Note: Before you go over these slides, make sure you
have watch the these videos first:
1. Climate Change 2013 – The Physical Science Basis
2. Gavin Schmidt – The emergent patterns of
climate change
Three key messages from IPCC WGI AR5:
“…
1. Warming in the climate system is unequivocal
• based on observations at the multiple lines of independent evidence
2. Human influence on the climate system is clear
• Resulting from the combination of model simulations with the
observed climate change
3. Continued GHG emissions cause further climate change and
constitute a multi-century commitment in the future
• Limiting climate change requires substantial and sustain reductions in
GHG emission
…”
IPCC website: [Link]
Climate Models
“… all models are wrong, but some are usefull …”
(Box & Draper 1987)
Challenges for computer models à
• How to describe the earth system?
• How do we predict what our political decisions will be?
• How much energy we’ll use, from which sources?
• How we’ll use the land?
• What new technologies will emerge?
Instead we make climate predictions for a set of
different possible futures.
Models are a simplified representation of
the Earth’s Climate Systems …
A model is a simplified representation of reality, which is
elaborated to facilitate its comprehension and study, that allow
to see in a clear and simple way the different variables and the
relations that are established between them.
Examples:
Fuente: [Link]
Fuente: [Link]
Using the earth as a black box system we think of it as a system in which energy
enters and leaves. The energy that enters is the solar radiation (light) and the
energy that leaves is the infrared radiation (heat) from the earth's surface.
Fuente: [Link]
Using the earth as a white box system, we consider a system in radiative equilibrium
in which a change in any component of the system will requite a change in some
other component to restore that equilibrium.
… with climate models we make
predictions for different scenarios
What we try to answer with
What we would like to know models:
1. What will the world population 1. What would the population
be in the future? be if fertility rates were to stay the
• depends on future fertility same? Or decrease a little, or a lot?”
rates, which are impossible to
predict
2. How would the climate change if
2. How will the climate change? greenhouse gas emissions were to
keep increasing in the same way? Or
decrease a little, or a lot?”
Global Climate Models (GCMs)
• Numerical models representing physical
processes in the atmosphere, ocean,
cryosphere and land surface.
• Useful to simulate the response of global
climate system to increasing GHG
concentrations.
• GCMs use a 3-D grid over the globe with
250-600 km resolution, 10-20 layers in the
atmosphere and ~30 layers in the oceans.
• Sources of uncertainty from GCMs:
• Coarse scale and many processes happen at
smaller scales (method: parameterization)
Regional Climate Models (RCMs) and
• Simulation of feedbacks (water vapour,
downscaling methods (i.e., Delta)
clouds, ocean circulation and ice-snow
GCMs may simulate different responses to
albedo
the same forcing
Global Climate Models (GCMs)
Please take a look at this link for more information
about GCMs:
[Link]
[Link]/guidelines/pages/gcm_guide.html
Scenarios for CO2 emissions:
SRES (Special Report on Emissions Scenarios in 2000)
Extracted from Intergovernmental Panel on Climate Change, Fourth Assessment Report Working Group 1 Figure
10.26
The scenarios describe CO2 and other greenhouse gases, and other industrial emissions
(such as sulphur dioxide) that affect climate. Each scenario has its own “story”. For example:
A1: “rapid and successful economic development, in which regional average income per
capita converge – current distinctions between “poor” and “rich” countries eventually
dissolve”. The A1F1 scenario is the most fossil-fuel intensive.
B1: “a high level of environmental and social consciousness combined with a globally
coherent approach to a more sustainable development”
…climate predictions for each
story
Extracted from Intergovernmental Panel on Climate Change, Fourth Assessment Report Working Group 1 Figure
10.26
• Each broad line is an estimate of a conditional probability
• The probability of a temperature increase, given a particular scenario of
emissions
• Predictions are often called projections that explains “what would happen if…”, not
a “what will happen”.
• useful to distinguish the effects of our possible choices
• There is uncertainty (width of each line) about these effects
“Prediction is very difficult, especially if it’s
about the future.”
(Niels Bohr)
“Trying to predict the future is a mug’s game.”
(Douglas Adams)
Five criteria that should be met by climate scenarios if they are
to be useful for impact researchers and policy makers are
suggested:
• Criterion 1: Consistency with global projections. They should be consistent with a broad range of
global warming projections based on increased concentrations of greenhouse gases. This range
is variously cited as 1.4℃ to 5.8℃ by 2100, or 1.5℃ to 4.5℃ for a doubling of atmospheric CO2
concentration (otherwise known as the "equilibrium climate sensitivity").
• Criterion 2: Physical plausibility. They should be physically plausible; that is, they should not
violate the basic laws of physics. Hence, changes in one region should be physically consistent
with those in another region and globally. In addition, the combination of changes in different
variables (which are often correlated with each other) should be physically consistent.
• Criterion 3: Applicability in impact assessments. They should describe changes in a sufficient
number of variables on a spatial and temporal scale that allows for impact assessment. For
example, impact models may require input data on variables such as precipitation, solar
radiation, temperature, humidity and windspeed at spatial scales ranging from global to site and
at temporal scales ranging from annual means to daily or hourly values.
• Criterion 4: Representative. They should be representative of the potential range of future
regional climate change. Only in this way can a realistic range of possible impacts be estimated.
• Criterion 5: Accessibility. They should be straightforward to obtain, interpret and apply for
impact assessment. Many impact assessment projects include a separate scenario development
component which specifically aims to address this last point. The DDC and this guidance
document are also designed to help meet this need.
Source: [Link]
Weighting the Uncertainty
• Uncertainty in climate change
• Reasons for scientific uncertainty
• Sources and sinks of GHG—Affect future predictions
• Clouds—Affect magnitude of climate change
• Oceans—Affect timing and patterns of climate change
• Polar ice sheets—Affect predictions of sea level rise
• “Even though they may feel uncertain about tomorrow’s weather they
cannot refuse to make a forecast. If they do refuse they withhold from
the public most of the useful information they possess.”
Regarding uncertainty in climate
change projections…
• “Models cannot capture all the factors involved in a
natural system, and those that they do capture are
often incompletely understood.” (Maslin and Austin
2012)
• Climate models can never really be validated (Oreskes
et al. 1994)
Definitions:
Uncertainty: Lack of knowledge of the true value of a variable that can be
described as a probability density function (PDF) characterizing the range and
likelihood of possible values. Uncertainty depends on the analyst’s state of
knowledge, which in turn depends on the quality and quantity of applicable data
as well as knowledge of underlying processes and inference methods.
Probability density function (PDF): The PDF describes the range and relative
likelihood of possible values. The PDF can be used to describe uncertainty in the
estimate of a quantity that is a fixed constant whose value is not exactly known, or
it can be used to describe inherent variability.
Variability: Heterogeneity of a variable over time, space or members of a
population (Morgan and Henrion, 1990; Cullen and Frey, 1999).
Quantitative uncertainty analysis is performed by estimating the 95 percent
confidence interval of the emissions and removals estimates for individual
categories and for the total inventory.
Sources of uncertainty:
Parameters vs. GCMs
GCMs represented the majority
of uncertainty in the analysis
• GCM uncertainty: 66%
deviation from the mean
• Hydrologic model parameter
uncertainty only adds a small
amount to total uncertainty
in results (8% deviation from
the median) but gives
tremendous power for
interpretation and
representation of model
results.
Stationarity
Stationary process is a stochastic process whose
unconditional joint probability distribution does not
change when shifted in time. Consequently, parameters
such as mean and variance, if they are present, also do
not change over time.
Let’s be honest about stationarity
Modelling large regional areas requires efficiencies in
the development of model parameters.
However…
variation in runoff processes are more strongly
driven by climate than by vegetation and soils (Liang
and Guo, 2003; Demaria et al., 2007; van Werkhoven et al.,
2008; Merz et al., 2011)
Stationarity… it gets worse
• interactions between parameters can be more
important than the parameters themselves,
interactions vary across sites and across models
(Rosero et al. 2010)
• substantial uncertainties and biases can result from
calibration of hydrologic models to historical
records, biases appear to grow with time since the
calibration period (Merz et al., 2011)
why and how: talking about uncertainty
S. Santiam
at Cascadia
When predicted change is
consistent for all percentiles, greater
confidence can be placed on
conclusions regarding shifts in
modeled hydrologic responses.
“within the limits of our uncertainty
Thomas Creek
assessment, a 59 percent decrease
was most frequently predicted for
July runoff for the 2080 time period.
However, the decrease was shown
to be as low as 19 percent for a
substantial number of GCM and
model innovations.”
