7% by the 2050s. In winter precipitation increases in all regions with the
largest changes of up to 23% occurring in the east. Conversely, in summer precipitation decreases across
virtually all of the UK with the largest changes of up to –32% occurring in the south.
Socio-economic
Future socio-economic change can be as significant, if not more significant, than climate change for some sectors
(Holman et al., 2005b), so that the widespread practice of studying the impacts of future climate changes as if
they were imposed on today’s society is becoming increasingly untenable. Socio-economic scenarios are, to a
certain extent, based on assumptions that are known to be tenuous (Gaffin et al 2004). These scenarios are not
intended to predict the future, but are tools for helping to evaluate societal responses to major environmental
changes such as global climate change, assuming that (Morris et al 2004):
1. The future is unlike the past, and is shaped by human choice and action;
2. The future cannot be foreseen, but exploring the future can inform present decisions;
3. There are many possible futures, so scenarios map ‘possibility space’;
4. Scenario development involves a mix of rational analysis and subjective judgement.
Within RegIS2, the original socio-economic scenarios as used in RegIS (Holman et al., 2005a) were refined,
updated and an improved spatial resolution implemented. The regional descriptive storylines of alternative
futures of the RegIS scenarios (Shackley & Deanwood, 2003) were derived from an iterative procedure between
researchers and stakeholders with interests in the sectors under consideration at national, regional and local
levels and are broadly similar to the related UK national (Berkhout et al. 2002) and global narratives (Nakicenovic
and Swart, 2000; Arnell et al., 2004). In accordance with the original preferences of the RegIS funders and the
UKCIP, formally independent socio-economic scenarios were developed, also referred to as ‘non-climate change’
scenarios, as they do not permit interaction between climate and socio-economic changes, such as occurs in coevolutionary socio-economic scenarios (Lorenzoni et al., 2000).
Using non-climate change socio-economic scenarios has the benefit that it permits a clearer distinction between
the effects of physical climate change, and socio-economic changes (and hence identifies more clearly the role of
response to the impacts identified in the 2050s). The feedbacks between climate change and socio-economic
change within co-evolutionary socio-economic scenarios (e.g. Lorenzoni et al., 2000) make the relative impact of
physical climate change, socio-economic change and socio-economic/political responses difficult to untangle.
From the qualitative scenarios for the region, projections for a number of quantitative non-spatial and spatial
indicators have been developed for use as inputs to the numerical models. These indicators are: population,
Gross Value Added (GVA), persons per household and population density. The quantification has been based on
the qualitative storylines and informed by regional trends and Government projections. Spatial indicators are
urbanization, forest and protected areas.
SID 5 (2/05)
Page 12 of 23
Scenario linkage –to link or not to link global and regional scenarios
Because formally independent socio-economic scenarios were used, it was necessary to decide how to link the
climate and socio-economic (non-climate) scenarios. Two alternative options were possible:
1) To formally link the climate and socio-economic (non-climate) scenarios through their associated
emissions scenarios to create internally consistent visions of the future. This was the approach taken in
RegIS (Holman et al. 2005a) and was based on the rationale that a strong market-driven SE scenario
implies higher emissions of carbon, whilst a community-based SE scenario implies lower emissions of
carbon. In this way only futures in which the emissions of the regional socio-economic scenario were
consistent with the global emissions scenario behind the climate change scenario could be simulated;
2) To provide no formal linkage between the two sets of scenarios, but to leave the choice to the user. This
approach allows regions to develop along divergent socioeconomic paths from those associated with the
emissions scenario behind the climate change scenario.
The second approach was chosen because of the greater flexibility afforded to the User. Because the climate
change scenarios are dependent upon global emissions of carbon, but the SE scenarios operate at a range of
scales from the local to the global, allocation of relationships between the SE scenarios and the climate change
scenarios assumes that the same local to national socio-economic strategy is underway everywhere in the world.
At the scale of the RegIS2 regions this is a highly uncertain assumption, and so the ability to investigate
alternative trajectories of regional socio-economic development that are not necessarily consistent with global
developments was allowed. Guidance is given to the User regarding the consistency of climate and socioeconomic scenario linkages.
Metamodels
Description of the sectoral metamodels
Because the focus of the RegIS2 project was on the development of the IA software, the environmental
compartments modelled (agriculture, biodiversity, coasts and floodplains and water resource management) were
those as used in RegIS. These were originally selected as a result of stakeholder concerns about environmental
change impacts in the study areas (Science and Policy Associates & ESYS, 1996). However, the scope of the
models, range of indicators and linkages between models were greatly expanded from those described in Holman
et al (2005a).
Flooding (Fluvial and coastal environments)
Climate change is predicted to change the frequency of flooding in both the fluvial and coastal flood plains. It will
also affect valuable ecosystems such as saltmarsh and coastal and fluvial grazing marsh. The flood metamodel
considers the impacts of both sea-level rise and changing flood flows in rivers, and the effects on the distribution
of these ecosystems. The present location and standard of flood defences for both coastal and fluvial flood plains
are determined from the National Flood and Coastal Defence Database (NFCDD) and the National Flood Risk
Assessment (NaFRA). Where data on the standard of defence is unavailable it has been calculated based on the
land use and ODPM Guidance (PPG 25).
The flooding methodology uses the 2003 Indicative Floodplain Map (Environment Agency) to identify the flood
hazard zones for the 200 year and 100 year events for coastal and fluvial floods, respectively. The effect of
relative sea level rise on coastal flooding is estimated by raising the surge heights by the magnitude of relative
sea-level rise (i.e. including regional land uplift/subsidence), leading to a reduction in the defence standards.
A similar approach is used to determine the effect of changing the flood flow on flood defence standards of
protection. The High Flows metamodel is based upon the Catchment Descriptors method (Bayliss, 1999) in the
Flood Estimation Handbook, and simulates the QMED or the median annual flood Indicator (Robson et al., 1999).
An increase in peak river flow is translated into an increase in the risk of flooding, based on a relationship
between the change in flow, defence standard and change in defence standard.
The impact on coastal habitats is modelled based on the capacity for saltmarshes to respond vertically and
horizontally to sea-level rise. In the vertical domain, losses of saltmarsh are assumed to occur once the critical
value of the rate of sea-level rise as a function of tidal range has been exceeded. Hence saltmarshes in areas
with a large tidal range are less vulnerable than those in areas with low tidal range. In the horizontal domain,
saltmarsh gains may occur due to managed realignment or unplanned coastal abandonment. Saltmarsh increase
due to migration will often be at the expense of areas that are presently coastal grazing marsh, resulting in a
SID 5 (2/05)
Page 13 of 23
decline in these areas, although the creation of compensatopry inland fluvial grazing marsh is explored by the
model.
Agriculture
A metamodel approach was developed which rapidly simulates the results of the full agricultural land use model
described in Rounsevell et al. (2002). The areas of each grid available for agriculture are initially modified
according to urban and forest extent (using the Land Cover Map 2000 for the baseline and the socio-economic
scenarios for the future), and flooding (from the flood metamodel). The method combines a metamodel which
predicts the yield of a wide range of crops as a function of climate and soil type, with a farm model which
determines the optimum cropping given crop gross margins, labour and machinery costs and soil workability. The
yield metamodel used a neural network which was fitted to spatial modelled yield data from Holman et al. (2005),
derived using the ACCESS model, using soil data (rooting depth and availability of water), weather parameters,
latitude, carbon dioxide level and amount of irrigation.
For the farm model metamodel which calculates the area of cropping which maximises profit, the data was
characterised by the crop gross margin of each crop, the soil type and workable hours in the winter, the maturity
date of each crop, spring and autumn evaporation and rainfall, and the change in labour and machinery costs. A
regression approach was used to predict the percentage of the farm in a given crop and the profit. The results of
the agricultural metamodel then give the level of production of each commodity in each region. The user
specifies the level of production change in each scenario and the level of irrigation available. The model thus
iterates, altering the prices of crops and water, to obtain the correct level of production and irrigation. Based on
the final level of farm profit, land was also classified as either cropped, extensive or abandoned, from which levels
of production, water use and nitrate leaching are calculated.
Water resources management
The water resources management metamodel is divided into a number of modules operating at different spatial
scales. For every sub-catchment and catchment in the two regions, the flow duration curve (FDC) for each is
simulated based upon the Low Flow Estimation methodology, according to the distribution of soil types,
urbanization (which comes from the socio-economic scenarios), climate, cropping and land use classes (from the
Agricultural metamodel). From the flow duration curve, the availability of water for abstraction in each
subcatchment is calculated, after allowing for the environmental flow requirements of the river. Of the many
methodologies available for calculating environmental flow allocations (e.g. reviews by Acreman and Dunbar,
2004; Tharme, 2003), the approach chosen is based upon that used in the Environment Agency's Resource
Assessment Management (or RAM) methodology. This limits abstraction to protect low flows and flow variability
within the flow duration curve, whilst also recognising that abstraction is not unlimited at the high flow end of the
FDC due to water quality and practicality considerations (e.g. turbidity, faecal contamination etc).
The regional water supply availability is derived from the water availability in the catchments, allowing for water
imports/exports and reservoir storage. The regional water demand is the sum of the agricultural water demand
(from the agricultural metamodel), domestic, industrial and commercial demand (based upon the socio-economic
scenarios) and leakage from that proportion of the water demand supplied by main water. Finally a supplydemand balance for the regional provides a measure of the robustness of the supply-demand system to the
pressures exerted in the scenario.
Biodiversity
The SPECIES model (Spatial Estimator of the Climate Impacts on the Envelope of Species; Pearson et al., 2002;
Harrison et al., in press) is used to simulate the impacts of climate change on the potential geographical
distribution of 46 species associated with 8 habitats, chosen to interact with the agricultural, coastal and water
environments. SPECIES uses an artificial neural network to characterise the currently suitable climate space for
a species and to estimate likely changes in its spatial distribution. This way it is possible to see how the climate
space for individual species could contract, expand or shift as the climate changes. Five bioclimatic variables are
used within the model: growing degree days > 5°C; absolute minimum temperature expected over a 20-year
period; annual maximum temperature; accumulated annual soil water deficit; and accumulated annual soil water
surplus. The model is trained using existing empirical data on the European distributions of species to enable the
full climate space of a species to be characterised and to capture their response to climatic conditions that might
be expected under future scenarios. Tests of model performance show good agreement between observed and
simulated distributions for all 46 species. Specifically, all modelled species have statistics based on the Receiver
Operating Characteristics curve (ROC) greater than 0.9, indicating very good discrimination ability, and 37
SID 5 (2/05)
Page 14 of 23
species have Kappa statistics of similarity greater than 0.7 indicating very good agreement whilst 11 have Kappa
statistics between 0.4 and 0.7 indicating reasonable agreement between observed and simulated distributions.
Predictions of potential climate space for each species are combined with data on the distribution of habitats,
protected sites and outputs from the other sectoral models to simulate the impacts of climate and socio-economic
changes on species’ suitability at the regional scale. For the coastal species, model outputs of changes in the
area of salt marsh and coastal and floodplain grazing marsh are used. For fens and bog species, outputs from
the water resources model of indicators of low and high flows are related to species’ preferences for dry and wet
conditions, respectively. For lowland heath and upland hay meadows, output from the agricultural meta-model on
land use distribution and from the socio-economic scenarios of the distribution of protected sites is used to assess
habitat loss. Habitat re-creation is simulated according to rules associated with historic habitat distributions,
relevant soil properties (pH for lowland heath and brown earths for upland hay meadows), output from the
agricultural meta-model on land use distribution and proximity to existing habitat. Finally, for cereal field margins,
outputs from the agricultural meta-model on nitrogen inputs and the distribution of arable land use are related to
species’ tolerances to nitrogen stress for plants, whilst output on switches from spring to autumn sowing is related
to the availability of over-winter food supplies for higher organisms.
Results
The value of the Regional Impact Simulator is not so much in the results of the simulations carried out by the
RegIS2 Team, but in the ability it gives Stakeholders to explore future scenarios and therefore in their simulation
results, interpretations and actions. However, example results from using the Regional Impact Simulator are
given below from simulations of the following common scenario futures:
 Baseline;
 2050s Low climate + Global Markets future;
 2050s High climate + Global Markets future;
 2050s High climate + Regional Stewardship future.
Flooding under future climate change and socio-economic scenarios in East Anglia and
North West England
The Median Annual Flood (QMED) increases under all climate scenarios, with the increases being greatest in the
2050s and for the North West. Under the Regional Stewardship scenario, urbanization decreases so that the
increase of QMED due to climate change is slightly diminished to increases of 5-14% in QMED. The impacts of
climate change on QMED are generally exacerbated in both regions by the Global Markets scenario, in which
urbanization increases, for both the 2050s Low and High scenarios, with QMED increasing by 7-16% and 1021%, respectively.
Without adaptation, climate change will lead to the area at risk of flooding and the economic damage for both
regions increasing with the increase in the climate pressures. For the 1 in 75 year event, the area at risk of
flooding will maximise under the 2050s High scenario and the Global Market scenario to reach 297215 and
109649 hectares in East Anglia and North West, respectively. The estimated economic damages will be £8606
and £ 6600 million in East Anglia and North West respectively. The number of people affected by flooding has a
different pattern. The highest figures of people affected will be reached under the 2050s High and Regional
Stewardship scenarios; this is anticipated because the number of household occupancy under the Regional
Stewardship scenario would be 3 while it would be 2 under the Global Market scenario.
Given that the increase in sea level and river peak flow can cause a decline in the standard of protection of the
sea and river defences respectively, responding to climate change by raising the tidal and fluvial defences will
result in increasing their standard of protection, and lead to a great reduction in the areas at risk of flooding and a
decline in economic damages despite the climate change pressures. For example, the adaptation measures
under the 2050s High and Global Market scenario in East Anglia will decrease all impacts of flooding: the total
areas at risk of flooding will decrease from 95605 to 46788 hectares for the 1 in 10 year event, the economic
damages will go down from £ 2304 to £1414 million, and the number of people affected will decrease from 54090
to 30232.
Upgrading flood defences to existing Defra guidance and maximising the creation of habitats under the Regional
Stewardship combined with the climate pressures under the 2050s High scenario suggest that the areas at risk of
flooding and the estimated damages are less than those experienced today in both regions. This indicates that
although climate change is the biggest factor in increasing areas at risk and consequently damages, our adaptive
responses can generate a significant decrease in areas at risk and damages. However, under this scenario there
SID 5 (2/05)
Page 15 of 23
is a cost of displacing people as the tidal defences are retreated to allow the saltmarsh in the coastal flood plain to
migrate inland in order to maximise stocks. This is not quantified in the flood metamodel and the inundated areas
are not considered at risk of flooding.
Impact of future socio-economic and climate changes on regional agricultural land use
and the consequences for the wider environment
The results of the study show that in East Anglia cropping remains quite stable, though grassland is eliminated in
the 2050s High scenarios – almost certainly due to the low yield in the drier conditions. Sunflowers are only
significant in the 2050High + Regional Stewardship scenario, which is probably related to the high flat rate
subsidy in the Regional Stewardship scenario. Nitrate leaching increases as fertiliser use increases with the large
increases in crop yields in the Global Markets scenarios, because fertiliser requirement is proportional to
expected yield. In the North West there is a very much greater range of outcomes though all future scenarios
suggest a reduction in grassland with the greatest in the 2050s High climate scenario combined with the Regional
Stewardship socio-economic scenario. In the Regional Stewardship scenario, the low level of demand for milk
products, which can easily be supplied from a smaller area, and the flat subsidy structure encourages the widest
range of break cropping.
None of the six cereal field margin species modelled lose climate space in East Anglia, but they all gain some
climate space in North West England ranging from 2% for Papaver rhoeas to 62% for Legousia hybrida under the
2050s High scenario. The socio-economic scenarios show little difference in the arable area. However, the effects
of nitrogen fertiliser use on species suitability show much greater differences between the socio-economic
scenarios. Species’ suitability is lowest in East Anglia under the Regional Enterprise scenario where between 19
and 27% of grid cells are classified as being marginal due to nitrogen stress, whilst between 3 and 6% are
classified as unsuitable. This is followed by the Global Markets scenario, where between 10 and 19% are
classified as marginal due to nitrogen stress, whilst between 0.5 and 2% are unsuitable. Finally, the Global
Sustainability and Regional Stewardship scenarios show all grid cells to be suitable. Species’ suitability is
slightly worse under the 2050s Low climate scenario combined with the different socio-economic scenarios than
the 2050s High scenario, because there are slightly more N input in the former case. For the North West region
under the 2050s Low and Regional Enterprise scenario. Only 2% of grid cells are classified as marginal due to
nitrogen stress for the most tolerant species, Galium aparine. For Papaver rhoeas (medium tolerance), 13% are
marginal and 2% become unsuitable due to nitrogen increases. The least tolerant species, Legousia hybrida,
shows 12% of grid cells becoming marginal and 15% becoming unsuitable due to nitrogen stress.
In order to reduce the impacts on species with low and medium tolerances to nitrogen, a possible adaptation
option would be to restrict chemical inputs, although the adoption of such a policy though would depend on the
socio-economic scenario under consideration, being less likely under Global Markets.
Linking water resource availability and agricultural irrigation demand
Climate change impacts the naturalised flows on which the abstraction availability is based, with the reduced
annual rainfall in the 2020s and 2050s being reflected in reduced river flows. By the 2050s under a High climate
change scenario, abstraction availability in East Anglia is reduced by over 300 Ml/d, a reduction of around 30%
compared to current availability. Similar reductions in abstraction availability are seen in the North West region,
although abstraction availability is much higher in the North West compared to East Anglia.
In parallel with the reduction in abstraction availability, there are variable changes to irrigation demand in both
region. The greatest increases are in East Anglia under a 2050s High climate (40% increase in demand
compared to the baseline) and in the North West under 2050s Low and High climates where demand increases
by between 200 and 250 %. The larger percentage rise in the North West reflects the very low baseline irrigation
demand in the region.
The net result of the decrease in abstraction availability coupled by the increase in irrigation demand differs
between the two regions. Even under a 2050s climate, with its resultant decrease in abstraction availability, there
is still a positive supply-demand balance in the North West, indicating that increased irrigation demand can be
met from the available resource. However, in East Anglia, the supply-demand balance under baseline socioeconomic conditions shows a progressively increasing supply deficit.
With changing futures, the profitability of irrigation changes, due to changes in water costs, demand for produce,
gross margins etc. The extent to which the environment will be protected is a function of the socio-economic
context, and differs greatly between the socio-economic futures. Under the GM and RE scenarios the Regional
Environmental Priority is low but the interaction of supply with demand differs. Under the GM scenario and
sufficient water can be abstracted to satisfy demand, giving a positive supply-demand balance. Under GS and RS
SID 5 (2/05)
Page 16 of 23
scenarios, the Regional Environmental Priority is high and the necessary water is allocated to protect the river
ecology at the expense of abstraction availability. The net result is that despite reduced water demand in these
futures, there is a significant water resources deficit which is likely to lead to further restrictions on water demand,
particularly in drier than average years.
The adaptive capacity of the water sector in these futures is, however, fairly high, although this is for contrasting
reasons. Under the GM and RE scenarios, the focus of water management is on supply, which means there is
potential adaptive capacity in demand management. Under the RS and GS scenarios, demand management has
very effectively lowered demand so that there is little further capacity to adapt. However, increasing supply
remains an option if more water were to be abstracted from the environmental allocation of water.
The efficacy of pricing to control water demand also varies according to the socio-economic conditions prevailing
within a future. Under futures where local sustainable agricultural production is a priority, total irrigation water
costs need to rise significantly before significant reductions in irrigation abstraction were observed. In such socioeconomic futures, constraints on irrigation demand are best managed through volumetric restrictions on the
abstraction licences, and trigger limits on summer abstraction.
Wetland habitat changes from an integrated assessment linking water resources and
biodiversity
Changes to the high and low flows within each catchment are a consequence of the changing climate, the
changing landscape (both urban and rural) and the changing non- consumptive water demand patterns of society.
However, the relative sensitivity of a catchment’s flows to climate and socio-economic changes depend largely
upon changes in urbanisation and the components of water demand (especially per capita water demand) which
are closely linked to the socio-economic scenario chosen. In catchments where there are no dramatic changes in
urbanisation, the flows are more sensitive to climate change rather than to the socio-economic changes. The
results therefore show the significant influence of future urbanization, population water requirements and regional
environmental allocation of water on the hydrological responses of catchments to future change. In general,
climate change in the 2050s will produce decreases in naturalized flows, the decreases being greater under a
2050s High climate change scenario and in East Anglia. Whether the effects of the socio-economic scenario
moderate or amplify these reductions depends on the characteristics of the scenario and the region:
 In those catchments with low levels of urbanization (in much of the northern and central parts of the North
West and much of East Anglia to a greater extent) and under GS and RS scenarios with low water
demand (especially per capita water consumption), flows will be little affected by these socio-economic
factors, and real flows will be determined primarily by climate change.
 In those catchments with low levels of urbanization and under GM and RE scenarios in which water
demand is high (especially per capita water consumption), increased abstraction will lead to the
reductions in naturalized flows caused by the impacts of climate change being amplified;
 Finally, in those catchments with significant urbanized areas, the influence of urbanization and population
water requirements can be highly significant. Decreases in low flows, as given by the daily naturalised
Q95, will be partially or in some cases completely moderated by the effects of effluent returns. High
flows, as given by the Q5 may increase to a lesser extent than Q95 due to the increased returns, leading
to potential increases in waterlogging stress.
Three of the six fen species modelled show virtually no change in climate space in East Anglia (P. australis, V.
moulinsiana and G. maxima). The other three species show no gains in climate space, but significant loses of
28% (E. palustris), 40% (V. dioica) and 56% (R. scleratus) centred in the west and south of the region under the
2050s High scenario. In all cases this is likely to be due to increased summer dryness. Four of the 11 bog
species show only small changes in climate space in the North West (E. vaginatum, M. gale, R. alba and C.
vulgaris). The other seven bog species show either no or small gains in climate space combined with larger
losses, ranging from 11% for B. pendula to 86% for C. tullia, under the 2050s High scenario.
The impacts of high and low flows on species’ suitability under the climate and socio-economic scenarios
depends on the species’ sensitivities to drought and waterlogging stress. As fens are a very restricted habitat, the
habitat mask led to only 14% of the squares being suitable for either fen species. P. australis, a low sensitivity
species, shows little change under the scenarios, while R. scleratus, a medium sensitivity species, shows that 2%
of grid cells become unsuitable under the 2050s Low scenario, with 0.5% becoming marginal due to dryness
under GM and one square marginal due to wetness under RS. Under the 2050s High and GM scenario, a further
4% of cells become unsuitable due to loss of climate space or habitat, 3.9% of grid cells become unsuitable due
to dryness and 1.1% become marginal, for RS these figures are 4%, 0% and 1.2% respectively.
In the North West, species with a high drought tolerance show very little loss under any of the scenarios, E.
vaginatum, for example, only loses 4% of grid cells under the 2050s High GM scenario in the Lake District. For
those with a medium drought tolerance, such as L. cordata, 14.5 % of grid cells become unsuitable under the
SID 5 (2/05)
Page 17 of 23
2050s Low GM and RE scenarios. A further 8% are lost under 2050s High for these scenarios, mostly from the
southern part of the region. Under 2050s High GM an additional 4.4% became marginal due to dryness along the
Eden valley. For S. cuspidatum, under the 2050s Low GM and RE scenarios between 14 and 17% of cells
become marginal due to dryness in northern Lancashire and Cumbria and under the GM 2% of cells in the Eden
valley become unsuitable due to dryness. This pattern continues for the 2050s High GM and RE scenarios with
6% of grid cells becoming unsuitable due to climate, and 4% under GM due to dryness, with the consequence
that the number of marginal cells decreases.
The possibilities for adaptation are limited and being based on increasing water demand and supply that could be
difficult to meet, especially in East Anglia.
Regional scale assessment of the impacts of climate change on coastal and fluvial
ecosystems and the scope for adaptation
The results for the two regions demonstrates the differences that the critical value of relative sea level rise has on
the viability of coastal habitats under future sea level rise. The North West has a generally higher sediment and
biomass availability than East Anglia, which is reflected in the current (baseline) condition of less threatened
habitats, and lower predictions of future losses. Within East Anglia, Suffolk experiences increased future losses
of saltmarsh in comparison to Norfolk, for the same reasons of lower sediment and biomass availability,
demonstrated by a lower critical value.
The socio-economic choices can have more effect on habitats and their recreation than the future changes in
climate. Currently, there are insufficient areas for the recreation of habitats that are predicted to be lost under
future sea level rise, however, this can be mitigated by the provision of more suitable areas. Specific examples
include subjecting agricultural land and low density urban areas to increased flooding frequencies through
removal or abandonment of defences, or relocation of properties. The conversion of both these areas to coastal
and fluvial grazing marsh through an increase in flooding frequency is subject to the adaptation choice that these
areas will not be protected to the same degree as dense urban areas. Under the Global Markets scenario it
would be more evident that arable land would not be protected, as in the global context the retention and
protection of these areas would be uneconomic.
There is intense competition between saltmarsh and coastal grazing marsh for space around the intertidal zone,
due to coastal squeeze limiting the area available to saltmarsh and coastal grazing marsh limited to the extent of
the floodplain and the available land within that. Competition for space ultimately leads to the loss of grazing
marsh over the 2050s timescale due to the difficulties of maintaining coastal grazing marsh in-situ. An adaptation
option is the creation of fluvial grazing marsh as a replacement for coastal grazing marsh. The creation of both
coastal and fluvial grazing marsh does presume that management will be undertaken to preserve these habitats,
through the provision of some defences or protection and active grazing management. These aspects are not
included in the metamodel and under which socio-economic scenarios this kind of management is likely has not
been investigated.
Discussion
Use of the Regional Impact Simulator
It was never intended that the Regional Impact Simulator would be a Decision Support System, but rather it would
have a wider educational / information role. It was hoped that, through allowing the user to explore the uncertain
future and consequently improve their understanding of the issues and the implications of the multiple interactions
between the sectors driving landscape change, it would inform better decision making and policy development.
In considering the use of the Regional Impact Simulator within adaptation frameworks, a broad distinction can be
made between its use in adaptive capacity building and its use in informing actions that implement operational
adaptation decisions. The wider use of the Regional Impact Simulator can support elements of adaptive capacity
building by communicating climate change information and building awareness of potential impacts at an
appropriate regional level.
Implementing adaptation decisions focuses on avoiding anticipated or reducing cumulative impacts of climate
change, ensuring that distribution impacts of adaptation are minimised and ensuring that adaptive measures
taken by one organization or sector do not adversely impact upon others (Adger et al., 2005). The Regional
Impact Simulator can inform many of these issues, through spatially displaying the impacts and consequences of
regional adaptation responses. However, it is its use to inform the integration of adaptation actions and policies
across sectors, such that unintentional adaptation resulting from actions in one sector does not reduce the
SID 5 (2/05)
Page 18 of 23
effectiveness of purposeful adaptation in another sector (which remains a key challenge to achieve affective
adaptation -Adger et al., 2005) which may be the most significant use of the Regional Impact Simulator.
Assessing the success of the Regional Impact Simulator in delivering the needs of
stakeholders
From a scientific perspective, RegIS2 was successful in developing the use of computationally simpler modelling
techniques for IA, and has produced new understandings which are the subject of the main report. However, the
question needs to be asked as to whether a user friendly interface was successfully developed, and whether the
stakeholder community will make use of the Regional Impact Simulator and the RegIS methodology.
The latter question will only be answered in due course, after the Regional Impact Simulator is made freely
available to the wider Stakeholder community and time to judge its use has been elapsed. However, a beta
prototype of the Regional Impact Simulator was released to a group of approximately 60 volunteer stakeholders
for external evaluation. These stakeholders came from national Government (8), regional Government (5), local
Government (6), NGO’s (18, ranging from National Park authorities and environmental regulators to sustainability
fora), charities (5), commercial organizations (11, ranging from water and power utilities to consultancies) and
academic/research institutions (6) from within the two regions, from elsewhere in the UK and from outside the UK.
The evaluation process was intended to get independent feedback on the design and functionality of the Regional
Impact Simulator, rather than on the numerical values of the output itself. Questions addressed the design and
operation of the Interface, the usefulness of the model output and potential future developments. Users were
asked to score their answers between 1 (Low) and 5 (High) or to identify strengths and weaknesses. The results
of the feedback provide an important metric for whether the Regional Impact Simulator achieved its aims
Key determinands of the success of the interface were:
1.
Whether it was deemed to be easy to use and intuitive;
2.
Whether the runtime was considered acceptable;
3.
Whether the output indicators were useful.
Ease of use and intuitiveness
All bar one of the evaluators considered that the ease of use/intuitiveness of the Interface was adequate or better,
although none rated it as ‘very easy’. The intuitive feel of the Interface was not something that the evaluators
could assess prior to installing and running the software. Therefore we do not believe that the respondents were
positively self-selecting with regard to this metric.
There was an approximately even distribution of responses to how long it took to understand how to use the
Regional Impact Simulator, with no relationship between their assessment of this and the intuitiveness. There
was an even split between those who needed to use the help tutorial and those who didn’t. All those who
required more than 1 hour to understand the interface used the Tutorial, as did some who required half an hour or
less.
It is not possible to provide a substantive assessment of whether the design fully met it’s objective. However, we
believe that the generally positive assessment of its intuitiveness when considered with the even distribution of
times taken and need for the tutorial, which suggests that the stakeholders had a broad range of backgrounds
and technical ability, provides an affirmation for the interface design. Whilst a single design will never fully meet
the diverse needs of all stakeholders, the hierarchical nature of the three screens within the Regional Impact
Simulator appears to have satisfied most respondants.
Run times
The run times for the Regional Impact Simulator are longer than was initially hoped for, prior to model
development. This partly relates to the finer spatial resolution adopted by the flooding metamodel (1km) and
partly to the reading and writing of files between the metamodels. The runtime is also computer-dependent, as
can be seen from Table 2 which shows the run-times for 2 contrasting PCs representing an old PC and a current
typical PC. Despite this unavoidable variability in performance, all but 2 of the respondents felt that the runtime
was adequate or better. Although the ‘typical’ new PC had a run time of around 20 seconds, this is reduced to
around 10 seconds when run without the biodiversity models- the last in the cascade of models.
Usefulness of outputs
All of the evaluations rated both the usefulness of the indicators and model results as adequate or better
SID 5 (2/05)
Page 19 of 23
Table 2 System performance of the Regional Impact Simulator
PC ‘type’
Processor
RAM
Windows
East
Anglia*
Operating
runtime (secs)
system
‘old’ (4 years Pentium
4 512Mb
2000
291 [54]2
old)
1.6 GHz
‘Typical’ (1
Pentium
4 1Gb
XP
14 [28]
year old)
2.8 GHz
* 2050s High climate + Regional Enterprise scenarios- all settings at default
1 Only saltmarsh habitat selected
2 All habitats selected
3 Only blanket/raised bog habitat selected
North
West*
runtime (secs)
363 [65] 2
19 [31]
Credibility of using reduced form or metamodels within the Regional Impact Simulator
It is important that the fast run-times above do not affect the perceived credibility of the Regional Impact
Simulator. The multiple facets that determine the credibility of an IAM in the eyes of stakeholders dictate that it
has been approached from a variety of perspectives:
 Basing the metamodels on widely accepted pre-existing models;
 All of the metamodels have either been validated against results from their widely accepted pre-existing
model or against observed data;
 The slider bars with which the User changes the Pressures or Responses show the numerical values, so
that the magnitude of changes being made are clear;
 The fast runtimes enable Users to carry out many simulations to rapidly investigate the sensitivity of
system under baseline conditions, and to compare this, where appropriate, with their own knowledge;
 In the Influencing the Future screen, the user is automatically guided to the Responses that that affect the
chosen Indicator. It is therefore transparent to the user which choices will have greatest effect;
 A hierarchy of User support - tooltips (for immediate provision of a short definition or explanation);
concise higher level Help files which provide basic background understanding for the majority of Users,
and full documentation within the Help files for technical users;
 Explicitly communicating that the very model inputs themselves which derive from the scenarios are
uncertain. The colour coding approach used in Regional Impact Simulator, and the actual breadth of the
coloured uncertainty bands on the slider bars conveys clearly that different values (and the assumptions
behind them) are likely to give rise to differing outcomes which the User can rapidly explore due to the
rapid run times.
Future Developments
During the evaluation process, it was also asked whether it would be useful to develop the Regional Impact
Simulator further. There were no negative responses, with most being highly positive (scoring 4 or 5). It was
then asked whether the development should be in expanding or changing the range of sectors, expanding or
changing the regions modelled or extending the simulation to provide national coverage. Although a small
number of respondents felt that the scope of any further development should be determined by a wider
consultation exercise, most respondents selected “more or different regions” and “national coverage”.
Two issues are evident from this. Firstly that there is a wider demand for regional integrated assessment- many
of the respondents were in regional or national organizations and could see the benefits of such a tool for their
own area. Secondly, although there is a perception that the regional scale is the most appropriate for decision
making with regard to climate change impacts and adaptation, respondents felt that national coverage was
required. This may reflect the small size of the UK and the importance of national organizations such as the
regulatory organizations (e.g. Environment Agency, English Nature).
A number of developments would enhance the capacity and utility of the system:
a)
application to new regions and/or coupling regional and national scale applications- this was
suggested by most respondents in the evaluation process,
b)
refinements to the user interface- e.g. different ways of visualising and classifying the output data,
such as enhanced GIS and data classification tools. It would also be interesting to explore the
potential of allowing users to control/modify the spatial input data within the interface (e.g. adding
zones of environmental protection, planning restrictions, new housing development, etc.), rather than
SID 5 (2/05)
Page 20 of 23
c)
d)
e)
implementing aggregate effect on the whole of the region. Allowing the user to do this is clearly of
relevance given the importance of location for some model outputs;
refinement to existing models- e.g. changing spatial resolution, improving the temporal resolution by
running dynamic simulations. Dynamic simulations would better account for the path dependency of
socio-economic futures;
addition of new meta-models (e.g. urban development, forestry);
new assessment concepts (risk, valuation, visualisation).
Conclusions
The RegIS2 project has developed a user-friendly tool (the Regional Impact Simulator) to allow stakeholders, for
the first time, to carry out their own impact and adaptation studies. The major advantage of the metamodel
approach used in the Regional Impact Simulator is the large number of analyses that can be carried out in a short
space of time. A single run of the whole system typically takes about 20 seconds, though speed of computer has
a large impact, as does the number of habitats selected for analysis. This enables impacts to be studied in far
more detail, though this ironically leads to as many questions as to why a response happens. This should
however be viewed in a positive light as, where previously the answers were simply a black box to be taken on
trust, it is now possible for a user to study what is happening in response to changes in some detail.
Integrated, regional assessment represents a powerful tool in environmental change impact and adaptation
studies that has many advantages over ‘traditional’ sectoral approaches. IA models such as RegIS allow the
relationships between sectors to be evaluated and thus the consequences of changes in one sector to be
assessed with respect to another sector. Single sector studies are clearly limited by their inability to treat all of the
processes that may be important in affecting a sector. It becomes clear when undertaking assessments in this
way that the various regional-scale sectors are interdependent with changes and potential adaptation options for
one sector sometimes having profound effects on other sectors. The use of the Regional Impact Simulator to
inform the integration of adaptation actions and policies across sectors, such that unintentional adaptation
resulting from actions in one sector does not reduce the effectiveness of purposeful adaptation in another sector,
may be the most significant use of the Regional Impact Simulator.
The Regional Impact Simulator addresses both autonomous and societal adaptation strategies. However, whilst
autonomous adaptation is dealt with explicitly by the modelling approaches, the Regional Impact Simulator does
not indicate the policy mechanisms that would be needed to implement such societal adaptation strategies in
practice. The need remains for stakeholders to make decisions themselves about what is or is not feasible,
achievable and/or desirable in terms of adaptation responses, as they are best placed to do this, informed by
using the Regional Impact Simulator to explore the effect of acting on the pressure variables through regional
responses. In such a way, the Regional Impact Simulator should promote better informed decision makers and
decision-making.
References
Acreman, M. and Dunbar, J. (2004). Defining environmental river flow requirements- a review. Hydrology and
Earth System Sciences, 8(5), 861-876.
Adger, W.N., Arnell, N.W. and Tompkins, E.L. (2005). Successful adaptation to climate change across scales.
Global Environmental Change, 15, 77-86.
Arnell, N.W., Livermore, M.J.L., Kovats, S. Levy, P., Nicholls, R., Parry, M.L. and Gaffin, S.R. (2004). Climate
and socio-economic scenarios for global-scale climate change impacts assessments: characterising the
SRES storylines. Global Environmental Change, 14(1), 3-20.
Berkhout, F., Hertin, J. and Jordan, A. (2002). Socio-economic futures in climate change impacts assessment:
using scenarios as learning machines. Global Environmental Change, 12, 83-95.
Berry, P.M., Harrison, P.A. and Dawson, T.E. (2005). Modelling Natural Resource Responses to Climate Change
(MONARCH2): A Local Approach. Development of a conceptual and methodological framework for
universal application. Environmental Change Institute, University of Oxford, 275pp.
Carmichael J, Tansey J, Robinson J (2004). An integrated assessment modelling tool. Global Environmental
Change 14, 171-183
EEA [European Environment Agency] (2003). Europe’s environment: the third assessment. Environment
Assessment Report No. 10, European Environment Agency, Copenhagen, pp. 343.
SID 5 (2/05)
Page 21 of 23
Gaffin, S.R., Rosenzweig, C., Xing, X. and Yetman, G. (2004). Downscaling and geo-spatial gridding of socioeconomic projections from the IPCC Special Report n Emissions Scenarios (SRES). Global Environmental
Change 14, 105-123.
Holman I.P., Rounsevell M.D.A., Shackley S., Harrison P.A., Nicholls R.J., Berry P.M. and Audsley E. (2005a). A
regional, multi-sectoral and integrated assessment of the impacts of climate and socio-economic change in
the UK: I Methodology. Climatic Change, 71, 9-41.
Holman, I.P. Nicholls, R.J. Berry, P.M. Harrison, P.A. Audsley, E. Shackley, S. and Rounsevell, M.D.A. (2005b).
A regional, multi-sectoral and integrated assessment of the impacts of climate and socio-economic change in
the UK: II Results. Climatic Change, 71, 43-73.
Holman, I.P., Loveland, P.J., Nicholls, R.J., Shackley, S., Berry, P.M., Rounsevell, M.D.A., Audsley, E., Harrison,
P.A., and Wood, R. (2001). RegIS – Regional Climate Change Impact Response Studies in East Anglia and
North West England. DEFRA, London, p.20, (available from www.ukcip.org.uk).
Holman, I.P. and Loveland, P.J. (eds.) (2001). Regional Climate Change Impacts in East Anglia and the North
West (the RegIS project). Final report of MAFF Project No. CC0337 (available from www.ukcip.org.uk).
Hulme, M., Jenkins, G.J., Lu, X., Turnpenny, J. R., Mitchell, T.D., Jones, R.G., Lowe, J., Murphy, J.M., Hassell,
D., Boorman P., McDonald, R. & Hill, S. (2002). Climate Change Scenarios for the United Kingdom: The
UKCIP02 Scientific Report. Tyndall Centre for Climate Change Research, School of Environmental
Sciences, University of East Anglia, Norwich, UK, 120 pp.
La Jeunesse I. M., Rounsevell, M., Vanclooster, M. and Romanowicz, A.A. (2003). ‘Delivering a Decision Support
System tool to a river contract : a way to implement the participatory approach principle at the catchment
scale?’ Physics and Chemistry of the Earth, 28, 547-554
Lorenzoni, I., Jordan, A., Hulme, M., Turner, R.K., and O'Riordan, T. (2000). A co-evolutionary approach to
climate change impact assessment: Part I. Integrating socio-economic and climate change scenarios. Global
Environmental Change, 10, 57-68.
Morris J., Gomez M., Vasileiou K. And Berbel J. (2004). WADI scenario definition. In Vecino J.B. and Martin
C.G. (eds). Sustainability of European Irrigated Agriculture under Water Framework Directive and Agenda
2000. Final report of 5th Framework project EVK1-2000-00057, ISBN 92-894-8005-X.
Nakicenovic, N. and Swart, R. (eds.) (2000). Special Report on Emissions Scenarios. A Special report of Working
Group III of the Intergovernmental Panel on Climate Change. Intergovernmental Panel on Climate Change,
Cambridge University Press Cambridge, p. 599.
Organization of Economic Co-operation and Development (1993). OECD Core set of indicators for environmental
performance review. Environmental Monograph No. 83, OECD, Paris.
Pearson, R.G., Dawson, T.P., Berry, P.M. and Harrison, P.A. (2002). SPECIES: A spatial valuation of climate
impact on the envelope of species. Ecological Modelling 154(3), 289-300.
Rounsevell, M.D.A. Annetts, J.E., Audsley, E. Mayr, T. and Reginster, I. (2003). Modelling the spatial distribution
of agricultural land use at the regional scale. Agriculture, Ecosystems and Environment, 95(2-3), 465-479
Science and Policy Associates Inc and ESYS: 1996. Proposal for a UK Integrated Climate Change Impacts
Assessment. Report prepared for the Department of the Environment and the Environment Agency.
Department of the Environment, London, p. 28.
Shackley, S. & Deanwood, R. (2003), Constructing social futures for climate-change impacts and response
studies: building qualitative and quantitative scenarios with the participation of stakeholders. Climate
Research, 24 (1), 71-90.
Tharme, R.E. (2003). A global perspective on environmental flow assessment: emerging trends in the
development and application of environmental flow methodologies for rivers. River Research and
Applications, 19, 397-441.
Turner, R.K. et al. (1998) Coastal management and environmental economics: analysing environmental and
socio-economic change on the British coast. The Geographical Journal, 164, 269-281.
van Asselt MBA and Rijkens-Klomp N (2002). A look in the mirror: reflection on participation in integrated
assessment from a methodological perspective. Global Environmental Change 12, 167-184
References to published material
SID 5 (2/05)
Page 22 of 23
9.
This section should be used to record links (hypertext links where possible) or references to other
published material generated by, or relating to this project.
SID 5 (2/05)
Page 23 of 23