Risk, uncertainty and the estimation of the economic impacts of
climate change
Francisco Estrada Porrúa1,2
1Institute for Environmental Studies-VU, Amsterdam, 2Centro de Ciencias de la Atmósfera-UNAM, México
The Economics of Climate Change
The estimation of the potential economic costs of climate change involves a
range of natural and social systems and entails projecting their future states in
the near, medium and long term, under a variety of possible scenarios. As
such, this task is characterized not only by uncertainty from a wide range of
sources and types, but also by ignorance due to the current lack of knowledge
and to the limits of what can be known.
Integrated Assessment Models (IAM) are one of the tools widely used for
studying the economics of climate change (impacts, mitigation and adaptation).
IAM provide a very simplified mathematical representation of some of the
relevant natural and human systems, their processes and interactions. These
models provide a feasible and flexible manner to explore the consequences of
different scenarios and policy measures.
Nevertheless, most of IAM fail to reflect the inherent uncertainty of the problem
they address and that is produced by the oversimplifications they are based on.
To date, the most widely used IAM completely ignore uncertainty, being the
PAGE2002 and PAGE2009 model an exception due to the probabilistic
representation of its parameters, although it assumes deterministic inputs for
which their uncertainty is not accounted. As part of my PhD project a new
stochastic IAM is being developed that takes into account both the aleatory and
epistemic uncertainty in its climate and impact modules.
Probabilistic climate scenarios
Figure 1. Probability of an increase larger than 2ºC emulating 20 of the GCM used in the IPCC’AR4
(IAM-CCA; Estrada, 2011; Gay and Estrada, 2010)
Probabilistic damage functions
A large part of the available probabilistic climate change scenarios as well as
some of the risk measures that are derived from them are based on a
frequentist approach (see, for example the IPCC AR4). This approach can be
seen as an extension of the framework developed for producing weather and
climate forecasts, completely ignoring the conditions for interpreting frequencies
as probabilities and the dominance of epistemic uncertainty (see, for example,
Gay and Estrada, 2010; and Estrada et al., 2012a).
The climate module in the IAM that is being developed offers the possibility of
generating probabilistic climate change scenarios that integrate the range of
possible changes in global temperatures with expert opinion for producing
tailored probabilistic scenarios. From this global temperature scenarios,
regional projections of temperature and precipitation with a spatial resolution of
2.5ºx2.5º are generated, emulating 20 of the IPCC’s AR4 general circulation
models. The regional scenarios are weighted by different probability
distributions reflecting, for example, model performance (Fig. 1).
The basic structure and parameterization of the damage functions in the new
IAM is taken from the PAGE2002 (Hope, 2006), although significant
modifications are being introduced to improve, for example, the temporal and
spatial dynamics of impacts.
The impact functions explicitly address the existence of epistemic uncertainty in
their parameters and are fully stochastic. Uncertainty in this case is represented
by triangular distributions parameterizad to reflect the state of knowledge of the
potential impacts of climate change in both the market and non-market sectors.
Fig 2 shows the mean economic impact as percent of regional GDP lost in year
2100 under the A2 emissions scenario.
Economic growth, vulnerability, resilience, adaptation,
poverty traps and large scale discontinuities
Several relevant processes and their uncertainties are represented very poorly
or are completely missing in current IAM. Vulnerability, expressed in the
parameterizations of the damage functions, is seldom considered uncertain and
is conceived as invariant in time. Adaptation is now included in some IAM but in
a very limited manner and its effects are treated as deterministic.
Economic growth, technology and development in most cases are exogenous
and no uncertainty is considered, despite these are variables on which the
results of the evaluation of climate change impacts heavily relies on. The
existence of poverty traps and other nonlinear responses to shocks are, until
now, not considered.
Some other processes are implicitly included but in an unintended manner, and
in some cases misrepresenting important characteristics of the systems under
analysis. This is the case of the resilience of natural and human systems to
shocks, where most of the current IAM assume that the effects of impacts
dissipate from one period to the next (Estrada et al., 2012b).
Current IAM in most cases are limited to simulating global temperature change,
ignoring other variables such as precipitation and regional changes in climate.
Extreme events and the occurrence of large scale discontinuities are also
poorly represented.
IVM INSTITUTE FOR ENVIRONMENTAL STUDIES
VU University Amsterdam
De Boelelaan 1087
1081 HV Amsterdam
t +31-20-5983 992
f +31-20-5989 553
www.ivm.vu.nl
Info.ivm.vu.nl
Figure 2. Mean economic impact as percent of regional GDP lost in 2100 under the A2 emissions
scenario (IAM-CCA; Estrada, 2011)
References
Estrada F (2011) PhD research project. IVM-VU.
Gay C, Estrada F (2010) Objective probabilities about future climate are a matter of opinion,
Climatic Change, 99 (1-2) 27-46.
Estrada F, Gay C, Conde C (2012a) A methodology for the risk assessment of climate
variability and change under uncertainty. A case study: coffee production in Veracruz, Mexico.
Climatic Change 113(2): 455-479.
Estrada F, Tol RSJ, Gay C (2012) The persistence of shocks in GDP and the estimation of the
potential economic costs of climate change. Working paper.
Hope, C. W. 2006. The Marginal Impact of CO2 from PAGE2002: An Integrated Assessment
Model Incorporating the IPCC’s Five Reasons for Concern. Integrated Assessment Journal,
6(1): 19–56.