Effects of Climate Change on Hydrology and Hydropower Systems in the Italian Alps

1
I N T R O D U C T I O N
1.1 setting the context
Climate change is considered to be a key factor in the availability
of water resources during the XXI century [IPCC, 2014]. The rise of
temperature and the shift in the distribution of precipitation at the
global scale will affect the hydrological cycle and thus the water re-
lated human activities. The hydrology of the Alpine regions is likely
to be affected more then others since they are characterized by a high
presence of snow and glaciers and are more sensitive to climate con-
ditions [Zierl and Bugmann, 2005; Beniston, 2003]. The temperature
increase will cause an earlier snowmelt and a shift in temporal and
spatial precipitation patterns will considerably change water avail-
ability. Furthermore, the impact of climate change on hydrology will
be accentuated by the glaciers retreat [Haeberli and Beniston, 1998].
The hydropower plants installed in the Alps play a key role in the sup-
ply of electricity and, due to their flexibility compared to other elec-
tricity sources, they also provide a certain stability in the international
network [Gaudard et al., 2014]. Moreover, their importance as a key
resource is growing with the increase of power installed in intermit-
tent renewable energy sources such as wind and solar power. Since
the patterns of production of wind and solar power are very irreg-
ular, and other traditional sources of electricity (e.g., nuclear power
and fossil fuel) are less flexible, hydropower is a strategic source for
the future. Hydropower in a mountainous country like Switzerland
represents 59.7% of electricity generation [Energiebundesamt, 2012],
while in Italy this value decreases to 13.2% [Terna, 2012], mantaining
anyhow a considerable share of the national production. The changes
currently taking place in the electricity market due to the increasing
share of renewable energies and the implementation of an energy
stock exchange are leading to several transformations in which hy-
dropower will be one of the main players.
In such a context, it is important to investigate the complex relation-
ship occurring among climate, water availability, and hydropower
production.
When dealing with climate change, the uncertainty related to future
projection can not be neglected. Climate change impact studies are
the result of a complex modelling chain, which comprises the defi-
nition of the Representative Concentration Pathway (RCP), the Gen-
eral Circulation Model (GCM), Regional Circulation Model (RCM), the
downscaling procedure, the hydrological modelling, and the mod-
1
2 introduction
elling of the reservoirs. In recent years considerable effort has been
spent in order to characterize the uncertainty related to climate and
hydrology. Since an uncertainty analysis can be addressed in several
different ways, a large number of approaches have been proposed in
the past years, to tackle specific aspects of the problem. For example,
Murphy et al. [2004] analyzed the changes in the probability density
functions of some climate indicators, with a probabilistic approach.
Hawkins and Sutton [2009] tried instead to quantify how the different
sources of uncertainty change with the lead time of the projection. An-
other approach was proposed by Finger et al. [2012] that attempted,
through an analysis of the variance, to quantify how different sources
of uncertainty affect the climate during the twelve months of the year.
1.2 objectives of the thesis
The main objective of this thesis is to assess the impact of climate
change on hydrology and hydropower production in the Italian Alps.
In particular, we focus on the Lake Como catchment. We adopt the
classical workflow of climate change impact studies, known in the
literature as “scenario-based”. The first step is the analysis of climate
change scenarios. More precisely, we consider temperature and pre-
cipitation as projected by an ensemble of climate models forced with
two different Representative Concentration Pathways RCPs (4.5 and
8.5). These scenarios refer to the EURO-CORDEX project and the
Intergovernamental Panel for Climate Change (IPCC) Fifth Assess-
ment Report (AR5). The next step is to apply a statistical downscal-
ing, since the spatial resolution of the climate scenarios is not fine
enough for the hydrological modelling. Then, a fully distributed and
physically-based hydrological model is calibrated and simulated. The
importance of employing a spatially distributed hydrological model
is related to the possibility of assessing the response of hydrology to
climate change in every single part of the catchment, allowing spa-
tial analyses on river network, glaciers, and reservoirs. The last step
comprises the assessment of the impact of hydrological scenarios on
the management of the reservoirs that can, thanks to the spatially dis-
tributed hydrological modelling, be jointly taken into account. Along
with this workflow, an uncertainty characterization is carried out in
order to assess the contribution of the single modelling components
to the global uncertainty.
Ultimately, the main innovative contributions of this thesis are the
following:
• The analysis of the predicted impact of climate change on the
Southern Alps and on the Lake Como catchment, within the
EURO-CORDEX framework.
1.3 thesis structure 3
• The employment of a fully distributed hydrological model, in
order to analyze the complexity of the response of hydrology
to climate change, together with the impact on the reservoirs
management.
• The uncertainty characterization, carried out in order to assess,
within the “scenario-based” workflow, where most of the uncer-
tainty is located.
1.3 thesis structure
This thesis is structured in the following parts:
• The next chapter (2) contains a description of the methods and
tools used in the thesis: the climate models, the statistical down-
scaling technique and the hydrological model.
• Chapter 3 provides a comprehensive description of the study
area of the Lake Como catchment.
• Chapter 4 is about the impact of climate change on the study
site. The IPCC AR5 is introduced and the EURO-CORDEX project
with its climate models is described. Then the results of our
analysis on the climate change are shown.
• Chapter5 is intended to describe the adopted downscaling pro-
cedure, starting with the datasets used and concluding with
some comparisons between historical observations and down-
scaled climate scenarios.
• Chapter 6 describes the application of the hydrological model
Topkapi-ETH to the Lake Como catchment. At the beginning a
comprehensive description of the model properties and input
data for setup and calibration is given to the reader. In the sec-
ond part the hydrological scenarios obtained via simulation of
Topkapi-ETH on the case study are analyzed.
• Chapter 7 tackles the issues related to uncertainty. It is shown
how the problem has been approached in the past and which
procedure is adopted in this work to give a quantitative de-
scription of the single modelling component contribution to the
global uncertainty.
2
M E T H O D S A N D T O O L S
2.1 methodology
The general framework used in this thesis to assess the impact of
climate change is usually addressed in literature as "top-down" or
"scenario-based" approach [Wilby and Dessai, 2010]. This approach
consists on the application of a cascade of models, from the demo-
graphic development to the management of a water system. Gener-
ally, in the field of water resources management this modelling cas-
cade includes:
• The definition of a Green House Gases (GHGs) emission sce-
nario.
• The global climate modelling via General Circulation Models
(GCMs).
• The regional climate modelling via Regional Circulation Models
(RCMs).
• The application of statistical downscaling in order to possibly
refine even more the resolution of the climate variables.
• The employment of a hydrological model, to evaluate the stream-
flow scenarios.
• The modelling of the impact on water resources management.
Another possible way to approach climate change impact studies is
the so-called "bottom-up" or "vulnerability-based" approach [Wilby
and Dessai, 2010]. In this different approach the perspective is re-
versed since it relies mainly on the observation of the current water
system and less on the future scenarios. It usually implies the two
following main steps:
• The identification of the current water system vulnerabilities.
• The definition of better strategies to deal with them.
The integration of the two methods is probably the best way to set a
comprehensive analysis and approach policy design in climate change
conditions. Nevertheless, as the main goal of this thesis is to assess
the impact of the climate change on existing hydropower reservoirs,
the first approach ("scenario-based") is applied. The models and tools
applied to the Lake Como catchment to implement the workflow of
5
6 methods and tools
the scenario-based approach are graphically shown in Figure2.1 and
listed here:
• We consider two Representative Concentration Pathways in the
framework of IPCC Fifth Assessment Report, they are RCP4.5 and
RCP8.5.
• We take into account an ensemble of combinations of GCMs and
RCMs to evaluate the effect of climate change on the variables of
temperature and precipitation on the Lake Como catchment.
• As the resolution of RCMs is still to coarse for a physically-based
hydrological model we apply a statistical downscaling using the
Quantile Mapping technique.
• We calibrate a fully distributed and physically-based hydrologi-
cal model (Topkapi-ETH) on the catchment and simulate it, fed
by the downscaled scenarios, in order to assess the impact of
climate change on the hydrology.
• Within Topkapi-ETH we apply an operative rule to the reser-
voirs in the catchment, to evaluate how changes in hydrology
will reflect on the reservoirs.
2.2 models and tools
In the next sections we describe the models and tools used in the the-
sis, namely the RCPs, the GCMs and RCMs, the statistical downscaling
technique and the hydrological model.
2.2.1 Representative Concentration Pathways and IPCC AR5
The Representative Concentration Pathways are Green House Gases
(GHGs) concentration trajectories adopted by the IPCC for its AR5 in
2014. They describe possible climate futures on the basis of the radia-
tive forcing values (changes in balance between incoming and outgo-
ing radiation to the atmosphere, caused by its composition) relative to
the pre-industrial period. RCPs substitute the Special Report on Emis-
sion Scenarios (SRES) projections published in 2000, and used in IPCC
Third Assessment Report (TAR) e Fourth Assessment Report (AR4).
The SRES describes emission scenarios. Emission scenarios are a repre-
sentation of the future discharges in the atmosphere of GHGs that pro-
vide input to climate models. To be produced they require assump-
tions about patterns of economic and demographic growth, technol-
ogy development and future energy consumption. The SRES are com-
plemented by socio-economic storylines, which help in their interpre-
tation. Although they have been widely used, after over a decade of
2.2 models and tools 7
Figure 2.1: Flow chart of the main steps followed in the analysis. On the
left side of the graph, are shown the traditional steps of a
scenario-based workflow in water resources management. On
the right side are listed the sources of the tools used in each
specific step: the RCPs were considered in the framework of
the IPCC AR5; the climate scenarios were retrieved from the
EURO-CORDEX project; the downscaling technique used was
the Quantile Mapping; the hydrological model employed was
Topkapi-ETH, which comprises reservoir operative rules.