THE WORLD ENDS THIS WEEKEND - PROFECIES VON SOARES Adapting to Climate Change Thresholds, Values, Governance Making adaptation happen for the common good Adaptation has always taken place, and is likely to continue doing so. Human beings have been able to adapt to changing environments and societies, surviving and flourishing overall. However, if we hold a lens to the adaptation process and analyse it further in detail, it becomes clear that environmental and social change does not affect everyone equally. Less resilient communities – and more vulnerable individuals – can be severely affected by change, thus limiting their opportunities for adaptation. The prospect of climatic changes of greater magnitude and frequency than those experienced throughout most of human history beg the question of whether adaptation is possible and how adaptation to present and future changes may be facilitated. In very simple terms, adaptation entails an adjustment to changing conditions. On a social level, this can be interpreted as some form of cognitive or behavioural response at individual and collective levels, both being invariably entwined. Understanding adaptation in the context of climate change requires careful consideration of two dimensions: scale (Who is responding where, to what?) and purpose (Why are we responding? What are the aims of adaptation?). Let us consider these in turn. Adaptation occurs at different but related levels. Policies shaped by national and international circumstances set objectives to be achieved at local and regional levels. Individuals and organisations however do not operate in isolation. Interpretation of information and its translation into decisions and behaviours are affected by social context, individual characteristics and direct experiences. In other words, adaptation is a multi-scalar process of multi-level governance, concerned with the interaction of individual and collective behaviours acting from the bottom–up and the top–down in response to changinG ...Climate change and agricultural regime shifts In terms of limiting the ability of humans to adapt to climate change, it is the transformation of much of the Earth’s terrestrial surface to agricultural lands that is likely to be the most substantial. Not only does it represent a massive modification of Earth’s ecological functioning, but its continuation is considered to be essential for human well-being. Agricultural ecosystems cover an estimated 40% of Earth’s surface, but they also create impacts on other ecosystems. One of the major ways that agriculture affects distant ecosystems is through its modification of global water flows. Agriculture does this in some obvious ways. About two-thirds of the water removed from rivers is used for irrigation (Scanlon et al., 2007), and the water that flows from agricultural lands into rivers and lakes carries with it agricultural fertilizers that reduce water quality in aquatic ecosystems (Bennett et al., 2001; Galloway et al., 2004). However, less obviously, agriculture alters atmospheric flows of water due to the impacts of irrigation and deforestation on global evapotranspiration (Gordon et al., 2005). It is via both direct and indirect impacts on other ecosystems that agriculture has increased the supply of desired ecosystem services, such as food and fibre, but at the same time led to unintended declines in non-agricultural ecosystem services, such as fisheries, flood regulation and downstream recreational opportunities ( Millennium Ecosystem Assessment, 2005). Managing trade-offs is difficult due to the social and ecological complexities involved, and managing them will be made even more difficult in a changing climate. However, while these changes would be difficult to cope with even if gradual and predictable, ecological research suggests that declines in ecosystem services may also be abrupt and surprising – and difficult to reverse. Abrupt changes in ecosystem services can occur due to shifts between different ecosystem regimes, and this presents a substantial challenge to ecosystem management and development goals Characterisation of adaptation options Table 3.2 summarises adaptation options that have been identified by water companies, the Environment Agency, pressure groups and local councils as potentially feasible in the Medway catchment. Some of these are specific resource schemes (which will also serve other catchments), whilst others are options applicable to water resources more generally. Many of these options have been incorporated into the Water Resources Management Plans of the water companies responsible for water resources in the Medway catchment. The complex responsibility for water resources in the catchment means it is necessary to consider schemes across the Kent region. The table provides indicative estimates of the potential contribution of each option, where these are available (estimates are in many cases very generalised, and not to be taken too literally Decision pathways and defining adaptation thresholds It has been clear from the outset of TE 2100 that to deal with uncertainty associated with the likely effects of climate change there was a need to move away from reactive flood defence towards the proactive adaptive management of future flood risk. Historically, London’s flood defences have been raised and improved in the aftermath of various flood catastrophes – typically to a height just above that of the flood that had just been experienced. These incremental raisings can be readily seen in many of the flood walls flanking the River Thames, whose present-day crest height was largely defined by the 1953 flood event. The proactive management of risk promoted within TE 2100 sees a series of timed interventions seeking to manage flood risk within an acceptable zone. This vision recognises that if the risk were to be left unmanaged, it would increase in the future as the impacts of climate change along with development pressures on the floodplain become more acute and as the asset base deteriorates with time (Figure 4.1). However, through the implementation of risk management responses at different points in the century, this risk can be managed within the appropriate bounds. The appropriate bound for flood risk is largely determined through interpretation of the government’s guidance on flood risk management and will include an element of cost–benefit analysis.

 Discussion: five key factors in shaping vulnerability
Emerging from the results above are five key factors that shape vulnerability to the
effects of heat waves among the participants of this study. This section explores
these factors in more detail.
Many elderly do not recognise their vulnerability
It has been argued that at the international level, standards of responsibility and
accountability tend to be defined by prevailing ideological paradigms, hampering
drives to create institutions for global environmental governance based on shared
ethics, justice and equity considerations (Okereke, 2008). It appears that although
there are diverse forms of governance that could be combined into novel hybridised
forms, adaptable and flexible to context-specific needs and changing circumstances,
the prevalent drive appears to denote an incremental change of the status
quo
Responses
In all, a total of 17 engineering adaptation responses to rising sea levels were
modelled, including the existing system of defences and modification to the way
this is operated. For each of the 17 engineering responses, it was also possible to
assess the effect of raising defences. Table 4.1 provides a brief description of the
17 scenarios of responses used in the model.
Model simulations
Model simulations were undertaken using the ISIS one-dimensional hydrodynamic
river modelling software (www.halcrow.com/isis). The models were schematised as
in-bank, meaning water did not spill out onto the floodplain. This meant that the
results for each scenario could be reused to define overtopping thresholds for different
defence levels. Once the simulations were completed, the maximum water levels
from each design event were extracted and stored in a spreadsheet, which was then
used to facilitate analysis. Various overtopping thresholds were then calculated for
each response, both with defences at current levels, and with defences raised by 1 m.
Assumptions
The large number of engineering adaptation responses, combined with the many
potential future extreme sea levels, necessitated that a series of assumptions be
made in order to keep the study to a manageable size. Are there limits to climate prediction?
The accuracy of climate predictions is limited by fundamental, irreducible uncertainties.
Uncertainty means that more than one outcome is consistent with expectations.
For climate prediction, uncertainties can arise from limitations in knowledge
(for example, cloud physics), from randomness (for example, due to the chaotic
nature of the climate system), and also from intentionality, as decisions made by
people can have significant effects on future climate and on future vulnerability (for
example, future greenhouse gas emissions, population, economic growth, development
etc.). Some of these uncertainties can be quantified, but many simply cannot,
meaning that there is some level of irreducible ignorance in our understandings of
future climate (Dessai and Hulme, 2004).
A ‘cascade’ or ‘explosion’ of uncertainty arises when conducting climate change
impact assessments for the purposes of making national and local adaptation decisions
(Jones, 2000). In climate projections used for the development of long-term adaptation strategies, uncertainties from the various levels of the assessment
accumulate. For example, there are uncertainties associated with future emissions
of greenhouse gases and aerosol precursors, uncertainties about the response of the
climate system to these changes (due to structural, parameter and initial conditions
uncertainty) and uncertainties about impact modelling and the spatial and temporal
distributions of impacts. Wilby (2005) has shown that the uncertainty associated
with impact models (in his case a water resources model) arising from the
choice of model calibration period, model structure, and non-uniqueness of model
parameter sets, can be substantial and comparable in magnitude to the uncertainty
in greenhouse gas emissions.
Recent increases in computational power have allowed the partial quantification
of model uncertainty in climate projections using techniques such as perturbedphysics
ensembles (Stainforth et al., 2005), multi-model ensembles (Tebaldi and
Knutti, 2007), statistical emulators (Rougier and Sexton, 2007) and other techniques.
This has partially moved the science from deterministic climate projections
to probabilistic climate projections, but the interpretation of the latter are
much disputed (Stainforth et al., 2007). Most of this work is done with GCMs of
coarse resolution (for example 300–500 km grids), but ensembles of regional climate
model simulations (for example 25–100 km grids) are also being developed
(Murphy et al., 2007, which includes the next set of national UK climate scenarios,
UKCIP 09). Studies that have propagated these various uncertainties for the purposes
of adaptation assessments (sometimes called end-to-end analysis) have found
large uncertainty ranges in climate impacts (Whitehead et al., 2006; Wilby and
Harris, 2006; Dessai and Hulme, 2007; New et al., 2007). They have also found
that the impacts are highly conditional on assumptions made in the assessment,
for example with respect to weightings of GCMs (according to some criteria, such
as performance against past observations) or to the combination of GCMs used.
Some have cautioned that the use of probabilistic climate information may misrepresent
uncertainty and therefore lead to bad a daptation decisions (Hall, 2007). Hall
(2007) warns that improper consideration of the residual uncertainties of probabilistic
climate information (which is always incomplete and conditional) in optimization
exercises, could lead to maladaptation and be far from optimal.
Future prospects for reducing these large uncertainties are limited for several
reasons. Only part of the modelled uncertainty space has been explored up to now
(due to computational expense) so uncertainty in predictions is likely to increase
even as computational power increases. It has proved elusive to find ‘objective’
constraints with which to reduce the uncertainty in predictions (see Allen and
Frame, 2007; Roe and Baker, 2007, in the context of climate sensitivity). The problem
of equifinality (sometimes also called the problem of ‘model identifiability’
or ‘
non-uniqueness’) – that many different model structures and many different parameter sets of a model can produce similar observed behaviour of the system
under study – has rarely been addressed in climate change studies except in some
impact sectors such as water resources (see, for example, Wilby, 2005).
It is also important to recognize that when considering adaptation, climate is only
one of many processes that influence outcomes, sometimes important in certain
decision contexts, other times not (Adger et al., 2007). Many of the other processes
(for example, globalization, economic priorities, regulation, cultural preferences
etc.) are not considered to be amenable to prediction. This raises the question of why
climate should be treated differently, or why accuracy in one element of a complex
and dynamic system would be of benefit given that other important elements are
fundamentally unpredictable. One answer is that we currently live in a society with
a strong emphasis on science- and evidence-based policy-making. This has led
predictive scientific modelling to be elevated above other evidence base because it
can be measured and because of its claimed predictive power