CC-WARE Mitigating Vulnerability of Water Resources under Climate Change, WP3 - Vulnerability of Water Resources in SEE

Full text

CC-WARE
Mitigating Vulnerability of Water
Resources under Climate Change
WP3 - Vulnerability of Water
Resources in SEE
Report
Version 5.0

2/82 WP3 report v.5 30.6.2014
CC-WARE –
Mitigating Vulnerability of Water Resources under
Climate Change
SEE Project, supported by the means of the ERDF
(European Regional Development Fund) & by the Instrument of Pre-Accession
Assistance (IPA)
Lead Partner:
Austrian Federal Ministry of Agriculture, Forestry, Environment and Water
Management, Forest Department (AT)
Hubert Siegel, Head of Subdivision and Project Coordinator
Contact:
www.ccware.eu
Editor of the WP3 report:
Barbara Čenčur Curk
University of Ljubljana, Faculty of Natural Sciences and Engineering,
Department of Geology
barbara.cencur@ntf.uni-lj.si
Authors of particular chapters:
Part II Chapter 1 Climate and climate change
Sorin Cheval
Part II Chapter 2.1.1 2) Water demand, 3) Local water explotation index
Barbara Čenčur Curk, Petra Vrhovnik, Timotej Verbovšek
Part II Chapter 2.1.1.3.2) LWEI for summer season, 2.1.1.3.3) LWEI for winter
season, 2.1.1.3.4) Annual local water exploitation index corrected for
seasonality
Mathiew Herrnegger, Hans Peter Nachtnebel
Part II Chapter 2.2 Water quality
Prvoslav Marjanović

3/82 WP3 report v.5 30.6.2014
Partners and Persons working in WP3 within the CC-WARE project
LP
Austrian Federal Ministry of Agriculture, Forestry, Environment and Water Management,
Forest Department (AT)
Hubert Siegel, Elisabeth Gerhardt
Associated Organisations:
University of Natural Resources and Life Sciences, Vienna (BOKU), Department of Forest-
and Soil Sciences
Eduard Hochbichler, Roland Koeck
PP1
Municipality of the City of Vienna, MA31 Vienna Waterworks
(AT)
Gerhard Kuschnig
Associated Organisations:
University of Natural Resources and Life Sciences, Vienna (BOKU), Department of Water,
Atmosphere and Environment
Hans-Peter Nachtnebel, Mathew Herrnegger, Tobias Senoner, Johannes Wesemann
University of Natural Resources and Life Sciences, Vienna (BOKU), Department of Forest-
and Soil Sciences
Eduard Hochbichler, Roland Koeck
PP2
Municipality of Waidhofen
van
der Ybbs (AT)
Markus Hochleitner
University of Natural Resources and Life Sciences, Vienna (BOKU), Department of Forest-
and Soil Sciences
Eduard Hochbichler, Roland Koeck
PP3
University of Ljubljana (SI)
Faculty of Natural Sciences and Engineering, Department of Geology
Barbara Čenčur Curk, Timotej Verbovšek, Petra Vrhovnik, Petra Žvab Rožič, Mihael Brenčič,
Nina Zupančič
PP4
Public Water Utility Ljubljana JP Vodovod
-
Kanalizacija d.o.o.
(SI)
Branka Bračič Železnik
PP5
National Institute for
Environment (HU)
László Perger, Agnes Tahy, Enikő Becsákné Tornay
Associated Organisations:
SMARAGD-GSH Ltd.
Zoltan Simonffy, Istvan Bogardi
PP6
National Forest Administration (RO)
Adam Crăciunescu, Ion Codruţ Bîlea, Petrişor Vică
Associated Organisations:
- Petroleum-Gas University of Ploieşti – Faculty of Petroleum Refining and
Petrochemistry, Department: Engineering of Petroleum Processing and Environmental
Protection

4/82 WP3 report v.5 30.6.2014
Ion Onuţu
,
Caşen Panaitescu
- Forest Research And Management Institute Bucureşti
Cristinel Constandache
- Plobil Consulting Ploieşti
Aurel Bilanici
PP7
National Meteorological Administration (RO)
Sorin Cheval, Alexandru Dumitrescu, Madalina Baciu, Traian Breza, Lenuta Marin, Cristina
Draghici, Cerasela Stoica
PP8
Executive Forest Agency (BU)
Albena Bobeva, Lubcho Trichkov, Denitsa Pandeva
Associated Organisations:
National Institute of Meteorology and Hydrology
Valery Spiridonov, Irena Ilcheva, Krasimira Nikolova, Snejanka Balabanova
PP9
Thessaloniki Water Supply & Sewerage Co sa (GR)
Athanasios Soupilas, Spachos Thomas, Kostas Zambetoglou
PP10
Decentralised Administration of Macedonia and Thrace, Water Directorate of Central
Macedonia (GR)
Konstantinos Papatolios, Stelios Michailidis, Charicleia Michalopoloy
Associated Organisations:
Aristotle University of Thessaloniki, Civil Engineering Department
Margaritis Vafeiadis
PP11
Regional Agency for Environmental Protection in the Emilia
-
Romagna region (IT)
Marco Marcaccio, Demetrio Errigo, Donatella Ferri, Franco Zinoni
Associated Organisations:
- University of Modena and Reggio Emilia, Department of Chemical and Geological
Sciences
Alessandro Corsini, Francesco Ronchetti, Margarit Nistor
- University of Bologna, Department of Civil, Chemical, Environmental and Materials
Engineering
Lisa Borgatti, Federico Cervi, Francesca Petronici
IPA1
Jaroslav Cerni Institute for the Development of Water Resources (RS)
Dejan Dimkić, Prvoslav Marjanović,
Branislava Matić, Dragana Pejović,
Vladimir Lukić,
Milutin Stefanović, Dušan Đurić, Marko Marjanović, Miodrag Milovanović, Đulija Boreli-
Zdravković, Goran Mitrović, Nenad Milenković
Associated Organisations:
University of Belgrade, Faculty of Mining and Geology, Department of Hydrogeology
Zoran Stevanović,
Saša Milanović

5/82 WP3 report v.5 30.6.2014
Acknowledgements
We want to express our gratitude to the Lead Partner, to all Project Partners and to the
whole CC-WARE project consortium for the project funding, for the spririt of transnational
cooperation and for the unbending efforts for accomplishing all project goals.
Furthermore we want to express our appreciation to the European Union, to the ERDF
(European Regional Development Fund) and to the IPA (Instrument for Pre-Accession
Assistance) for their support of CC-WARE.

6/82 WP3 report v.5 30.6.2014
TABLE OF CONTENTS
PART I – Description of the Workpackage 3 (WP3) ................................................................ 8
1
WP3 Structure ................................................................................................................. 8
2
Partners Involved in WP3 ................................................................................................ 9
3
WP3 meetings ................................................................................................................. 9
4
WP3 Outputs ................................................................................................................. 10
PART II – Vulnerability of Water Resources in SEE ............................................................... 12
1
Climate and climate change .......................................................................................... 15
1.1
Determination of climate variables and indicators ................................................... 16
1.2
Homogeneous areas .................................................................................................. 18
2
Water resources vulnerability to CC ............................................................................. 25
2.1
Water quantity .......................................................................................................... 25
2.1.1
Water quantity indicators .................................................................................. 25
2.1.1.1
Local total runoff ............................................................................................ 26
2.1.1.2
Water demand ................................................................................................ 27
2.1.1.3
Local water exploitation index (LWEI) ............................................................ 32
2.1.1.3.1
Annual local water exploitation index (LWEI
a
) ............................................... 33
2.1.1.3.2
LWEI for summer season (LWEI
s
) ................................................................... 35
2.1.1.3.3
LWEI for winter season (LWEI
w
) ..................................................................... 38
2.1.1.3.4
Annual Local Water Exploitation index corrected for seasonality (LWEI
asw
) . 40
2.1.1.4
Local Water Surplus (LWS) ............................................................................. 42
2.2
Water quality ............................................................................................................. 44
2.2.1
Water quality indicators ..................................................................................... 44
2.2.2
Present potential pollution load (exposure of water resources to land use
impacts) 45
2.2.3
Future potential pollution load .......................................................................... 49
2.2.4
Surface water quality index (WQI
SW
) ................................................................. 57
2.2.5
Groundwater quality index (WQI
GW
) .................................................................. 58
3
Adaptive capacity .......................................................................................................... 62
3.1
Socio-Economic adaptive capacity factors ................................................................ 62
3.1.1
Water retention and water distribution infrastructuresError! Bookmark not
defined.
3.2
Natural adaptive capacity factors .............................................................................. 63

7/82 WP3 report v.5 30.6.2014
3.2.2
Ecosystem services ............................................................................................. 63
3.2.3
Large groundwater systems acting as reserves .... Error! Bookmark not defined.
4
Integrated assessment of water resources vulnerability to climate change ................ 65
4.1
Integrated vulnerability according to composite programming formula (HU-method)
66
4.1.1
Water Resources Index (WR_HU) ...................................................................... 67
4.1.2
Adaptive Capacity Index (AC_HU) ...................................................................... 68
4.1.3
Integrated vulnerability (IV_HU) ........................................................................ 69
5
Drinking water sources vulnerability to climate change .. Error! Bookmark not defined.
6
References ..................................................................................................................... 72
ANNEX 1 – Handling with water demand data..................................................................... 75
ANNEX 2 – The PRELUDE scenarios for land use changes in the future .............................. 78
Remark: All maps in the report are in small format and are informative. All maps are
available for partners and public also separately as pictures (png) on the CC-WARE project
web page (www.ccware.eu). Project partners can download also GIS files from the CC-WARE
project internal web page.

8/82 WP3 report v.5 30.6.2014
PART I – Description of the Workpackage 3 (WP3)
Concern about the potential effects of climate change on water supply and water demand is
growing. Water resources vulnerability is a critical issue to be faced by society in the near
future. Current variability and future climate change are affecting water supply and demand
over all water-using sectors. Consequently, water scarcity is increasing.
The objective of WP3 is to assess present and future vulnerability of water resources based
on a jointly elaborated methodology. In particular the work package focuses on the
identification of drivers influencing vulnerability, the evaluation of the vulnerability of water
resources as well as the assessment and classification of drinking water risks under climate
change.
Vulnerability of freshwater resources is characterized by several indicators: describing water
availability and increasing demand and the future qualitative state of the system compared
to drinking water standards. Vulnerability indicators (e.g. climatic, hydrologic...) representing
the SEE were selected. Consequently, jointly elaborated maps of water quantity and water
quality vulnerability as well as of integrated vulnerability represent the SEE conditions under
climate change. By classifying the water resources vulnerability, critical areas can be
identified, where water resources stay under risk.
The knowledge of the areal distribution of vulnerable water resources is an important
prerequisite for sustainable management of the relevant areas.
1 WP3 Structure
WP3 is divided into three activities:
ACT 3.1 – Climate change as vulnerability indicator
ACT 3.2 – Evaluation of water quantity and quality vulnerability
ACT 3.3 – Integrated assessment and classification of drinking water risks under CC
The WP3 structure is presented in Table 1. Yellow fields illustrate WP and ACT leader.

9/82 WP3 report v.5 30.6.2014
Table 1: The WP3 structure.
2 Partners Involved in WP3
All partners of the CC-WARE projects were involved in WP3. Partners and persons working
on WP3 are listed in the beginning of the report.
There is an exception in ACT3.3. where PP7 was not invoved.
3 WP3 meetings
For WP3 three meetings were planned in the preparatory phase:
- TWG 3.1: 17 April 2013 in Vienna, AT,
- TWG 3.2: 14 – 15 October 2013 in Beograd, SRB and
- TWG 3.3: 22 January 2014 in Budapest, HU.
Regarding experiences from other projects a need for more joint work always arises.
Discussions regarding WP3 were held also at:
- Plenary Workshop in Thessaloniki, GR (4 - 6 June 2013) and

10/82 WP3 report v.5 30.6.2014
- SC and KT1 in Modena, IT (18 - 20 March 2014) – presentation and discussion of WP3
outputs.
There were several workshops among the WP3 and WP4 leader groups:
- Belgrade, SRB (12 – 13 October 2013),
- Belgrade, SRB (9 – 11 January 2013)
and also two additional workshops for finalizing WP3 with broader working group:
- Radenci, SI (6 – 7 March 2014),
- Radenci, SI (5 – 6 June 2014).
4 WP3 Outputs
Qualitative and quantitative description of the WP3 outputs and results from the Application
Form (AF) are listed in Table 2.
Table 2: Outputs of WP3.
ACTIVITY OUTPUT CONTRIBUTING
PARTNERS
REALIZATION
3.1 Map of homogeneous reference areas in SEE according to the
selected indicators
LP,PP1,3,4,5,6,7,8,
9,10,11, IPA1
Part II
Chapter 1.2
3.2 Common methodology on water quantity and quality
vulnerability mapping (considering climate change and
socio¬economic conditions in the present and future)
LP,PP1,3,4,5,6,7,8,
9,10,11, IPA1
Part II
Chapter 2
3.2 Report on a common methodology for selecting indicators
(climatic, hydrological, geographical and socio-economic) for
water quantity and quality vulnerability determination
transnational SEE vulnerability map (quantity, quality
considering climate change and socio-economic conditions in
present and in the future)
LP,PP1,3,4,5,6,8,9,
10,11, IPA1
Part II
Chapter 2
3.2 Transnational SEE vulnerability map (quantity, quality
considering climate change and socio-economic conditions in
present and in the future)
LP,PP1,3,4,5,6,8,9,
10,11, IPA1
Part II
Chapter 2
3.3 Joint methodology for integrated vulnerability mapping
LP,PP1,3,4,5,6,7,8,
9,10,11, IPA1
Part II
Chapter 4
3.3 Integrated vulnerability map for drinking water resources in SEE
region
LP,PP1,3,4,5,6,7,8,
9,10,11, IPA1
Part II
Chapter 4

11/82 WP3 report v.5 30.6.2014
According to AF ten outputs should be delivered (Table 2). There are two additional
deliverables, which are workshops reports from two additional workshops for finalizing WP3,
which were held in Radenci, SI (6 – 7 March 2014 and 5 – 6 June 2014).
Important remark:
In the AF it was forseen that for each single activity a separate report will be prepared. All
these reports are now incorporated in the final WP3 report, which is presented here. Each
activity and subactivity is presented in separate chapter in PART II of this report.
Maps in the report are informative. All maps are available for partners and public also
separately as pictures (png) in bigger format. Project partners can download also GIS files.

12/82 WP3 report v.5 30.6.2014
PART II – Vulnerability of Water Resources in SEE
Concern about the potential effects of climate change on water supply and water demand is
growing. Water resources vulnerability is a critical issue to be faced by society in the near
future. Current variability and future climate change are affecting water supply and demand
over all water-using sectors. Consequently, water scarcity is increasing.
The objective of WP3 is to assess present and future vulnerability of water resources based
on a jointly elaborated methodology. In particular the work package will focus on the
identification of drivers influencing vulnerability, the evaluation of the vulnerability of water
resouces as well as the assessment and classification of drinking water risks under climate
change.
Vulnerability of freshwater resources is characterised by several indicators: describing water
availability and increasing demand and the future qualitative state of the system compared
to drinking water standards.
Land use may significantly influence the quantity of the water resources, water demand and
overall water quality. A methodology for determining water resources vulnerability
regarding quantity and quality shall take into account also extreme natural events and the
multiple impact of the land use. By classifying the water resources vulnerability, critical areas
can be identified, where water resources stay under risk.
The knowledge of the areal distribution of vulnerable water resources is an important
prerequisite for sustainable management of the relevant areas.
The Intergovernmental Panel on Climate Change (IPCC) describes vulnerability as a function
of impact and adaptive capacity and 'the degree to which a system (water resources) is
susceptible to, or unable to cope with, adverse effects of climate change, including climate
variability and extremes' (IPCC, 2003). 'Vulnerability is a function of the character,
magnitude and rate of climate variation to which a system is exposed, its sensitivity and its
adaptive capacity' (IPCC 2007). The methodology applied in CC-WARE builds on this
description of vulnerability by examining the exposure (predicted changes in the climate),
sensitivity (the responsiveness of a system to climatic influences) and adaptive capacity (the
ability of a system to adjust to climate change) of a range of indicators in a SEE region.
Climatic, hydrological, geological and socio-economic factors influencing vulnerability need
to be identified and appropriate indicators selected.
Exposure, sensitivity, potential impact and adaptive capacity (Figure 1) are all considered in
the evaluation of vulnerability to a defined climate change stressor such as temperature
increases (Local Government Association of South Australia, 2012).
In CC-WARE project impacts of climate, land use and demographic changes on water
resources were analyzed.

13/82 WP3 report v.5 30.6.2014
Figure 1: Components of Vulnerability
Exposure is the change expected in the climate for a range of variables including
temperature and precipitation. Sensitivity is the degree to which systems respond to the
changes. For example less precipitation may reflect in substantial reduction of water
availability in a small river basin or aquifer.
Adaptive capacity describes how well a system can adapt or modify to cope with the climate
changes to which it is exposed to reduce harm. Examples of natural systems with low
adaptive capacity are those with a limited gene pool and as a result a limited capacity to
evolve, over extraction of ground or surface water, salinity or environmental pollutants that
do not have the resilience to adapt. Economic systems that have minimal opportunities to
increase income would also struggle to adapt to climate changes. Social systems that are
disrupted have poor communication networks etc. are also likely to be limited in their
capacity to adapt. When the adaptive capacity of a system is reduced, it is considered to be
more vulnerable to the impacts of climate change. By considering adaptive capacity it is
possible to avoid attending to impacts that may be reduced by the system itself with
minimal outside help, or putting systems that have no capacity to adapt as a low priority
with the result that more harm occurs than expected. (Local Government Association of
South Australia, 2012)
In CC-WARE project ecosystem services and GDP were applied as adaptive capacity
indicators. When the ecosystem services are high (e.g. the ecosystem is in a sound state and
provides a lot of services at low costs) the society saves financial resources while in the
opposite case we find a degraded ecosystem where the society needs large investments to
replace the ecosystem functions by technical measures.
Integrated water resources vulnerability is an overall indicator characterized by set of
indicators refering to water quantity, water quality and adaptive capacity (Figure 1).
INTEGRATED VULNERABILITY
High - low
ADAPTIVE CAPACITY
Economic systems
GDP
E.g. low GDP countries have low adaptation capacity
Natural systems
Ecosystem services
Present state
E.g. Retention of water and pollutants
E.g. over extracted aquifers are limited in adaptation
capacity, salinity or pollutants do not have resilience to adapt
EXPOSURE
CLIMATE CHANGE
Precipitation
Temperature
Actual evapotranspiration
A1B scenario
Data from 3 RCM:
ALADIN, RegCM3, PROMES
WATER RESOURCES
STATUS INDICATORS
WATER QUANTITY
Water exploitation index
WATER QUALITY
Water quality index
POTENTIAL IMPACT
WATER QUALITY
Climate change induced land use
changes reflecting in water quality
Pollutants accumulation in dry periods
WATER QUANTITY
Water scarcity due to reduction of
water availability and increased
demand, above all in summer
WATER RESOURCES
VARIABLES
WATER QUANTITY
Water availability (total runoff)
Water demand
WATER QUALITY
Surface and groundwater pollution
load index due to land use

14/82 WP3 report v.5 30.6.2014
From water resource management perspective, vulnerability can be defined as: the
characteristics of water resources system’s weakness and flaws that make the system
difficult to be functional in the face of socioeconomic and environmental change (UNEP
2009). Thus, the vulnerability should be measured in terms of:
(i) exposure of a water resources system to stressors at the river basin scale; and
(ii) capacity of the ecosystem and society to cope with the threats to the healthy
functionality of a water system (UNEP 2009).
Vulnerability corresponds to changes, which can be compared to a reference situation (e.g.
differences between the past/present and future state). However the determination of the
changes needs the estimation of the present and the future values of the relevant indicators.
Besides, vulnerability cannot be measured, but can be assessed with the help of indicators.
During the CC-WARE workshop discussion it was decided to use “Overlay/index method” for
assessment of vulnerability on a national scale (FOOTPRINT 2006). This method is easier to
understand than the more complex physical based models and therefore more suitable to
use for none-modelers and also more appropriate to enhance the participatory process. To
discriminate between different levels of vulnerability (e.g. three classes low/moderate/high),
it is necessary to combine all quantities into a single measure.

15/82 WP3 report v.5 30.6.2014
1 Climate and climate change
The climate is the main natural driver of the variability in the water resources, and
atmospheric precipitation, air temperature and evapotranspiration are commonly used for
assessing and forecasting the water availability. Generally, the precipitation deficit
associated with high temperature and evapotranspiration values define meteorological,
agricultural and hydrological drought, while the precipitation amounts exceeding the
multiannual averages over an area refill the water resources.
The main objective is to provide climatic indicators relevant for analysing the water
resources vulnerability in the SEE Europe. The data will be available for the activities focused
on assessing the vulnerability of the water resources.
For climate change data results from CC-WaterS project were used. Climate change data
were obtained from three RCMs (RegCM3 – ITCP, Aladin – CNRM, Promes – UCLM), based on
A1B scenario.
The CC-WaterS data base comprises daily and monthly temperature and precipitations
derived from three RCMs, namely RegCM3, ALADIN-Climate and PROMES, extended from
1961 to 2100, at 25-km spatial resolution. RegCM3 is the third generation of the RCM
originally developed at the National Center for Atmospheric Research during the late 1980s
and early 1990s. The model is driven by the GCM ECHAM5-r3, it uses a dynamical
downscaling, and it is nowadays supported by the Abdus Salam International Centre for
Theoretical Physics (ICTP) in Trieste, Italy (Elguindi et al. 2007). ALADIN-Climate was
developed at Centre National de Recherche Meteorologique (CNRM), and it is downscaled
from the ARPEGE-Climate as a driver for the IPCC climate scenarios over the European
domain (Spiridonov et al 2005; Farda et al 2010). PROMES is a mesoscale atmospheric model
developed by MOMAC (MOdelizacion para el Medio Ambiente y el Clima) research group at
the Complutense University of Madrid (UCM) and the University of Castilla-La Mancha
(UCLM) (Castro et al 1993; Gaertner et al 2010), and it is driven by the GCM HADCM3Q0.
The initial simulation results of RegCM3, ADALDIN-Climate and PROMES were available from
the ENSEMBLES project (Hewitt 2004), and they were selected because (1) their spatial
extent covers the full study area of CCWaterS, (2) they provided good performance in the
simulation of historic climate conditions, and (3) each of them uses a different driving GCM.
A1B Scenario
A1B SRES IPCC scenario, which presumes balanced energy sources within a consistent
economic growth, into the context of increasing population until the mid-21st century, and
rapid introduction of more efficient technologies (IPCC TAR WG1 2001).
BIAS Correction
The RCMs outputs were bias corrected using the quantile mapping technique (Déqué 2007;
Formayer and Haas 2010) based on daily observations extracted from the E-OBS data base
v2.0 (CCWaterS 2010). E-OBS is an European 25 km-spatial resolution gridded temperature
and precipitation data set compiled from daily weather station measurements. Their ability
to reproduce the temperature and precipitation was tested both locally (Busuioc et al. 2010)
and at European scale (CCWaterS 2010). The results showed that differences between both
observations and model control runs exist and the results of different RCMs may differ

16/82 WP3 report v.5 30.6.2014
significantly especially in mountainous areas (CC-WaterS 2010). The quantile mapping
technique was used to calibrate each RCM for the control period 1951-2000. The correction
method is based on using the differences of the empirical cumulative density functions (CDF)
of each model and observation data (E-OBS; Haylock et al 2008)) and it is applied to the
model data such that the statistics of the observations are retained. For the scenario period,
the CDFs were calculated for the periods 2001-2025, 2026-2050, 2051-2075 and 2076-2100
and applied in a way, that allows the production of continuous bias corrected time series
from 1951-2100 (1951-2050 for PROMES) (CCWaterS 2010). It has to be stressed that in CC-
WARE the periods analysed were slightly different (1961-1990; 1991-2020; 2021-2050), but
this should not affect the results.
The use of the updated E-OBS data sets (v10.0, released in April 2014) in the project CC-
WARE improved the bias corrected precipitation in some areas (e.g. Northern Carpathians),
while the general pattern remained similar at regional scale.
Ensemble
Considering the objectives of the project, the outputs of the three models were aggregated
for each season by calculating the arithmetic mean for every grid cell.
In CC-WARE project the following time intervals were used:
- 1961-1990 (baseline climate; B);
- 1991-2020 (present climate; P);
- 2021-2050 (future climate; F).
Far future period 2071-2100 was not selected for the CC-WARE study due to large
uncertainties.
1.1 Determination of climate variables and indicators
Main climate variables are:
• precipitation (RR),
• temperature (T) and
• potential and actual evapotranspiration (PET and AET).
Additional climate variables, which were used for the description of climate, are:
• De Martonne’s Index of Aridity
Precipitation (RR) and temperature (T) data were obtained from the ensemble data set
from three RCM models (RegCM3, ALADIN-Climate and PROMES), as described in
introduction to this chapter.
Potential evapotranspiration (PET)
The potential evapotranspiration (PET) is the maximum possible amount of water resulted
from evaporation and transpiration occurring from an area completely and uniformly

17/82 WP3 report v.5 30.6.2014
covered with vegetation, with unlimited water supply without advection and heating
(Dingman 1992; McMahon et al 2013). The potential evapotranspiration is calculated using
the Thornthwaite approach (1974), utilizing solely temperature data of the regional climate
models. We used the R-Package SPEI (Beguería and Vicente-Serrano 2010; Vicente-Serrano
et al. 2010) to calculate the PET using the Thornthwaite's formula (Thornthwaite 1948):


 












(1)
with
PET
m
= monthly potential evapotranspiration [mm];
L = average day length of the month being calculated [h];
N = number of days in the month being calculated [-];



= average monthly temperature [°C]; PET
m
=0 if



<0
I = heat index:
I









(1.1)
α = (6.75·10
-7
)·I
3
- (7.71·10
-5
) ·I
2
+ (1.791·10
-2
)·I + 0.49239 (1.2)
Actual evapotranspiration (AET)
The actual evapotranspiration (AET) is a key component for catchment and water balance
studies, representing the real evapotranspiration occurring over a certain area in a specific
period. The AET was calculated with the Budyko's original equation (Budyko 1974, Gerrits et
al. 2009) according to annual PET and precipitation:





    


    



(2)
where RR
a
denotes mean annual rainfall and φ is Budyko Aridity Index:
 




(2.1)
where PET
a
is annual potential evapotranspiration.
The Budyko framework is frequently applied to assess actual evapotranspiration on a
catchment scale (e.g. Oudin et al., 2008; Roderick et al., 2011; Zhang et al., 2008, 2004,
2001) and has shown satisfactory results. The condition of the application to larger regions is
met in CC-WARE. The spatial scale the method is applied to is defined by the 0.25° grid of the
climate data, resulting in an area of about 625 km² being evaluated. Furthermore long term
annual values of rainfall and potential evapotranspiration are used (1991-2020; 2021-2050)
as a basis. Therefore the precondition, that the storage term within an area can be
neglected, is also considered.
Budyko (1974) considered watersheds with area larger than 1000 km
2
to minimize the
effects of groundwater flows that he assumed to be negligible. Under these conditions he
obtained empirically the Budyko curves by plotting the watershed data and fitting with a
smooth curve. This is a tool to estimate total runoff from such watersheds. In CC-WARE
project AET was calculated for 25 km x 25 km grids (area of 625 km
2
) and it is assumed that
Budyko curves can be applied, since the methodology has been applied also to smaller

18/82 WP3 report v.5 30.6.2014
catchments (Oudin et al. 2008, Zhang et al. 2001, 2004, 2008), where validation using
observed data show reasonable results. Nevertheless we have to be aware that this is an
approximation, since for more precise results Budyko curves have to be modified on the
basis of runoff observations, which are not available for the whole SEE region.
Additional uncertainties of AET results arise because AET is derived from modelled
precipitation data, which were bias corrected with E-OBS data base. In spite of that in some
regions AET show significant errors, which is especially the case for some mountainous
areas. Therefore in these areas results have to be additionally interpreted.
De Martonne’s Index of Aridity
At almost 90 years since its creation, de Martonne Aridity Index (MA) still proves its utility
for evaluating the water availability in an area (Baltas 2007; Maliva and Missimer 2012). The
annual value of the index was calculated by the equation (4) (Doerr 1963), while the
corresponding precipitation amounts and climatic classification can be followed in the Table
3 (Baltas 2007).
 


(3)
where RR [mm] is the annual precipitation and T [°C] the annual mean temperature.
Table 3: De Martonne index aridity classification and corresponding precipitation amounts (Baltas 2007).
Aridity classification MA Precipitation (mm)
Dry < 10.0 < 200.0
Semi-dry 10.0 - 19.9 200.0 - 399.9
Mediterranean 20.0 - 23.9 400.0 - 499.9
Semi-humid 24.0 - 27.9 500.0 - 599.9
Humid 28.0 - 34.9 600.0 - 699.9
Very humid 35.0 - 55.0 700.0 - 800.0
Extremely humid >55.0 >800.0
1.2 Homogeneous areas
For climate variables and indices homogenous regions were elaborated based on grids and
interpolation. Spatial resolution is 0.25
o
, which is approximately 25 km when projected.
Due to many local coordinate projected systems (e.g. Gauss-Krüger D48 used in Slovenia,
another local Gauss-Krueger projected system for Serbia etc.) it was decided to use the most
common geographic system WGS1984. Units of this geographic system are latitude and
longitude degrees. Consequently, cell size of all raster data was fixed to 0.25
o
x 0.25
o
to be
consistent with other raster data and snapping of the raster cells was set in ArcGIS
Environmental settings. For some layers, data was received or calculated in geographic
system ETRS89, using slightly different ellipsoid (GRS80 ellipsoid) than WGS84 system
(WGS84 ellipsoid), but the differences in ellipsoid is less than a millimeter in the polar axis,
leading to maximum half of the meter in projection, and is as such completely negligible for
the purpose of the project data, having cell size of 0.25
o
x 0.25
o
.

19/82 WP3 report v.5 30.6.2014
In ESRI grid data, the first six lines indicate the reference of the grid, followed by the values
listed in "English reading order" (left-right and top-down).
For estimation of impact of climate change on climate variables, relative changes of absolute
values were calculated as:



(4.1)



(4.2)
where Var is climate variable (P, AET, PET) and indexes F mean future (2021 – 2050), P
present (1990 – 2020) and B base period (19961-1990).
Temperature
Differences in the seasonal temperature (
o
C) according to ensemble of RegCM3, ALADIN and
PROMES models between future (2021-2050) and present (1991-2020) period are presented
in Figure 2.
The data retrieved by the ensemble models show that the air temperature will increase in all
the seasons, and in all the regions of the SEE area. Comparing the future and present mean
temperatures, one can remark that the highest temperature increase occur during the
summer, from 0.8 to 1.7°C, with the lowest changes in the Alps and the highest change on
the Balkan Peninsula. The increasing trend is present in all other seasons, but at slightly
lower rates. The most concerning is the temperature rise in winter (up to 1.3°C), which is the
highest in northern part and the Alps, which means less snow in the future and less water
reserves for spring and summer river discharges.

20/82 WP3 report v.5 30.6.2014
Figure 2: Differences in average temperature values (
o
C) between future (2021-2050) and present (1991-
2020) period for fall, winter, spring and summer based on mean ensemble values of RegCM3, ALADIN and
PROMES models.
Annual precipitation
The ensemble precipitation for base (1961-1990), present (1991-2020) and future (2021-
2050) period according to ensemble of RegCM3, ALADIN and PROMES models are presented
in Figure 3. Distribution of precipitation in all periods is generally following the
geomorphological aspects of the area. The lowest precipitation in the project area is in
southeast Italy, the Aegean part of Greece and in Danube delta in Romania. Low
precipitation is in Pannonian basin (Hungary, N Serbia, W Romania), Transylvanian plateau
(Romania), E and S Romania, E Greece, central and E Bulgaria and S Italy. The highest
precipitation is in the Alpine areas of Austria, N Italy and Slovenia. Relatively high
precipitation is also in Carpathians including SW Transylvania (southern Carpathians), Dinaric
area of Serbia and Pindus mountain chain in western Greece.
Relative differences in precipitation between the present (1991-2020) and base (1961-1990)
period (Figure 4) reveal that the analysed region is at the zone between the northern areas
expecting increasing amounts, and the southern ones where the models show a decrease in
precipitation in the next decades. On the other hand relative differences in precipitation

21/82 WP3 report v.5 30.6.2014
between the future (2021-2050) and present (1991-2020) period (Figure 4) show more
scattered pattern, with areas with significant precipitation decrease up to 15%, such as
western and northern Greece and southern Bulgaria. Lower precipitation decrease (5-10%)
can be observed in Po plain and some parts of south-eastern Italy (Gargano Promontory and
Salentino Peninsula).
Figure 3: Annual precipitation amount for baseline (B), present (P) and future (F) period based on mean
annual ensemble values of RegCM3, ALADIN and PROMES models.
Figure 4: Relative changes in annual precipitation amount between present - base period and future -
present period based on mean annual ensemble values of RegCM3, ALADIN and PROMES models.
Potential annual evapotranspiration
Annual potential evapotranspiration (PET) values calculated according to Thornthwaite
formula (see eq. 2) on the basis of T derived by the ensemble of RegCM3, ALADIN and
PROMES models for baseline, present and future period are presented in Figure 5. The
pattern is similar to temperature, because of its calculation. Low PET in the base period
(1961-1990) is in the mountainous regions (Alpine, Carpathian, Dinaric, Balkan, Rhodope
mountains) and high values are in S Italy and SE Greece. In present (1991-2020) period high

22/82 WP3 report v.5 30.6.2014
PET values are also in Po plain (N Italy) and at Romanian Bulgarian border (Romanian plain).
Low PET values are higher than in base period, especially in Pannonian plain (Hungary, N
Serbia and W Romania). In future (2021-2050) period PET is generally higher than in base
and present period, which is due to raise of temperatures in the future. Higher PET values
are in the most part of Italy (except mountains), S and E Romania, SE Bulgaria and Greece.
Relative differences in potential evapotranspiration between the present (1991-2020) and
base (1961-1990) period (Figure 6) show that the relative changes are small (only up to 6 %).
The highest changes in PET are predicted for S Romania (Wallachian plain), central Bulgaria
(Maritsa valley) and E Greece with islands. Relative differences in potential
evapotranspiration between the future (2021-2050) and present (1991-2020) period (Figure
6) show that in the southern part of the SEE area relative changes of PET are predicted to be
up to 9 % (S Italy, entire Greece, S Romania, N and S Bulgaria).
Figure 5: Annual potential evapotranspiration based on mean annual ensemble values of RegCM3, ALADIN
and PROMES models for base, present and future period.
Figure 6: Relative changes of annual potential evapotranspiration between present - base period and
future - present period based on mean annual ensemble values of RegCM3, ALADIN and PROMES models.

23/82 WP3 report v.5 30.6.2014
Annual actual evapotranspiration
Annual actual evapotranspiration values calculated on the basis of PET and precipitation
estimates derived by the ensemble of RegCM3, ALADIN and PROMES models for baseline,
present and future period are presented in Figure 7. High annual AET for all periods occur is
in mid-northern and S Italy, N Austria (Danube valley), SE Hungary, W Serbia, N and SW
Romania and Slovenia and W Greece.
Figure 7: Annual actual evapotranspiration based on mean annual ensemble values of RegCM3, ALADIN
and PROMES models for present and future period.
The present AET pattern will be preserved in the future, but some fluctuations in the
absolute values are estimated (Figures 7 and 8). Relative differences in precipitation
between the present (1991-2020) and base (1961-1990) period (Figure 8) show that the
annual AET will probably increase for up to 7 % (Figure 8) in the northern part of the SEE
area (N Italy, Austria, Slovenia, SE Serbia, N and S Romania). Relative change between the
future (2021-2050) and present (1991-2020) period will be in some areas slightly lower than
between present and base period (S Italy, N Italy, N Austria, E Hungary, NW Bulgaria, and in
some areas higher (up to 10 %): SE Bulgaria and SE Greece (Attica, Central Greece, and N
Peloponnese).

24/82 WP3 report v.5 30.6.2014
Figure 8: Relative changes of annual actual evapotranspiration between present - base period and future -
present period based on mean annual ensemble values of RegCM3, ALADIN and PROMES models.
De Martonne’s Index of Aridity
De
Martonne’s Index of Aridity
(see eq.
3
) based on the
e
nsemble of RegCM3, ALADIN and
PROMES models for baseline, present and future period is presented in Figure 9.
According to the De Martonne’s Index of Aridity extremely humid areas are in the Alps and
some parts of Carpathians. Very humid areas are also in the Apennines, Carpathians and
Dinaric mountains. Humid areas are in W Serbia, Romanian Subcarpathians, N Moldavian
plateau, and Transylvanian Depression and W Greece (Pindus mountains). Semi humid areas
(yellow) are in SE Hungary, W and S Romania and N Bulgaria. Semi-dry areas are in the S
Italy, E Greece and Danube delta in Romania. Dry area is only the most E part of Crete.
Figure 9: De Martonne’s Aridity Index based on mean annual ensemble values of RegCM3, ALADIN and
PROMES models for present and future period.

25/82 WP3 report v.5 30.6.2014
2 Water resources vulnerability to CC
2.1 Water quantity
According to UNEP methodology (2009), vulnerability is a function of water availability, use
and management parameters. The latter will be discussed in WP4. One of the parameters is
water exploitation index (WEI) or water stress, which is the ratio of total water demand
(domestic, industrial and agricultural) to the available amount of renewable water resources
that consists of surface water and groundwater safe yield (river discharge or runoff and
groundwater recharge). Values from 0.2 to 0.4 indicate medium to high stress, whereas
values greater than 0.4 reflect conditions of severe water limitations (Vörösmarty et al.
2000).
Water demand is estimated as water withdrawal by sectors. Future water demand can be
estimated regarding population growth (domestic water use), GDP changes (industrial water
use) and land use changes (agricultural water use). Nevertheless, all these are also subject to
policy. Future water demand will be assessed applying different scenarios. Uncertainty can
be expressed as differences among min, plausible and max values.
2.1.1 Water quantity indicators
Variables and indicators for water quantity sensitivity to CC are presented in Table 4. Water
quantity indicators were calculated for the present (P; 1991-2020) and future (F; 2021-2050)
periods. As climate data results from CC-Waters project were used (see chapter 1). No
indicators were calculated for the baseline time period (1961-1990), since no comprehensive
data sets for land use (CLC), covering the whole SEE-are, exist. Furthermore, after the major
political changes in the 1990’s in the SEE area, some water demand parameters changed
significantly.
Table 4: Variables and indicators for water quantity.
SYMBOL UNITS DATA SOURCES & FORMULAS
VARIABLES
Precipitation RR mm/yr = (l/m
2
)/yr CC-WaterS SEE Project
Actual evapotranspiration AET mm/yr = (l/m
2
)/yr Budyko method
Water demand - total WD mm/yr = (l/m
2
)/yr WD = DWD + AGRWD + INDWD
Water demand - domestic DWD (l/m
2
)/yr EUROSTAT, Partner Countries
Water demand - agriculture AGRWD
(l/m
2
)/yr Partners countries, FAO, EUROSTAT
Water demand - industry INDWD (l/m
2
)/yr EUROSTAT, Partner Countries
INDICATORS
Local Total Runoff LTR mm/yr = (l/m
2
)/yr LTR = RR - AET
Local Water Exploitation
Index
LWEI ND LWEI = WD / LTR
Local Water Surplus LWS mm/yr = (l/m
2
)/yr LWS = LTR – WD

26/82 WP3 report v.5 30.6.2014
Generally all indicators are calculated as long term mean annual values. To account for
uneven seasonal distribution of water demand and water availability, a seasonal water
exploitation index is additionally considered (see chapter 2.1.1.2 and 2.1.1.3).
2.1.1.1 Local total runoff
Water availability was calculated as a simplified water balance:
Q = RR – AET+ΔS (5)
where Q is total runoff (surface and groundwater), RR is precipitation, AET is actual
evapotranspiration and ΔS is a storage change term. Since long term annual values are used,
the storage term ΔS is neglected.
Calculations of total runoff were elaborated based on grids with spatial resolution of 25 km
(0,25
o
). Deficits of the grid by grid calculations exist, since inflowing and outflowing runoff to
and out of the cells is not taken into consideration with this approach. The headwaters and
upper basins as a source for water supply (e.g. from surface water, bank filtration and
regional groundwater systems etc.) are neglected. Basically only direct runoff recharge (from
precipitation) was taken into consideration. Based on these considerations, the indicator
was named LOCAL TOTAL RUNOF (LTR) instead of water availability. Local total runoff is
calculated as:
LTR = RR – AET (6)
Precipitation and actual evapotranspiration input mean values were obtained from selected
RCM’s, which has some bias (see chapter 1).
Figure 10 presents baseline, present and future local total runoff. In all periods it is obvious
that in the Alps total runoff is high, whereas in all other parts it is relatively low, which
means that in those areas there is very low annual recharge.
Differences between the periods are very low, therefore the relative changes of absolute
values of local total runoff (ΔLTR) were calculated (see equations 4.1 and 4.2). With relative
change impact of climate change on local total runoff can be estimated.
Relative changes of LTR between present (1991-2020) and base (1961-1990) period (Figure
11) show that there will be higher LTR (more recharge; up to 16 %) in mountainous areas of
the Alps and Carpathians and in W Pannonian plane (Slovenia, Hungary), N Italy and middle
Serbia. On the other hand in Greece, S Bulgaria, some parts in SE Romania, E Hungary and S
Italy scenarios show that local total runoff would diminish (up to 40 % in Greece).
Relative changes of LTR between future (2021-2050) and present (1991-2020) period (Figure
11) show that there are less areas where LTR would be higher in the future, mostly only in
the Alps, in some areas in Carpathian and central Serbia. Scenario shows that in Italy, NE
Austria, E Hungary, SE Romania, W and E Bulgaria and Greece LTR would be lower in the
future. LTR would diminish up to 36 % in N Greece on the border to FYROM and Bulgaria.
Generally, scenarios show that there would be less recharge and water available in the
future in Italy and Greece (up to 36 %), whereas in other areas the increase of LTR is smaller
(up to 20%). Considering 10-20% uncertainty, all other parts of SEE region are inside this
range. Nevertheless, also small regional changes can influence local water supply.

27/82 WP3 report v.5 30.6.2014
Figure 10: Local total runoff (LTR) based on mean annual ensemble values of RegCM3, ALADIN and
PROMES models for baseline (B), present (P) and future (F) period.
Figure 11: Relative change of Local total runoff (ΔLTR) between present - base period and future - present
period based on mean annual ensemble values of RegCM3, ALADIN and PROMES models.
2.1.1.2 Water demand
Present water demand
Total water demand (WD) was evaluated as the sum of domestic (DWD), agricultural
(AGRWD) and industrial (INDWD) water demand:
WD = DWD + AGRWD + INDWD. (7)

28/82 WP3 report v.5 30.6.2014
All WD data have units m
3
/year; for further calculations these data were transformed to
mm/year (with division by area). Data sets of WD were provided on NUTS 3 level (where
data were available) or on country level for individual countries by the project partner.
Agricultural water demand was not easy to estimate since most of counties do not have geo-
referenced water use data. Moreover it is not easy to get industrial water use data with
separation of water use for hydro power plant and thermal and nuclear PP. Water use for
hydro power plant is in some countries very high, but this water use does not present
significant water loss and should be excluded.
Not all countries have available data on NUTS3 level; in such cases country data was used. In
this case weights were defined for particular WD in order to allocate country water demand
value to NUTS3 level (Table 5). For domestic water demand data weight is population
density (population number for each NUTS 3 respectively). Weight for agricultural water
demand is a percentage of agricultural areas in particular NUTS 3 and for industrial water
demand is a percentage of industrial areas in particular NUTS 3 area (Table 5). Scales of
water demand data and methodology for allocation of country level data to NUTS 3 regions
are presented in Table 6.
In case of Italy, we collected only data for eastern part of a country which belong to SEE
region, all other data were excluded from the further analyses. In case of Republic of Serbia,
which is not involved into EUROSTAT nomenclature system, all data were collected on
municipality level. Thus they also provided shape files for further analyses. In table 5 is
presented an overview of data levels and collected data sets obtained by CC-WARE partner
countries.
Water use data for partner countries as annual values of water use are presented in Table 1
of Appendix 1. CC-WARE water demand data was checked with the comparison of data
adopted by FAO (available at FAO online database) and World Bank. Discrepancies among
data are not big (Table 1 in Appendix 1).
Table 5: Methods for estimation of water demand for different sectors in NUTS 3 scale
Scale of
data sets DWD AGRWD INDWD
COUNTRY

 





















NUTS 3
Domestic water use [m
3
/yr]
for each NUTS 3
Agricultural water use (irrigation)
[m
3
/yr] for each NUTS 3
Industrial water use [m
3
/yr] for
each NUTS 3
Municipality
Domestic water use [m
3
/yr]
for each Municipality
Agricultural water use (irrigation)
[m
3
/yr] for Municipality
Industrial water use [m
3
/yr] for
each Municipality

29/82 WP3 report v.5 30.6.2014
Table 6: Overview of available and collected water demand data by sectors
Country
DWD AGRWD INDWD REMARKS
AT NUTS 3 level NUTS 3 level NUTS 3 level
ITA NUTS 3 level NUTS 3 level * NUTS 3 level
HU NUTS 3 level NUTS 3 level NUTS 3 level
RS Municipalities Municipalities Municipalities
BG country Irrigation by stations Country Country level was used
(WDa-NUTS2)
RO NUTS 3 level (lack of
data) NUTS 3 level (lack of data) NUTS 3 level (lack of
data)
Country level was used
due to lack of data
GR NUTS 3 level NUTS 3 level (lack of data) NUTS 3 level (lack of
data)
Country level was used
due to lack of data
SI country Country (+NUTS 3) Country Country level was used
due to lack of data
*
Data were extracted from NUTS 2 data based on the % of all agricultural lands in corine land cover in each NUTS 3 with respect to the
total all agricultural lands in Corine land cover (CLC) in the NUTS 2. This has some disadvantages: for instance, the sum area at NUTS 2
derived form CLC does not match with the irrigated area value of FAO.
Future water demand
For future water demand four scenarios of water demand changes have been applied:
• 10 % decrease of WD,
• no change in WD,
• 10 % increase of WD,
• 25 % increase of WD.
For calculating water demand in the future, factor ∆WD was introduced:
WD
Z[\[]^
 )DWD _ INDWD _ AGRWD/ ∗ ∆de (8)
where ΔWD is 0.9, 1.0, 1.1 and 1.25 for four water demand scenarios in the future.
Figure 12 presents domestic water demand for present and future scenarios for CC-WARE
countries within SEE area. It can be clearly seen that data was gathered on NUTS 3 level. The
pattern is following the population density. In mountainous regions there is very low
population density therefore domestic water demand is very low, whereas in plains
population density is high and domestic water demand is higher.

30/82 WP3 report v.5 30.6.2014
Figure 12: Domestic water demand (DWD) for present and future scenarios for CC-WARE countries within SEE
area.
Figure 13 presents agricultural water demand for present and future scenarios for CC-WARE
countries within SEE area. There is a very high water demand in the Black sea area in SE
Romania, because there are the main agricultural areas. Higher agricultural is also in Vienna
basin, Po plain, SW and SE Hungary, N Serbia, Greece and in SE Romania (Danube river
delta), where are the main agricultural areas. IN Serbia pattern is very scattered due to the
data scale on Municipality level.
Figure 14 presents industrial water demand for present and future scenarios for CC-WARE
countries within SEE area. High industrial water demand is in the Po plain and the most
southern parts of Italy, in eastern Austria, Slovenia, northern part of Hungary, central Serbia,
and north of Sofia, along Marica river (Plovdiv) and at the Black see (Varna) in Bulgaria.

31/82 WP3 report v.5 30.6.2014
Figure 13: Agricultural water demand (AGRWD) for present and future scenarios for CC-WARE countries
within SEE area.
Figure 14: Industrial water demand (INDWD) for present and future scenarios for CC-WARE countries within
SEE area.

32/82 WP3 report v.5 30.6.2014
Figure 15 presents total water demand for present and future scenarios for CC-WARE
countries within SEE area. High total water demand is in Po plain and SE part in Italy, E
Austria, central Hungary, central Serbia, in Romania around Bucharest and in the Black sea
area and north of Sofia, along Marica river (Plovdiv) and at the Black see (Varna) in Bulgaria
and in the most parts of Greece. In some countries agricultural and industrial water demand
is in different regions, therefore the total water demand is more distributed over a country
(e.g. Hungary).
Figure 15: Water demand for present and future scenarios for CC-WARE countries within SEE area.
2.1.1.3 Local water exploitation index (LWEI)
From WD maps and LTR maps, LOCAL WATER EXPLOITATION INDEX (LWEI) can be calculated
as a ratio between annual WD and LTR for all periods and scenarios, presented in GIS model
in Figure 11:
 


(9)
where LWEI is Local Water Exploitation Index, WD is Water Demand and LTR Local Total
Runoff.

33/82 WP3 report v.5 30.6.2014
The expression ‘local’ in Local water exploitation index is because total runoff was calculated
as direct runoff, not taking into consideration inflowing and outflowing runoff to and out of
the 0.25
o
x0.25
o
grid cell.
2.1.1.3.1 Annual local water exploitation index (LWEI
a
)
Considering annual values and different sectors contributing to water demand Annual Local
Water Exploitation Index (LWEI
a
) is then:














(10)
with
WD
a
... annual water demand [l/m
2
/yr=mm/yr],
LTR
a
... annual local total runoff [mm/yr],
ΔWD ... factor for change of WD in future scenarios (0.9, 1.0, 1.1, 1.25),
DWD ... domestic water demand [l/m
2
/yr=mm/yr],
AGRWD ... agricultural water demand [l/m
2
/yr=mm/yr],
INDWD ... industrial water demand [l/m
2
/yr=mm/yr],
RR
a
... mean annual rainfall [mm/yr],
AET
a
... mean annual actual evapotranspiration [mm/yr].
Local Water Exploitation Index values were classified into five stress classes:
< 0.2 very low water stress
0.2 – 0.4 low water stress
0.4 – 0.6 medium water stress
0.6 – 0.8 high water stress
> 0.8 very high water stress.
Values above 0.4 already signify severe water stress and measures for diminishing of water
stress have to be considered and applied
Figure 16 shows that there is medium water stress central Italy, around Budapest and in NW
Hungary, central Serbia and eastern Greece and very scattered also in the whole Greece.
High and very high water stress on annual level in the SEE region is in the Po plain and
southern Italy, around Vienna in Austria, around Budapest and eastern Balaton area in
Hungary, around Belgrade and in central Serbia, SE Romania at the Black sea, north from
Sofia, around Sofia, in Maritsa plain and at the Black sea in Bulgaria, in NE Greece, around
Athens and in Crete. In the future the pattern will be the same, only areas with medium,
high and very high stress will be larger.
The resulting maps are actually indicators for measures to be applied in a region with high
stress. These measures are discussed together with annual LWEI considering seasonality
(LWEI
asw
; see chapter 2.1.1.3.4).

34/82 WP3 report v.5 30.6.2014
Figure 16: Annual Local Water Exploitation Index (LWEI
a
) for present and future scenarios of water demand
for CC-WARE countries within SEE area.
Assessing the LWEI
a
on an annual basis neglects seasonality and extremes in demand and
availability. These factors are however frequent causes for water scarcity and need to be
addressed. Figure 17 and Figure 18 schematically illustrate this problem.

35/82 WP3 report v.5 30.6.2014
Figure 17: Hypothetical example of monthly water
demand and availability.
Figure 18: Demand to availability ratio of a
hypothetical example
Assessing the LWEI
a
on an annual basis would show no substantial deficits, as the mean
water demand is lower than availability (solid and dashed line in Figure 17). This fact is also
visible in Figure 18, where the annual mean ratio between demand and availability is lower
than 1. The hypothetical example in Figure 18 however shows, that in single months the
demand is higher than the availability, leading to ratios between demand and availability
larger than 1 (Figure 18).
For this reason it was decided to evaluate the LWEI for three different time periods:
(i) annual basis (LWEI
a
),
(ii) summer period (April – September) – LWEI
s
and
(iii) winter (October – March) period – LWEI
w
.
As a basis for further assessments within CC-WARE, the LWEI of the different time periods
was combined to final Local Water Exploitation Index (LWEI
asw
). The methodology for the
assessment of the summer and winter LWEI (LWEI
s
; LWEI
w
) is described in the following
sections. The procedure for estimating actual evapotranspiration for summer and winter
period, which is needed for the water availability term, is described beforehand.
2.1.1.3.2 LWEI for summer season (LWEI
s
)
The Local Water Exploitation Index for summer season (LWEI
s
) is estimated as the ratio
between water demand and availability (total runoff) in summer months. The months of
April to September are thereby included. Similar to the annual LWEI
a
, a multiplicative factor
ΔWD for considering water demand change in future is also used, which is set to 1 for the
recent period (1991-2020). To account for an increase in domestic water demand in summer
months, e.g. due to tourism, a water demand seasonality index (α
sD
) is introduced and
provided by project partners. It is defined as the ratio between domestic water demand in
summer with regard to winter season. The domestic water demand is then:













  



(10)








(11)
Jan Feb Mar Apr May Jun Jul Aug Sep Okt Nov Dec
Demand - Availability [e.g. m³]
Agricultural water demand Domestic water demand
Industrial water demand Mean water demand per month
Mean available water per month Available water resources
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
Jan Feb Mar Apr May Jun Jul Aug Sep Okt Nov Dec
Demand/Availabilty [-]
Demand/Availabilty Annual mean

36/82 WP3 report v.5 30.6.2014



 




 








(12)
with


as domestic water demand seasonality index (a ratio between domestic water demand
in summer months with regard to winter months), DWD
s
domestic water demand in summer
and DWD
w
domestic water demand in winter.
For agricultural water demand it was assumed that the most water for agriculture (irrigation)
is consumed in summer season, therefore annual value of agricultural water demand was
taken into account. For industrial water demand it is assumed that it is the whole year more
or less constant, therefore in summer season industrial water demand is a half of annual
industrial water demand. Total water demand in summer is then:


   





     

(13)
The Summer Local Water Exploitation index (LWEI
s
) is calculated as









(14)
with
LWEI
s
- Water exploitation index for summer season (Apr, May, Jun, Jul, Aug, Sept)
WD
s
- Water demand in summer season
LTR
s
- Local total runoff in summer season; calculated LTR
s
in summer (PP
s
-AET
s
) can be less
than 0, therefore 0.1 mm is set to be the lowest value.
ΔWD - Factor for change of WD in future scenarios (0.90, 1.00, 1.10, 1.25)
DWD - Domestic water demand
AGRWD - Agricultural water demand
INDWD - Industrial water demand.
The water availability (local total runoff) is calculated as the difference between summer
precipitation and AET in summer months:








(15)
with
LTR
s
- Local total runoff in summer season
AET
s
- Mean annual actual evapotranspiration for summer season
RR
s
- Mean summer rainfall
The Budyko formula only estimates mean annual AET values. To estimate summer AET
s
,
annual AET
a
was multiplied with a scaling factor (
sA
). It is the ratio between PET in summer
months and on an annual basis. Furthermore AET
s
was limited to the amount of summer
rainfall, since AET cannot be larger than available summer rainfall. AET
s
for summer months
is calculated as follows:
AETs =min(
AETa ∙
}
sA,
uu
q
)
(16)

37/82 WP3 report v.5 30.6.2014
}
q
=

q

j
(17)
with
}
q
– scaling factor for actual evapotranspiration for summer season
PET
a
- mean annual potential evapotranspiration
PET
s
- mean summer potential evapotranspiration
The approach for estimating summer AET assumes that the ratio between summer and
annual AET is similar to the ratio between summer and annual PET. This approach is feasible,
since the seasonal distribution of AET is similar to (scaled) PET. However water availability
may limit the AET value, which was explicitly considered in the above equation.
Figure 19 shows that there are in the most SEE area only two extreme classes of LWEI
s
for
summer season: either very low or very high stress. A very high water stress in summer
months is in the most SEE region already in the present state (P), except in the Alps, in
Slovenia, western Hungary, western Serbia and in the Carpathian. In the future the pattern
will be the same, but in some areas the water stress is smaller (Romania, W Hungary) and in
other areas higher (in Italy around San Marino and W Bulgaria).
LWEI
s
for summer months present the worst case scenarios regarding water stress, which
are very important in water resources management, since in summer season water demand
is much higher and droughts are more frequent in the last decades.
The resulting maps are actually indicators for measures to be applied in a region with high
stress. These measures are discussed together with annual LWEI considering seasonality
(LWEI
asw
; see chapter 2.1.1.3.4).

38/82 WP3 report v.5 30.6.2014
Figure 19: Summer Local Water Exploitation Index (LWEI
S
) for present and future scenarios of water demand
for CC-WARE countries within SEE area.
2.1.1.3.3 LWEI for winter season (LWEI
w
)
The winter Local Water Exploitation Index (LWEI
w
) for the months October to December and
January to March is calculated in similar manner compared to the summer value:


















(18)
with
LWEI
w
- water exploitation index for winter season (Jan, Feb, Mar, Oct, Nov, Dec)
WD
w
- water demand in winter season
LTR
w
- water availability in winter season
ΔWD - factor for change of WD in future scenarios (0.90, 1.00, 1.10, 1.25)

39/82 WP3 report v.5 30.6.2014
For agricultural winter water demand it is assumed that there is no water consumption (no
irrigation). For industrial water demand it is assumed that it is the whole year more or less
constant, therefore in winter season industrial water demand is a half of annual industrial
water demand. Winter water demand (WD
w
) is then:
with
DWD – domestic water demand
INDWD – industrial water demand


– domestic water demand seasonality index (increase of domestic water demand in
summer months with regard to winter months).
The water availability (local total runoff) is calculated as the difference between winter
precipitation and AET in winter months:









(20)
with
LTR
w
– Local total runoff in winter season
AET
w
- Mean annual actual evapotranspiration for winter season
RR
w
- Mean winter rainfall
Winter AET is calculated as the difference between annual and summer AET:
AETw =
AETa
-
AETs
(21)
Figure 20 shows that LWEI
w
in winter months is very similar to annual LWEI
a
, only that is
slightly lower, which is due to higher water recharge in winter months and lower water
demand (no agricultural water use and smaller domestic water use in touristic areas). None
the less there are some scattered areas with high water stress in the Po plain and southern
Italy, in Vienna, Austria, in Budapest and eastern Balaton area in Hungary, around Belgrade
in Serbia, north from Sofia, in Maritsa plain and at the Black sea in Bulgaria and in SE Greece
around Athens. In the future the pattern will be the same, only areas with medium, high and
very high stress will be slightly larger.
The resulting maps are actually indicators for measures to be applied in a region with high
stress. These measures are discussed together with annual LWEI considering seasonality
(LWEI
asw
; see chapter 2.1.1.3.4).

















(19)

40/82 WP3 report v.5 30.6.2014
Figure 20: Winter Local Water Exploitation Index (LWEI
w
) for present and future scenarios of water demand
for CC-WARE countries within SEE area.
2.1.1.3.4 Annual Local Water Exploitation index corrected for seasonality (LWEI
asw
)
For the further evaluation of water resources in the context of CC-WARE a single annual
value resembling of the water quantity sensitivity is needed. After the intersection of winter
LWEI
w
and summer LWEI
s
to a single seasonal value, a matrix is used to derive the Local
Water Exploitation Index (LWEI
asw
), utilizing the seasonal and annual LWEI
a
values.
To combine the winter and summer LWEI to a seasonal value (LWEI
sw
), the following
procedure is applied, assuming that the more critical value in respect to water exploitation is
relevant:













(22)
Annual water stress (LWEI
a
) was corrected with seasonal water stress (LWEI
sw
) in order to
obtain annual water stress considering seasonality (LWEI
asw
). The method is expert based
classification (Table 7). The classification in Table 7 reflects the fact that higher annual
sensitivity leads to high overall sensitivity values, since the overall water budget is limited.
Higher seasonal values can on the other hand be compensated by lower annual sensitivity

41/82 WP3 report v.5 30.6.2014
values, as technical measure, e.g. dams and reservoirs can enable a seasonal redistribution
of water resources.
Table 7: LWEI
asw
: Annual Local Water Exploitation Index (LWEI
a
) considering seasonality (LWEI
sw
)
Figure 21 shows that there is in the most SEE area high water stress in the present state (P),
except in the Alps, in Slovenia, western Hungary, western Serbia and in the NE Carpathian.
Very high water stress in the SEE region is in the Po plain and southern Italy, around Vienna
in Austria, in Budapest and eastern Balaton area in Hungary, around Belgrade and in central
Serbia, SE Romania at the Black sea, in Bulgaria north from Sofia, around Sofia in Maritsa
plain and at the Black sea, in NE Greece, around Athens and in Crete. In the future the
pattern will be the same; in case of no water demand change there are very slight
differences, whereas in case of 25 % of water demand increase LWEI
asw
there are more
areas with very high stress. Annual LWEI
asw
considering seasonality has similar pattern as
annual LWEI
a
with reflecting summer LWEI
s
.
The resulting maps are actually indicators for measures to be applied in a region with high
water stress. In some cases measures had already been applied some centuries ago: e.g.
Vienna has high water stress (LWEI
asw
), but the problem was solved already century ago,
because drinking water for Vienna is coming from faraway mountains (southwestern from
Vienna). In northeastern Greece, where the water stress is also very high, the water problem
was solved with the use of surface water. In Hungary there is very high water stress around
Budapest; the lack of water was solved with river Danube bank filtration.
very low low medium high very high
[0-0.2] [0.2-0.4] [0.4-0.6] [0.6-0.8] [>0.8]
1 2 3 4 5
very low A A1 A2 A3 A4 A5
low B B1 B2 B3 B4 B5
medium C C1 C2 C3 C4 C5
high D D1 D2 D3 D4 D5
very high E E1 E2 E3 E4 E5
very low low medium high very high
LWEI
asw
LWEI
a
LWEI
sw

42/82 WP3 report v.5 30.6.2014
Figure 21: Annual Local Water Exploitation Index considering seasonality (LWEI
asw
) for present and future
scenarios of water demand for CC-WARE countries within SEE area.
2.1.1.4 Local Water Surplus (LWS)
Annual local surplus of water resources is calculated as the difference of local total runoff
and water demand:
LWS = LTR– WD (23)
Similarly to LWEI, LWS for the future is calculated for all scenarios of Water Demand (no
change, -10 %, +10 %, +25 %).
Annual local surplus of water resources (LWS) for baseline and present period is presented in
Figure 22 and for different water demand scenarios in Figure 23. The most of the CC-WARE
project area has water surplus (positive values). High water surplus is in the Alps, Carpathian,
Dinaric and Pindus mountains. Only some areas have low water deficit: in the Po plain and
southern Italy, around Vienna in Austria, around Budapest and eastern Balaton area in

43/82 WP3 report v.5 30.6.2014
Hungary, around Belgrade and some scattered areas in central Serbia, SE Romania at the
Black sea, in Bulgaria north from Sofia, around Sofia in Maritsa plain, at the Black sea, in NE
Greece, around Athens and in Crete. This is mostly due to higher water demand in those
areas. In the future the pattern will be the same, only water deficit is supposed to be larger.
The resulting maps are indicators for measures to be applied in a region with water deficit.
These measures are discussed together with annual local water exploitation index
considering seasonality (LWEI
asw
; see chapter 2.1.1.3.4).
Figure 22: Annual local surplus of water resources (LWS) for baseline and present for CC-WARE countries
within SEE area.
Figure 23: Annual local surplus of water resources (LWS) for future with different water demand scenarios for
CC-WARE countries within SEE area.

44/82 WP3 report v.5 30.6.2014
2.2 Water quality
Quality problems may occur due to pollution caused by human activities or natural
conditions (geological settings). The indicator “water pollution index” describes the
tendency or likelihood for pollutants to reach water resources.
An important driver (exposure in Figure 1) for water quality vulnerability is land use. CORINE
data base provides information necessary for the evaluation of the existing land use and
estimation of potential pollution load for water resources, which is essential for determining
critical areas and consequently for prioritising activities needed for the sustainable
management of water resources in the SEE area. Applied data set for land use in CC-WARE
project is Corine Land Cover (CLC2006).
2.2.1 Water quality indicators
Main driver for water quality vulnerability is land use. Impact of land use on water quality is
expressed with land use load coefficients (Table 8), which are estimated for each particular
land use (CLC level 3) and present potential for pollution. Pollution load index for surface
water is a sum of particular land use load coefficient multiplied by the particular land use
area (CLC AREA in Table 8). Normalized Pollution load index is indicator for surface water
quality – Water quality index SW (WQI
SW
). Ground water quality indicators are a function of
pollution load and effective infiltration coefficient. Therefore HG factor is introduced. HG
factor is expressed as effective infiltration coefficient, which was determined according to
the International Hydrogeological Map of Europe (BGR & UNESCO 2014). Multiplying Surface
water quality index (WQI
SW
) with HG factor and normalizing we obtain indicator for
groundwater quality - Water quality index GW (WQI
GW
). The methodology for the
assessment of the surface and groundwater quality index is described in the following
sections.
Land use is changing due to climate change and socio-economic factors. Future land use
scenarios (% changes – storylines) were evaluated in accordance with EEA study “Land-use
scenarios for Europe: qualitative and quantitative analysis on a European scale” (EEA 2007).
Table 8: Variables and indicators (red rows) for water quality.
INDICATORS SYMBOL UNITS DATA SOURCES & FORMULAS
Land use load
coefficients
LUSLI
Non dimensional land use load coefficients for particular land
use - literature
Pollution load index PLI Non dimensional Normalized LUSLI
Water quality index SW WQI
SW
Non dimensional (PLI
j
· CLC AREA
i
) and normalized from 0 to 1
HG factor HG Non dimensional HG factor according to IHME map categories
Water quality index GW WQI
GW
Non dimensional (WQI
SW
·
HG) and normalized from 0 to 1

45/82 WP3 report v.5 30.6.2014
2.2.2 Present potential pollution load (exposure of water resources to land use impacts)
The core data set for the calculation of WQI Index is the CORINE land use data set for 2006
(CLC 2006) except for Greece where CORINE 2000 (CLC 2000) is used as 2006 data set is not
available.
For each CORINE land use class at LEVEL 3 an overall water pollution load index is assumed
to be proportional to nutrient export coefficients from a given land use in CORINE.
Nitrogen and Phosphorous export coefficients have been widely used in the assessment of
nonpoint sources of pollution in the past. On the basis of the literature review and expert
knowledge for each CORINE land use class an appropriate Pollution load index (PLI) has been
assigned (see Table 9).
To evaluate this concept the relative ranking after normalization of the assigned Pollution
load Index (LUSLI) is compared to the phosphorous export coefficients from the literature.
Figure 24 shows a plot of the Normalized pollution load index (PLI) and the normalized
phosphorous export coefficients for a given CORINE land use classes from literature. Only
those CORINE Land uses are shown, for which literature data is available. The data used
(Wochna et al. 2011) is shown in Table 10.
Table 9: CORINE Land use and land use load coefficients.
CLC
CODE CLC Description
VERSION 1
Upper range of values
from literature
*Expert interpretation of
literature data
VERSION 2
Lower range of values
from literature *Expert
interpretation of
literature data
*Adopted for CC
WARE Version 2 -
Normalized
between 0 and 1
LUSLI
j
- Relative index of
pollution Load_2006 (or
Nitrogen Export
Coefficients)
LUSLI
j
- Relative index of
pollution Load_2006
PLI
j
-Normalized
Index of pollution
Load_2006
111 Continuous urban fabric 7 6 0.400
112 Discontinuous urban fabric 6.3 5.5 0.367
121 Industrial or commercial units 8 5 0.333
122 Road and rail networks and
associated land 5.5 7.5 0.500
123 Port areas 7 7 0.467
124 Airports 7 7 0.467
131 Mineral extraction sites 9 9 0.600
132 Dump sites 14 14 0.933
133 Construction sites 7 7 0.467
141 Green urban areas 3.5 3.5 0.233
142 Sport and leisure facilities 4 4 0.267
211 Non-irrigated arable land 12 12 0.800
212 Permanently irrigated land 15 15 1.000
213 Rice fields 13.5 13.5 0.900
221 Vineyards 6 6 0.400

46/82 WP3 report v.5 30.6.2014
CLC
CODE CLC Description
VERSION 1
Upper range of values
from literature
*Expert interpretation of
literature data
VERSION 2
Lower range of values
from literature *Expert
interpretation of
literature data
*Adopted for CC
WARE Version 2 -
Normalized
between 0 and 1
LUSLI
j
- Relative index of
pollution Load_2006 (or
Nitrogen Export
Coefficients)
LUSLI
j
- Relative index of
pollution Load_2006
PLI
j
-Normalized
Index of pollution
Load_2006
222 Fruit trees and berry
plantations 5 5 0.333
223 Olive groves 4.5 4.5 0.300
231 Pastures 3.5 3.5 0.233
241 Annual crops associated with
permanent crops 9 9 0.600
242 Complex cultivation patterns 8.3 8.3 0.553
243
Land principally occupied by
agriculture, with significant
areas of natural vegetation
4 5.5 0.367
244 Agro-forestry areas 3 3 0.200
311 Broad-leaved forest 3.6 3.6 0.240
312 Coniferous forest 2.5 2.5 0.167
313 Mixed forest 2.8 2.8 0.187
321 Natural grasslands 2.5 2.5 0.167
322 Moors and heathland 2.7 2.7 0.180
323 Sclerophyllous vegetation 2.5 2.5 0.167
324 Transitional woodland-shrub 2.6 2.6 0.173
331 Beaches, dunes, sands 2.5 2.5 0.167
332 Bare rocks 1.5 1.5 0.100
333 Sparsely vegetated areas 2 2 0.133
334 Burnt areas 5 5 0.333
335 Glaciers and perpetual snow 0.1 0.1 0.007
411 Inland marshes 2.3 2.3 0.153
412 Peat bogs 2.3 2.3 0.153
421 Salt marshes 2.3 2.3 0.153
422 Salines 2.3 2.3 0.153
423 Intertidal flats 3 3 0.200
511 Water courses 3 3 0.200
512 Water bodies 3 3 0.200
521 Cooastal Lagoons 3 3 0.200
522 Estuaries 3 3 0.200
523 Sea and ocean 3 3 0.200

47/82 WP3 report v.5 30.6.2014
Figure 24: Relationship between Normalized Pollution Load Index (PLI) and Normalized Phosphorous export
coefficients for a particular CORINE land use.
Table 10: Relationship between assigned values of land use load coefficients and literature data on
phosphorous export coefficients (Wochna et al. 2011).
CLC Land use CLC CODE
Values from
different sources
and expert
judgment
Values from
literature. all
values single
source
Normalized TN Normalized TP
TN Export
Coefficient
TP Export
Coefficient Normalized TN Normalized TP
Continuous urban fabric 111 5 1.2 0.417 0.246
Industrial or commercial
units 121 6 2.5 0.500 0.512
Road and rail networks
and associated land 122 5.5 1.2 0.458 0.246
Port areas 123 7 2.5 0.583 0.512
Airports 124 7 2.5 0.583 0.512
Construction sites 133 7 2.5 0.583 0.512
Green urban areas 141 3.5 0.83 0.292 0.170
Sport and leisure
facilities 142 4 1.2 0.333 0.246
Non-irrigated arable
land 211 12 4.88 1.000 1.000
Pastures 231 3.5 0.83 0.292 0.170
Complex cultivation
patterns 242 8.3 2.33 0.692 0.477
Land principally
occupied by agriculture.
with significant areas of
natural vegetation
243 4 0.49 0.333 0.100
Broad-leaved forest 311 3.6 0.26 0.300 0.053

48/82 WP3 report v.5 30.6.2014
CLC Land use CLC CODE
Values from
different sources
and expert
judgment
Values from
literature. all
values single
source
Normalized TN Normalized TP
TN Export
Coefficient
TP Export
Coefficient Normalized TN Normalized TP
Coniferous forest 312 2.5 0.36 0.208 0.074
Mixed forest 313 2.8 0.26 0.233 0.053
Natural grasslands 321 2.5 0.62 0.208 0.127
Moors and heathland 322 2.7 0.13 0.225 0.027
Transitional woodland-
shrub 324 2.6 0.26 0.217 0.053
Beaches, dunes, sands 331 2.5 0 0.208 -
Inland marshes 411 2.3 0.23 0.192 0.047
Peat bogs 412 2.3 0.23 0.192 0.047
Water courses 511 3 0.5 0.250 0.102
For those CORINE Land uses for which literature data is not available, expert judgment
assignment of appropriate values was used. The complete data set used for the Normalized
Index of pollution Load_2006 (PLI) is shown in Table 11. Surface water quality index (WQI
SW
)
map for the baseline year 2006 is obtained with applying of the Normalized Index of
pollution Load_2006 (PLI) to CLC 2006 (level 3) map with multiplying PLI by the belonging
CLC 2006 area (see Table 8) and then normalizing form 0 to 1.
Table 11: Adopted values for the PLI for the baseline year 2006.
CLC CODE CLC Description
LUSLI
j
Relative index of
pollution Load_2006
PLI
j
Normalized
Index of
pollution
Load_2006
111 Continuous urban fabric 6 0.400
112 Discontinuous urban fabric 5.5 0.367
121 Industrial or commercial units 5 0.333
122 Road and rail networks and associated land 7.5 0.500
123 Port areas 7 0.467
124 Airports 7 0.467
131 Mineral extraction sites 9 0.600
132 Dump sites 14 0.933
133 Construction sites 7 0.467
141 Green urban areas 3.5 0.233
142 Sport and leisure facilities 4 0.267
211 Non-irrigated arable land 12 0.800
212 Permanently irrigated land 15 1.000
213 Rice fields 13.5 0.900
221 Vineyards 6 0.400

49/82 WP3 report v.5 30.6.2014
CLC CODE CLC Description
LUSLI
j
Relative index of
pollution Load_2006
PLI
j
Normalized
Index of
pollution
Load_2006
222 Fruit trees and berry plantations 5 0.333
223 Olive groves 4.5 0.300
231 Pastures 3.5 0.233
241 Annual crops associated with permanent crops 9 0.600
242 Complex cultivation patterns 8.3 0.553
243 Land principally occupied by agriculture. with significant areas
of natural vegetation 5.5 0.367
244 Agro-forestry areas 3 0.200
311 Broad-leaved forest 3.6 0.240
312 Coniferous forest 2.5 0.167
313 Mixed forest 2.8 0.187
321 Natural grasslands 2.5 0.167
322 Moors and heathland 2.7 0.180
323 Sclerophyllous vegetation 2.5 0.167
324 Transitional woodland-shrub 2.6 0.173
331 Beaches, dunes, sands 2.5 0.167
332 Bare rocks 1.5 0.100
333 Sparsely vegetated areas 2 0.133
334 Burnt areas 5 0.333
335 Glaciers and perpetual snow 0.1 0.007
411 Inland marshes 2.3 0.153
412 Peat bogs 2.3 0.153
421 Salt marshes 2.3 0.153
422 Salines 2.3 0.153
423 Intertidal flats 3 0.200
511 Water courses 3 0.200
512 Water bodies 3 0.200
521 Coastal Lagoons 3 0.200
522 Estuaries 3 0.200
523 Sea and ocean 3 0.200
2.2.3 Future potential pollution load
The methodology for the evaluation of future WQI follows the methodology used for the
present evaluation. Again the basis for the analysis is the land use.
Future Land Use scenarios for EUROPE were adopted from the EEA study on qualitative and
quantitative analysis of Land use scenarios for Europe (EEA 2007). However, this study does
not cover the whole CC-WARE project area (data for Serbia is missing) and in order to
overcome this problem it was decided to use the scenarios from the above study but to
apply them to a CORINE data set that is available for the whole CC-WARE project area. Since
land use categories from the EEA (2007; table 12) are not the same as CORINE land use

50/82 WP3 report v.5 30.6.2014
categories a corespondance table (table 13) for land use classification between CORINE and
EEA had to be prepared.
Table 12: EEA classification of landscape types (EEA 2007).
LAND USE ID
Landscape type according to EEA 2007 Landscape characteristics
(based on land cover classes in Table 11)
001 Urban landscapes Urban land use is dominant. All other land cover classes
are not dominant
002 Landscapes with urban character Urban land use is dominant but any other land use could
be dominant as well
003 Landscapes with agricultural character Cropland is dominant. any other land use is dominant
004 Rural landscapes with grassland dominance Grassland is dominant. any other land use is not dominant
005 Rural mosaic landscapes with agricultural
character
With majority of agricultural land. i.e. cropland and
grassland > 50% of model cell area
006 Rural mosaic landscapes with semi-natural
character
With majority of semi-natural land. i.e. other land. surplus
land and forest > 50% of model cell area
007 Landscapes with semi-natural to natural
character Other land category is dominant
008 Semi-natural landscapes with abandoned
character
Other land category in combination with surplus land is
dominant
009 Semi-natural landscapes with grassland
character
Grassland in combination with other land and surplus land
is dominant
010 Forest Landscapes Forest is dominant
Table 13: Comparison of CORINE and EEA (2007) land use types.
CORINE
CODE j CORINE LAND COVER - CLC_3 j CORRESPONDING EEA 2007 LAND USE
111 Continuous urban fabric Urban landscapes
112 Discontinuous urban fabric Urban landscapes
121 Industrial or commercial units Landscapes with urban character
122 Road and rail networks and associated land Landscapes with urban character
123 Port areas Landscapes with urban character
124 Airports Landscapes with urban character
131 Mineral extraction sites Semi-natural landscapes with abandoned character
132 Dump sites Semi-natural landscapes with abandoned character
133 Construction sites Semi-natural landscapes with abandoned character
141 Green urban areas Landscapes with semi-natural to natural character
142 Sport and leisure facilities Landscapes with semi-natural to natural character
211 Non-irrigated arable land Landscapes with agricultural character
212 Permanently irrigated land Landscapes with agricultural character
213 Rice fields Landscapes with agricultural character
221 Vineyards Landscapes with agricultural character
222 Fruit trees and berry plantations Landscapes with agricultural character

51/82 WP3 report v.5 30.6.2014
CORINE
CODE j CORINE LAND COVER - CLC_3 j CORRESPONDING EEA 2007 LAND USE
223 Olive groves Landscapes with agricultural character
231 Pastures Semi-natural landscapes with grassland character
241 Annual crops associated with permanent crops Rural mosaic landscapes with agricultural character
242 Complex cultivation patterns Rural mosaic landscapes with agricultural character
243 Land principally occupied by agriculture. with
significant areas of natural vegetation
Rural mosaic landscapes with agricultural character
244 Agro-forestry areas Rural mosaic landscapes with agricultural character
311 Broad-leaved forest Forest landscapes
312 Coniferous forest Forest landscapes
313 Mixed forest Forest landscapes
321 Natural grasslands Forest landscapes
322 Moors and heathland Forest landscapes
323 Sclerophyllous vegetation Forest landscapes
324 Transitional woodland-shrub Forest landscapes
331 Beaches, dunes, sands Landscapes with semi-natural to natural character
332 Bare rocks Landscapes with semi-natural to natural character
333 Sparsely vegetated areas Landscapes with semi-natural to natural character
334 Burnt areas Semi-natural landscapes with abandoned character
335 Glaciers and perpetual snow No corresponding category
411 Inland marshes Landscapes with semi-natural to natural character
412 Peat bogs Landscapes with semi-natural to natural character
421 Salt marshes Landscapes with semi-natural to natural character
422 Salines Landscapes with semi-natural to natural character
423 Intertidal flats Landscapes with semi-natural to natural character
511 Water courses No corresponding category
512 Water bodies No corresponding category
521 Coastal lagoons No corresponding category
522 Estuaries No corresponding category
523 Sea and ocean No corresponding category
In EEA (2007) study five PRELUDE land use change scenarios (table 14) were developed (see
Annex 2 for storylines descriptions):
- Scenario 1: Great Escape - Europe of contrast
- Scenario 2: Evolved Society - Europe of harmony
- Scenario 3: Clustered Networks - Europe of structure
- Scenario 4: Lettuce Surprise U - Europe of innovation
- Scenario 5: Big Crisis - Europe of cohesion.

52/82 WP3 report v.5 30.6.2014
For each scenario a separate future land use map was elaborated. For each land use type it
was calculated the percentage regarding to the CC-WARE project area. This was done for all
five EEA scenarios of land use change.
Table 14: Percentages of major landscape types in the five scenarios in 2035 and the base year in 2005.
Base Year
The Great
Escape
Evolved
Society
Clusters of
European
Networks
Lettuce
Surprise U
After the
Big Crisis
Urban landscapes 3.0 4.2 3.2 3.0 5.2 4.3
Landscapes with urban character 3.2 2.2 3.1 3.4 1.1 2.0
Landscapes with agricultural character
24.4 7.2 24.1 15.5 0.8 17.1
Rural landscapes with grassland
dominance 12.9 9.1 10.9 9.2 9.9 10.6
Rural mosaic landscapes with
agricultural character 10.3 4.5 10.2 5.4 16.5 13.1
Rural mosaic landscapes with semi-
natural character 13.7 25.2 15.7 14.0 30.0 19.3
Landscapes with semi-natural to
natural character 8.6 8.4 8.6 8.9 8.8 8.6
Semi-natural landscapes with
abandoned character 0.0 15.6 0.4 18.8 3.9 1.1
Semi-natural landscapes with
grassland character 0.6 0.5 0.6 0.5 0.7 0.7
Forest landscapes 23.0 23.1 23.2 23.1 23.2 23.2
The data from the table 14 is used to compute the % change in land use for a given CORINE
land use category in the correspondance Table 13. The result scenarios for CORINE land use
changes for 2050 are given in Table 15.
To calculate the PLI for the future under different scenarios for the future we use the land
use % change for each scenario:


 

 


(24)
where:
S is the modifying coefficient for PLI of future land use change scenarios and is calculated as:



   


 (25)
j identifies the CORINE land use class and
k identifies the future Scenario % land use change.



= corresponding values from Table 15.
The values of 


are given in the Table 16.

53/82 WP3 report v.5 30.6.2014
Table 15: CORINE Land Use Scenarios for 2050 expressed in percentage in land use for a given CLC category
(CC-WARE Project).
CORINE
CODE
j
CORINE LEVEL 3
j



% of land use change for a given scenario
(% of total land area)
The Great
Escape_S1
k=1
Evolved
Society_S2
k=2
Clusters of
European
Networks_S3
k=3
Lettuce
Surprise U_S4
k=4
After the
Big
Crisis_S5
k=5
111 Continuous urban fabric 1.20 0.20 0.00 2.20 1.30
112 Discontinuous urban fabric -1 -0.1 0.2 -2.1 -1.2
121 Industrial or commercial units -1 -0.1 0.2 -2.1 -1.2
122 Road and rail networks and
associated land -1 -0.1 0.2 -1.2
123 Port areas -1 -0.1 0.2 -2.1 0.22
124 Airports -1 -0.1 0.2 -2.1 0.22
131 Mineral extraction sites 15.6 0.4 18.8 3.9 1.1
132 Dump sites 15.6 0.4 18.8 3.9 1.1
133 Construction sites 15.6 0.4 18.8 3.9 1.1
141 Green urban areas -0.2 0 0.3 0.2 0
142 Sport and leisure facilities -0.1 0 -0.1 0.1 0.1
211 Non-irrigated arable land -17.2 -0.3 -8.9 -23.6 -7.3
212 Permanently irrigated land -17.2 -0.3 -8.9 -23.6 -7.3
213 Rice fields -17.2 -0.3 -8.9 -23.6 -7.3
221 Vineyards -17.2 -0.3 -8.9 -23.6 -7.3
222 Fruit trees and berry
plantations -17.2 -0.3 -8.9 -23.6 -7.3
223 Olive groves -17.2 -0.3 -8.9 -23.6 -7.3
231 Pastures -3.8 -2 -3.7 -3 -2.3
241 Annual crops associated with
permanent crops -5.8 -0.1 -4.9 6.2 2.8
242 Complex cultivation patterns -5.8 -0.1 -4.9 6.2 2.8
243
Land principally occupied by
agriculture. with significant
areas of natural vegetation
-5.8 -0.1 -4.9 6.2 2.8
244 Agro-forestry areas -5.8 -0.1 -4.9 6.2 2.8
311 Broad-leaved forest 0.1 0.2 0.1 0.2 0.2
312 Coniferous forest 0.1 0.2 0.1 0.2 0.2
313 Mixed forest 0.1 0.2 0.1 0.2 0.2
321 Natural grasslands -3.8 -2 -3.7 -3 -2.3
322 Moors and heathland 11.5 2 0.3 16.3 5.6
323 Sclerophyllous vegetation 11.5 2 0.3 16.3 5.6
324 Transitional woodland-shrub 0.1 0.2 0.1 0.2 0.2
331 Beaches, dunes, sands -0.2 0 0.3 0.2 0

54/82 WP3 report v.5 30.6.2014
CORINE
CODE
j
CORINE LEVEL 3
j



% of land use change for a given scenario
(% of total land area)
The Great
Escape_S1
k=1
Evolved
Society_S2
k=2
Clusters of
European
Networks_S3
k=3
Lettuce
Surprise U_S4
k=4
After the
Big
Crisis_S5
k=5
332 Bare rocks -0.2 0 0.3 0.2 0
333 Sparsely vegetated areas -0.2 0 0.3 0.2 0
334 Burnt areas 15.6 0.4 18.8 3.9 1.1
335 Glaciers and perpetual snow 0 0 0 0 0
411 Inland marshes -0.2 0 0.3 0.2 0
412 Peat bogs -0.2 0 0.3 0.2 0
421 Salt marshes -0.2 0 0.3 0.2 0
422 Salines -0.2 0 0.3 0.2 0
423 Intertidal flats -0.2 0 0.3 0.2 0
511 Water courses 0 0 0 0 0
512 Water bodies 0 0 0 0 0
521 Coastal lagoons 0 0 0 0 0
522 Estuaries 0 0 0 0 0
523 Sea and ocean 0 0 0 0 0
Table 16: The values of 


CLC
CODE CLC Description
Correction factor for future land use scenarios



(Values derived from EEA Land Use Scenarios for Europe)
s1 (k=1) s2 (k=2) s3 (k=3) s4 (k=4) s5 (k=5)
111 Continuous urban fabric 1.01 1.001 0.998 1.021 1.012
112 Discontinuous urban fabric 1.01 1.001 0.998 1.021 1.012
121 Industrial or commercial units 1.01 1.001 0.998 1.021 1.012
122 Road and rail networks and
associated land 1.01 1.001 0.998 1.021 1.012
123 Port areas 0.969 0.9992 0.9624 0.9922 0.9978
124 Airports 0.9688 0.9992 0.9624 0.9922 0.9978
131 Mineral extraction sites 0.9688 0.9992 0.9624 0.9922 0.9978
132 Dump sites 0.9688 0.9992 0.9624 0.9922 0.9978
133 Construction sites 0.9688 0.9992 0.9624 0.9922 0.9978
141 Green urban areas 0.988 0.998 1 0.978 0.987
142 Sport and leisure facilities 0.988 0.998 1 0.978 0.987
211 Non-irrigated arable land 1.172 1.003 1.089 1.236 1.073
212 Permanently irrigated land 0.885 0.98 0.997 0.837 0.944
213 Rice fields 1 1 1 1 1
221 Vineyards 1.002 1 0.997 0.998 1

55/82 WP3 report v.5 30.6.2014
CLC
CODE CLC Description
Correction factor for future land use scenarios



(Values derived from EEA Land Use Scenarios for Europe)
s1 (k=1) s2 (k=2) s3 (k=3) s4 (k=4) s5 (k=5)
222 Fruit trees and berry plantations 1.002 1 0.997 0.998 1
223 Olive groves 1.002 1 0.997 0.998 1
231 Pastures 1.038 1.02 1.037 1.03 1.023
241 Annual crops associated with
permanent crops 1.172 1.003 1.089 1.236 1.073
242 Complex cultivation patterns 1.058 1.001 1.049 0.938 0.972
243
Land principally occupied by
agriculture. with significant
areas of natural vegetation
1.172 1.003 1.089 1.236 1.073
244 Agro-forestry areas 0.999 0.998 0.999 0.998 0.998
311 Broad-leaved forest 0.999 0.998 0.999 0.998 0.998
312 Coniferous forest 0.999 0.998 0.999 0.998 0.998
313 Mixed forest 0.999 0.998 0.999 0.998 0.998
321 Natural grasslands 1.001 1 1.001 0.999 0.999
322 Moors and heathland 1 1 1 1 1
323 Sclerophyllous vegetation 0.885 0.98 0.997 0.837 0.944
324 Transitional woodland-shrub 0.885 0.98 0.997 0.837 0.944
331 Beaches, dunes, sands 1 1 1 1 1
332 Bare rocks 1 1 1 1 1
333 Sparsely vegetated areas 1 1 1 1 1
334 Burnt areas 1 1 1 1 1
335 Glaciers and perpetual snow 1 1 1 1 1
411 Inland marshes 1 1 1 1 1
412 Peat bogs 1 1 1 1 1
421 Salt marshes 1 1 1 1 1
422 Salines 1 1 1 1 1
423 Intertidal flats 1 1 1 1 1
511 Water courses 1 1 1 1 1
512 Water bodies 1 1 1 1 1
521 Coastal Lagoons 1 1 1 1 1
522 Estuaries 1 1 1 1 1
523 Sea and ocean 1 1 1 1 1
With the application of correction factors we get pollution load index PLI for five PRELUDE
future scenarios (Table 17).

56/82 WP3 report v.5 30.6.2014
Table 17: PLI
SW
values for five PRELUDE future scenarios of land use change.
CLC
CODE CLC Description
FUTURE VALUE OF PLI
SW
PLIj_2050_
S1
PLIj_2050_
S2
PLIj_2050_
S3
PLIj_2050_
S4
PLIj_2050_
S5
111 Continuous urban fabric 0.404 0.400 0.399 0.408 0.405
112 Discontinuous urban fabric 0.370 0.367 0.366 0.374 0.371
121 Industrial or commercial units 0.337 0.334 0.333 0.340 0.337
122 Road and rail networks and
associated land 0.505 0.501 0.499 0.511 0.506
123 Port areas 0.452 0.466 0.449 0.463 0.466
124 Airports 0.452 0.466 0.449 0.463 0.466
131 Mineral extraction sites 0.581 0.600 0.577 0.595 0.599
132 Dump sites 0.904 0.933 0.898 0.926 0.931
133 Construction sites 0.452 0.466 0.449 0.463 0.466
141 Green urban areas 0.231 0.233 0.233 0.228 0.230
142 Sport and leisure facilities 0.263 0.266 0.267 0.261 0.263
211 Non-irrigated arable land 0.938 0.802 0.871 0.989 0.858
212 Permanently irrigated land 0.885 0.980 0.997 0.837 0.944
213 Rice fields 0.900 0.900 0.900 0.900 0.900
221 Vineyards 0.401 0.400 0.399 0.399 0.400
222 Fruit trees and berry plantations 0.334 0.333 0.332 0.333 0.333
223 Olive groves 0.301 0.300 0.299 0.299 0.300
231 Pastures 0.242 0.238 0.242 0.240 0.239
241 Annual crops associated with
permanent crops 0.703 0.602 0.653 0.742 0.644
242 Complex cultivation patterns 0.585 0.554 0.580 0.519 0.538
243
Land principally occupied by
agriculture, with significant areas of
natural vegetation
0.430 0.368 0.399 0.453 0.393
244 Agro-forestry areas 0.200 0.200 0.200 0.200 0.200
311 Broad-leaved forest 0.240 0.240 0.240 0.240 0.240
312 Coniferous forest 0.167 0.166 0.167 0.166 0.166
313 Mixed forest 0.186 0.186 0.186 0.186 0.186
321 Natural grasslands 0.167 0.167 0.167 0.167 0.167
322 Moors and heathland 0.180 0.180 0.180 0.180 0.180
323 Sclerophyllous vegetation 0.148 0.163 0.166 0.140 0.157
324 Transitional woodland-shrub 0.153 0.170 0.173 0.145 0.164
331 Beaches, dunes, sands 0.167 0.167 0.167 0.167 0.167
332 Bare rocks 0.100 0.100 0.100 0.100 0.100
333 Sparsely vegetated areas 0.133 0.133 0.133 0.133 0.133
334 Burnt areas 0.333 0.333 0.333 0.333 0.333
335 Glaciers and perpetual snow 0.007 0.007 0.007 0.007 0.007

57/82 WP3 report v.5 30.6.2014
CLC
CODE CLC Description
FUTURE VALUE OF PLI
SW
PLIj_2050_
S1
PLIj_2050_
S2
PLIj_2050_
S3
PLIj_2050_
S4
PLIj_2050_
S5
411 Inland marshes 0.153 0.153 0.153 0.153 0.153
412 Peat bogs 0.153 0.153 0.153 0.153 0.153
421 Salt marshes 0.153 0.153 0.153 0.153 0.153
422 Salines 0.153 0.153 0.153 0.153 0.153
423 Intertidal flats 0.200 0.200 0.200 0.200 0.200
511 Water courses 0.200 0.200 0.200 0.200 0.200
512 Water bodies 0.200 0.200 0.200 0.200 0.200
521 Coastal Lagoons 0.200 0.200 0.200 0.200 0.200
522 Estuaries 0.200 0.200 0.200 0.200 0.200
523 Sea and ocean 0.200 0.200 0.200 0.200 0.200
2.2.4 Surface water quality index (WQI
SW
)
Surface water quality index is assessed for the present (WQI
2006
, based on CLC 2006) and for
the future (WQI
2050
, for each of the five PRELUDE land use scenarios). Surface water quality
index WQI
SW
was calculated with ArcGIS in vector format by multiplying area of particular
CLC land use category with PLI value (Table 17) for this CLC land use category (see Table 8)
and normalizing by scaling from 0 to 1.
Figure 25 presents water quality index for surface waters (WQI
SW
), which is a potential for
surface water pollution. Since WQI
SW
is based on land use activities, these are reflecting in
the water quality index. In the present, areas with areas with higher potential for surface
water pollution (WQI
SW
) are mostly in lowlands, where there are intensive agricultural
activities, industrial areas and large cities. On the contrary, areas with low surface water
quality index (WQI
SW
) are in mountainous areas, where there are not many activities
resulting in water pollution. In the future in the most scenarios a great portion of agricultural
land is supposed to be abandoned and the portion of agricultural land use is supposed to
decrease (see Annex 2), therefore the most scenarios show lower WGI
SW
.
WQI
SW
is an index, which represents potential for surface water pollution, therefore it is not
necessary that in areas with high WQI actual qualitative water status is bad. Actual surface
water quality can be checked from the EU member state reports, where qualitative status of
surface water bodies and water resources at risk are defined for each year. In particular area
surface water body status could be good, but high WQI
SW
indicates that there is possible
pollution hazard in that area because of the land use.

58/82 WP3 report v.5 30.6.2014
Figure 25: Potential pollution load – surface water quality index (WQI
SW
) for present situation and five
PRELUDE future land use change scenarios (EEA 2007).
2.2.5 Groundwater quality index (WQI
GW
)
Sensitivity of groundwater bodies to pollution depends, in first place, on aquifer type or,
more specifically, on their effective infiltration coefficient, which represents the part of
rainfall that infiltrates into groundwater and that will eventually carry pollution load into
groundwater. Ground water quality sensitivity indicators are a function of pollution load and
effective infiltration coefficient.
The basis for spatial determination of groundwater quality index is International
Hydrogeological Map of Europe 1:1.500.000 - IHME1500 (Figure 26), which was kindly made
available in digital version by BGR (BGR & UNESCO 2014). HG factor is expressed as effective
infiltration coefficient. High coefficient values indicate higher groundwater quality
vulnerability; e.g. highly productive porous aquifers are very permeable and therefore more
vulnerable to groundwater quality than areas with insignificant aquifers, which have very
low permeability. For calculation of groundwater quality vulnerability effective infiltration
coefficient (Table 18) was applied to each aquifer type (Figure 27). Additionally, there are
some important confined aquifers in Po plain, Pannonian basin and Greece, which are lying

59/82 WP3 report v.5 30.6.2014
below shallow surface porous aquifer and confining layer with low permeability. For this
reason additional aquifer type was introduced: confined exstensive aquifer, for which a
value of 0.2 was set (Table 18 and Figure 27).
Figure 26: International Hydrogeological Map of Europe 1:1.500.000 (BGR & UNESCO 2014).
Table 18: HG factor - effective infiltration coefficient.
Aquifer type
Effective
infiltration
coefficient
1 Aquifers in which flow is mainly intergranular
1.1 extensive and highly productive aquifers 0.6
1.2 local or discontinuous productive aquifers or extensive but only moderately productive aquifers 0.3
Confined exstensive aquifer 0.2
2 Fissured aquifers. including karst aquifers
2.1 extensive and highly productive aquifers 0.8
2.2 local or discontinuous productive aquifers. or extensive but only moderately productive aquifers 0.4
3 Strata (granular or fissured rocks) forming insignificant aquifers with local and limited groundwater resources or strata with
essentially no groundwater resources
3.1 minor aquifers with local and limited groundwater resources 0.1
3.2 strata with essentially no groundwater resources 0.05

60/82 WP3 report v.5 30.6.2014
Figure 27: Effective infiltration coefficient as HG factor.
By multiplying surface water pollution index WQI
SW
(Figure 25) with HG factor (table 18) in
each grid we obtained groundwater pollution index (WQI
GW
) map, which was normalized by
scaling between 0 and 1.
Figure 28 presents groundwater quality index (WQI
GW
). Since it is based on land use activities
and hydrogeological characteristics, these are reflecting in the water quality sensitivity,
which is the largest in extensive agricultural areas in karst regions, which is the case of NE
Bulgaria: area between Va and Osum rives south of Pleven and in the NE Ludogorsko plateau
and Dobruja, which is extending to SE Romania. There are also some such cases in SE Italy (in
Puglia region close to Bari and Foggia) and NW Hungary (karst aquifers W and SW from
Budapest). In the future in the most scenarios a great portion of agricultural land is
supposed to be abandoned and the portion of agricultural land use is supposed to decrease,
therefore the most scenarios show even lower WGI
GW
.
WQI
GW
is an index, which represents potential for groundwater pollution, therefore it is not
necessary that in areas with high WQI
GW
actual qualitative water status is bad. Actual
groundwater quality can be checked from the EU member state reports, where qualitative
status of groundwater bodies and water resources at risk are defined for each year. In
particular area groundwater body status could be good, but high WQI
GW
indicates that there
is possible pollution hazard in that area because of the land use.

61/82 WP3 report v.5 30.6.2014
Figure 28: Potential pollution load – groundwater quality index (WQI
GW
) for present situation and five
PRELUDE future land use change scenarios.

62/82 WP3 report v.5 30.6.2014
3 Adaptive capacity
Adaptive capacity describes how well a system (water resources quantity and quality) can
adapt or modify to cope with the climate changes. A low adaptive capacity will result in high
vulnerability and vice-versa.
Adaptive capacity might reflect socio-economic and natural conditions. It may include
physical, environmental and socio-economic features. In CC-WARE project the ecosystem
services index was used as natural adaptive capacity and GDP as socio economic indicator.
The former expresses the role of the ecosystem in providing water in sufficient quantity and
quality and the latter expresses the economic capacity of a region to compensate ecosystem
service losses by technical or management measures.
3.1 Socio-Economic adaptive capacity factors
Economic status has one of the major roles in adaptation of drinking water supply to climate
change and can be measured with indicator GDP. Lower the GDP, lower is adaptive capacity
and the system is more vulnerable to climate change impacts.
Socio-economic adaptive capacity factors are population density and economic status: GDP,
employment rate etc. Population density is included already in domestic water demand, land
use and potential water pollution load. Employment rate is related to GDP, therefore only
GDP has been applied as socio-economic indicator. Population density is used also for
downscale data from NUTS 2 to NUTS 3 scale.
The GDP (gross domestic product) data is an indicator of the output of a country or a region
and was obtained from EUROSTAT database for all SEE countries except for Serbia. The GDP
reflects the total value of all goods and services used for intermediate consumption in their
production and it is expressed in PPS (purchasing power standards) to eliminate differences
in price levels between countries. The GDP data on EUROSTAT was available on NUTS 2 level
and was therefore downscaled to NUTS 3 level using population density of each NUTS 3
polygon. For Serbia GDP data was received from IPA1 partner on municipality level.
GDP values for partner countries are higher in western countries, such as Italy, Austria and a
part of Greece around Thessaloniki (Figure 29) and lower in southeastern part. The lowest
GDP values are in Serbia and Romania. There are also some areas with very low GDP values
in Austria, Slovenia, Hungary and Greece, which is due to low population density in most of
these areas, because GDP data were downscaled to NUTS 3 by population density. The GDP
map was normalized by scaling from 0 to 1 for calculation of adaptive capacity and
integrated vulnerability (see chapter 4). In this case the distribution is very homogeneous
because of extreme GDP values in the most developed region in Europe (Po plain area).

63/82 WP3 report v.5 30.6.2014
Figure 29: GDP as indicator of adaptive capacity (values and normalized map for integrated vulnerability
calculation).
3.2 Natural adaptive capacity factors
3.2.1 Ecosystem services
Furthermore, natural system plays an important role for drinking water sources protection.
Therefore ecosystems can be natural indicator for adaptation capacity. E.g. wetlands have
high protective value for drinking water protection. Ecosystem services have three functions:
Provisioning Ecosystem Service, Water Regulation and Water Quality Regulation. ESS can
increase ability of a particular area to provide water supply, or a qualitative rank of potential
ability of a particular area to provide excellent (both quantity and water quality) water
supply, i.e., areas where ESS are more sensitive, have a higher vulnerability from water
supply perspective.
For estimation of ecosystem services potential for drinking water, to each land use category
and ESS type is assigned importance for water supply. With these relative weights for each
land use-ESS category pair is assigned. Sum of the weights for each CLC land use class for all
three ES services and their normalization create ESS value to water Supply index (Ecosystem
Services Index ESSI) with values between 0 and 1. Detailed explanation of the ESS index
determination is presented in WP4 report.
Figure 30 presents ES services in water resources perspective. Very high ESS index is mainly
in mountainous areas, which means that those areas present high ES service and therefore
high adaptive capacity. On the other hand very low ES services signify very low adaptive
capacity, which is in valleys and plains, where all human activities are present (settlements,
agriculture and industry). This is due the fact that ES services for water supply are the
highest in forested and wetland ecosystems, followed by grassland ecosystems and the
lowest in agricultural ecosystems.
In the future in the most scenarios a great portion of agricultural land is supposed to be
abandoned and the portion of agricultural land use is supposed to decrease (see Annex 2). In
spite of that the ES service in water resources perspective is not much changing in the
future.

64/82 WP3 report v.5 30.6.2014
Figure 30: ES services in water resources perspective for present situation and five PRELUDE future land use
change scenarios (EEA 2007).

65/82 WP3 report v.5 30.6.2014
4 Integrated assessment of water resources vulnerability to
climate change
The vulnerability index (VI) should be able to compare and rank vulnerability over SEE and
form the basis of analyzing mitigation actions (WP4) and developing transnational strategy
for national/regional action plans (WP5).
There are several methods to determine integrated (overall) vulnerability index, which is a
composite of multiple quantitative indicators. The indicators are aggregated into groups
according to function. In CC-WARE project two groups of indicators were selected:
- water resources indicators group with indicators:
- annual local water exploitation index considering seasonality (LWEI
asw
) and
- groundwater quality index (WQI
GW
)
and
- adaptive capacity indicators group with indicators:
- GDP and
- ecosystem services index (ESSI).
These indicators can be combined with diverse formulas or can be combined as combination
of vulnerability classes to determine integrated (overall) vulnerability index.
Combining water resources indicators with adaptive capacity indicators we get integrated
vulnerability of water resources. The vulnerability is high in case of high impact, which can
result from high local water exploitation Index (low local total runoff, increased water
demand ) and high pollution potential, and low adaption capacity, such as low GDP and ESS.
Integrated vulnerability was calculated for present and future. For the future we have four
scenarios of water demand (- 10 %, no change, + 10 % and +25 % of present water demand)
and five scenarios of land use change. Taking into consideration all these scenario
combinations would result in too many combinations. Therefore we decided for two
extreme water demand scenarios (- 10 % and + 25 %) and land use scenarios (S1: Great
Escape - Europe of contrast and S5: Big Crisis - Europe of cohesion). Also in this case several
combinations have to be calculated (Table 19).
Table 19: List of maps for Water Resources Index, Adaptive Capacity and Integrated Vulnerability.
Water Resources - WR Adaptive Capacity – AC Integrated Vulnerability - IV
WR_base AC_base IV_base
WR_future_S1_-10% AC_S1 IV_S1_-10%
WR_future_S5_-10% AC_S5 IV_S5_-10%
WR_future_S1_+25% IV_S1_+25%
WR_future_S5_+25% IV_S5_+25%

66/82 WP3 report v.5 30.6.2014
4.1 Integrated vulnerability according to composite programming formula
(HU-method)
A composite integrated vulnerability index is determined by a multi criteria method
(composite programming), which provides a transparent method of assessment and
organizes indicators into a hierarchical structure (Figure 31). The indicators may have various
importance in forming overall vulnerability. These may be represented by assigning weights
to the indicators. For comparability, these weights should be uniform over all regions, and
were assessed by the CC-WARE expert group. Some indicator group may balance the
indicators out, e.g. lower water quantity in a wealthy region, or low water quality in a less
populated region. On the other hand, other indicators may not balance them out, e.g.
enough water quantity and low water quality. This latter case indicates a “limiting factor” or
“veto” situation.
Figure 31: Determination of integrated vulnerability according to composite programming.
Calculation model takes into consideration weighting and balancing factors (Figure 31).
Weights represent the relative importance of each indicator within one group as viewed by
the expert. Balancing factors are assigned for each group of indicators. Balancing factors
reflect the relative importance that is assigned to the maximal deviations of the indicators
and limit the ability of one indicator to substitute for another. In other words, it reflects the
strength of the preference for a particular objective, defining its relative importance.
Generally, the balancing factors and weights are assessed by expert group.
Finally, integrated vulnerability index is calculated for each group of basic indicators using
the following equation:

67/82 WP3 report v.5 30.6.2014
j
j
j
p
n
i
p
ijijj
SaIV
1
1








⋅=
∑
=
(26)
where
S
ij
is the normalized value of basic indicator i in the group j of indicators,
n
j
is the number of indicators in group j,
a
ij
is the weight of expressing the relative importance of indicators in group j such that their
sum equals one,
p
j
is the balancing factor among indicators for group j.
Water resources and adaptive capacity are two groups in this calculation; according to
balancing factors and weights from Figure 31 integrated vulnerability is then:
    




   
(27)
    

   




(28)
            
(29)
Integrated vulnerability of water resources is then:
      

   




             
(30)
4.1.1 Water Resources Index (WR_HU)
The first step of determination of integrated water resources vulnerability is to consider
exposure to climate change and the sensitivity of the indicators for water quantity and water
quality to those changes. This step provides an understanding of the potential impacts of
climate change on water resources. In the CC-WARE project water quantity indicator is
LWEI
asw
and water quality indicator is WQI
HG
. Combining these two we obtain water
resources index (WRI, Figure 31). Resulting data set is normalized in order to bring
proportion with other data sets for calculations.
Water resources vulnerability in present period is very low in mountainous area in the Alps,
Carpathian and Dinaric area (Figure 32). Very high and high water resources index is in some
parts of Po valley and in SE Italy, around Vienna in Austria, around Budapest and SE from
Balaton in Hungary, around Belgrade and in central Serbia, in SE Romania at the Black sea,
around Sofia, NE from Sofia, Maritsa valley and in NE Bulgaria, NE and SE Greece. This is due
to combination of high water stress and potential pollution load. In the future period the
pattern remains the same (Figure 33), smaller changes are due to lower/higher water
demand (use) and land use changes.

68/82 WP3 report v.5 30.6.2014
Figure 32: Water Resources Index based on mean annual ensemble values of RegCM3, ALADIN and PROMES
models for present (P) period (1991-2020).
Figure 33: Water Resources Index based on mean annual ensemble values of RegCM3, ALADIN and PROMES
models for future (F) period (2021-2050) for different water demand and land use scenarios.
4.1.2 Adaptive Capacity Index (AC_HU)
The second step is assessment of adaptive capacity with combining GDP and ESSI (see Figure
31). Again, resulting data set is normalized in order to bring proportion with other data sets
for calculations.
GDP is dominating adaptive capacity (Figure 33), because GDP was normalized in order to
bring proportion with other data sets for calculations, with this the distribution is very
homogeneous because of extreme GDP values in the most developed region in Europe (Po
plain area). Very high adaptive capacity is therefore in northwestern Italy, in the region with

69/82 WP3 report v.5 30.6.2014
the highest GDP (Milano area). In the future (Figure 34) the changes are due to land use
changes (ESSI).
Figure 34: Adaptive capacity based on mean annual ensemble values of RegCM3, ALADIN and PROMES
models for present (P) period (1991-2020).
4.1.3 Integrated vulnerability (IV_HU)
Finally, integrated vulnerability index is calculated for each group of basic indicators using
the following the equation (30).
Integrated vulnerability index in the present (Figure 36) has similar pattern as local water
exploitation index (LWEI
asw
) and water resources index, but the adaptive capacity lower
vulnerability for one class. LWEI
asw
as indicator for water availability is dominating the
integrated vulnerability, which is actually good, since also if water quality is very good, we
cannot use these water resources in case there is not enough quantity. Impact of water
quality on integrated vulnerability can be seen in NE Bulgaria (area between Va and Osum
rives south of Pleven and in the NE Ludogorsko plateau and Dobruja), where the area with
high vulnerability is larger.
High integrated vulnerability (Figure 36) is in the most of the project area except in the Alps
in N Italy, Austria and in Slovenia, in small area in western Hungary, western Serbia and in
Figure 35: Adaptive capacity based on mean annual ensemble values of RegCM3, ALADIN and PROMES
models for future (F) period (2021-2050) for different land use scenarios (ESS).

70/82 WP3 report v.5 30.6.2014
the Carpathian, where integrated vulnerability is low. Very high integrated vulnerability is in
some scattered areas in southern Italy, south from Vienna in Austria, in eastern Balaton area
in Hungary, in central Serbia, SE Romania at the Black sea, in Bulgaria very scattered areas
north from Sofia, in Maritsa plain and at the Black sea, in NE Greece, around Athens and in
Crete.
In the future (Figure 37) the integrated vulnerability pattern is the same, there are slight
changes due to water demand change and land use changes.
Figure 36: Integrated vulnerability index based on mean annual ensemble values of RegCM3, ALADIN and
PROMES models for present (P) period (1991-2020).
Figure 37: Integrated vulnerability index based on mean annual ensemble values of RegCM3, ALADIN and
PROMES models for future (F) period (2021-2050) for different water demand and land use scenarios.

71/82 WP3 report v.5 30.6.2014

72/82 WP3 report v.5 30.6.2014
5 References
Baltas E. (2007)
Spatial distribution of climatic indices in northern Greece. Meteorol. Appl. 14: 69-
78.http://onlinelibrary.wiley.com/doi/10.1002/met.7/pdf
Beguería S. Vicente-Serrano SM (2007) Package ‚SPEI‘ - Calculation of the Standardised Precipitation-
Evapotranspiration Index. Program instructions. 16p.
http://cran.r-project.org/web/packages/SPEI/SPEI.pdf
Budyko MI (1974) Climate and life. Translated from Russian. by D. H. Miller. Academic Press. Orlando. FL. 508
pp.
Busuioc A. Dumitrescu A. Baciu M. Cazacioc L. Cheval S (2010) RCM performance in reproducing temperature
and precipitation regime in Romania. Application for Banat Plain and Oltenia Plain. Romanian J
Meteor. 10(2): 1-19.
Castro M. Fernandez C. Gaertner MA (1993) Description of a mesoscale atmospheric numerical model. In Diaz
JI. Lions JL (eds) Mathematics. climate and environment. Rech. Math. Appl. Ser.. vol. 27. Ser. Masson:
Paris: 230–253.
CC-WaterS (2010) Climate change and impacts on water supply. Final report. December 2010. 357 pp.
Déqué M (2007) Frequency of precipitation and temperature extremes over France in an anthropogenic
scenario: Model results and statistical correction according to observed values. Global and Planetary
Change 57: 16–26.
Dingman SL (1992) Physical Hydrology. Prentice Hall. Upper Savage. New Jersey. 646 pp.
Doerr A. H. (1963) De Martonne’s Index of Aridity and Oklahoma’s Climate, in: Proceedings of the Oklahoma
Academy of Science, Vol. 43, University of Oklahoma, p. 211-213.
EC DG Environment (2010) Background Document on Vulnerability Indicators for the project Climate
Adaptation – modelling water scenarios and sectoral impacts. Contract N° DG ENV.D.2/SER/2009/0034
EEA (2007) Land-use scenarios for Europe: qualitative and quantitative analysis on a European scale. ISSN
1725–2237. EEA Technical report No 9/2007.
EEA (2012) Water resources in Europe in the context of vulnerability; EEA 2012 state of water assessment. EEA
Report No 11/2012.
Elguindi N. Bi X. Giorgi F. Nagarajan B. Pal J. Solmon F. Rauscher S. Zakey A (2007) RegCM Version 3.1 User’s
Guide. Trieste: PWCG Abdus Salam ICTP. 57 pp.
Farda A. Déqué M. Somot S. Horányi A. Spiridonov V. Tóth H (2010) Model Aladin as Regional Climate Model
for Central and Eastern Europe. Stud Geophys AS CR: 313-332.
Formayer H. Haas P (2010) Correction of RegCM3 model output data using a rank matching approach applied
on various meteorological parameters. In Deliverable D3.2 RCM output localization method (BOKU-
contribution of the FP 6 CECILIA project on http://www.cecilia-eu.org/).
FOOTPRINT (2006) Characteristics of European groundwater vulnerability scenarios. DL10. 6FP programe.
Project no. 022704 (SSP).
Gaertner MA. Dominguez M. Garvert M (2010) A modelling case-study of soil moisture–atmosphere coupling.
Quarterly Journal of the Royal Meteorological Society 136(S1): 483-495. DOI: 10.1002/qj.541.
Gain AK. Giupponi C. Renaud FG (2012) Climate Change Adaptation and Vulnerability Assessment of Water
Resources Systems in Developing Countries: A Generalized Framework and a Feasibility Study in
Bangladesh. Water . 4. 345-366.
Gerrits AMJ. Savenije HHG. Veling EJM. Pfister L (2009) Analytical derivation of the Budyko curve based on
rainfall characteristics and a simple evaporation model. Water Res. Research 45. W04403.
doi:10.1029/2008WR007308.

73/82 WP3 report v.5 30.6.2014
Haylock MR. Hofstra N. Klein Tank AMG. Klok EJ. Jones PD. New M (2008) A European daily high-resolution
gridded data set of surface temperature and precipitation for 1950-2006. J Geophys Res. 113. D20119.
doi:10.1029/2008JD010201.
Hewitt CD (2004) Ensembles-based predictions of climate changes and their impacts. Eos. Transactions
American Geophysical Union 85(52): 566. doi: 10.1029/2004EO520005.
IPCC TAR WG1 (2001) Houghton JT. Ding Y. Griggs DJ. Noguer M. van der Linden PJ. Dai X. Maskell K. Johnson
CA. (ed) Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third
Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.
IPCC (2014)
Local Government Association of South Australia (2012) Guidelines for Developing a Climate Change
Adaptation Plan and Undertaking an Integrated Climate Change Vulnerability Assessment. November
2012.
http://www.lga.sa.gov.au/webdata/resources/files/Guidelines_for_Developing_a_Climate_Change_A
daptation_Plan.pdf
Maliva R., Missimer Th. (2012) Aridity and Drought, in: Arid Lands Evaluation and Management, Springer, p.21-
39.
McMahon TA. Peel MC. Lowe L. Srikanthan R. McVicar TR (2013) Estimating actual. potential. reference crop
and pan evaporation using standard meteorological data: a pragmatic synthesis. Hydrol Earth Syst Sci
17: 1331–1363. doi: 10.5194/hess-17-1331-2013.
Oudin L. Andréassian V. Lerat J. Michel C. (2008) Has land cover a significant impact on mean annual
streamflow? An international assessment using 1508 catchments. Journal of Hydrology. 357. 303–
316.
Vicente-Serrano SM. Beguería S. Lopez-Moreno JI (2010) A Multi-scalar drought index sensitive global warming:
The Standardized Precipitation Evapotranspiration Index – SPEI. Journal of Climate 23: 1696.
Roderick ML. Farquhar GD (2011) A simple framework for relating variations in runoff to variations in climatic
conditions and catchment properties. Water Resour. Res.. 47. W00G07. doi:10.1029/2010WR009826.
Spiridonov V. Somot S. Déqué M (2005) ALADIN-Climate: from the origins to present date. ALADIN Newsletter
n. 29 (nov. 2005).
Sullivan CA (2011) Quantifying Water Vulnerability: A Multi-Dimensional Approach. Stoch. Env. Res. Risk Assess.
p.627–640. 25. 2011.
Thornthwaite CW (1948) An approach toward a Rational Classification of Climate. Geographical Review 38(1):
55-94. Thornthwaite. C. W.. Mather. J. R.. 1957. Instructions and tables for computing potential
evapotranspiration and the water balance. Publ. Climatol. 10 (3). 311.
Thornthwaite CW. Mather JR (1957) Instructions and tables for computing potential evapotranspiration and the
water balance. Publ. Climatol. 10 (3). 311.
UNEP (1997). World atlas of desertification 2ED. UNEP (United Nations Environment Programme). London.
http://www.fao.org/geonetwork/srv/en/metadata.show?id=37040
UNEP (2009) METHODOLOGIES GUIDELINES; Vulnerability Assessment of Freshwater Resources to
Environmental Change. United Nations Environment Programme. Nairobi.
UNEP (2012) Vulnerability Assessment of Freshwater Resources to Climate Change: Implications for Shared
Water Resources in the West Asia Region. United Nations Environment Programme. Nairobi.
Vicente-Serrano SM. Beguería S. Lopez-Moreno JI (2010) A Multi-scalar drought index sensitive global warming:
The Standardized Precipitation Evapotranspiration Index – SPEI. Journal of Climate 23: 1696.
Vörösmarty CJ. Green P. Salisbury J. Lammers RB (2000) Global Water Resources: Vulnerability from Climate
Change and Population Growth. Science. vol. 289: 284-288.

74/82 WP3 report v.5 30.6.2014
Wochna A. Lange K. Urbanski J (2011) The influence of land cover change during sixty years on nonpoint source
phosphorus loads to Gulf of Gdansk. GIS Centre. University of Gdansk . Journal of Coastal Research.
ISSN 0749-0208. SI 64. Special Issue.
Zhang L. Potte. N. Hickel K. Zhang Y. Shao Q (2008) Water balance modeling over variable time scales based on
the Budyko framework – Model development and testing. Journal of Hydrology. 360. 117-131.
Zhang L. Hickel K. Dawes WR. Chiew FHS. Western AW. Briggs PR (2004) A rational function approach for
estimating mean annual evapotranspiration. Water Resour. Res.. 40. W02502.
doi:10.1029/2003WR002710.
Zhang L. Dawes WR. Walker GR (2001) Response of mean annual evapotranspiration to vegetation changes at
catchment scale. Water Resour. Res. 37. 701–708.

75/82 WP3 report v.5 30.6.2014
ANNEX 1 – Handling with water demand data
Water demand data for different sectors were gathered and unified in one large MS Excel
Spreadsheet, from where they were transformed into GIS environment. Data was collected
on NUTS3 statistical level from each country, with two exceptions. For Italy, only selected
NUTS3 regions were included in the project (not all of the Italy), and these regions were used
in the mask. For Serbia, municipalities were used instead of NUTS regions, as this country is
not in the statistical EU NUTS region. One must note that the exact borders of Serbia do not
match exactly the country borders of other NUTS3 regions, but the gaps on the border are
small and were disregarded in the rasterization process.
To assure the best quality of data they were also compared with data adopted by FAO
(available at FAO online database), EUROSTAT database and with WD data from World Bank
database. Water use data for partner countries as annual values of water use are presented
in Table 1. Discrepancies among data are not big.
All data was saved into a vector shape-file (SHP format) with a file name
SEE_NUTS3_WD_final_ITA.shp. Please note that in the GIS model picture (Figure 1), the file
name is shortened to NUTS3_SEE for the increased readability.
Figure 1: A GIS model of creating maps.
Shape-file contains following attributes: FID and Shape, STAT_LEV for NUTS level, NUTS_ID
and NUTS3 for NUTS3 identification, AGRWD for agricultural water demand, DWD for
domestic water demand, INDWD for industrial water demand, WD_tot for total water
demand (WD_tot = AGRWD + DWD + INDWD) and DWD_summer (

) as a correction factor
(Figure 2).

76/82 WP3 report v.5 30.6.2014
Table 1: Comparison of water demand data from the CC-WARE project with FAO and World Bank database.
Water demand AGRWD DWD INDWD
with
thermal
and
nuclear PP
without
thermal
and
nuclear PP
WD_tot
AGRWD
(Water
wit hdrawal fo r
irrig ation)
IRRIG
(Irr igation wate r
re quir em e nt)
IRRIG REQ INDWD
DWD
(Total
wat er
wit hdrawal pe r
capita)
DWD WD_tot WD_tot
Country
World Bank
(2011)
Unit m
3
/inhab/yr
10^9 m3/yr
10^9 m3/yr
0.1 (2002) 0.0599 (2007) 2.695 (2008) 0.608(2008) 489.5 (1990) 3.807 (1990)
2.889 (2000) 0.668 (1999) 457.1 (1997) 3.644 (1997)
1.548 (1989) 0.4028 (1989) 452.4 (2002) 3.657 (2002)
2,288
3,66
CC-WARE - data FAO - data
10^9 m3/yr
10^9 m3/yr
Austria
0,0879 0,5337 1,6667 / /
Italy
7,3775 2,3321 0,5714 / / 31,312
45,41
20.01 (2000) 12.89 (2007) 8.022 (2007) 12.5 (1990) 6.5 (1990) 789.8 (2000) 45.41 (2000)
12.89 (2007) 16.29 (2000) 9.111 (1999)
8.943 (2005)
9.095 (2008)
Italy
7,3775 2,3321 0,5714 / / 31,312
45,41
4,759
5,59Hungary
0,3849 0,7159 3,7168 3,8689 3,8198
0.68 (2002) 0.174 (2002) 0.0253 (2007) 4.279 (1999) 0.84 (2000) 570.8 (2002) 5.799 (2002)
0.305 (2006) 0.04 (2006) 4.611 (2007) 0.667 (2007) 555.9 (2007) 5.583 (2007)
4,759
5,59Hungary
0,3849 0,7159 3,7168 3,8689 3,8198
Serbia
0,1920 0,7900 3,0420 / / 4,026
4,12
0.129 (2007) 0.0661 (2011) 0.0212 (2011) 3.133 (2007) 0.692 (2007) 402.1 (2007) 3.954 (2007)
0.077 (2009) 3.361 (2009) 0.683 (2009) 418.5 (2009) 4.121 (2009)
Serbia
0,1920 0,7900 3,0420 / / 4,026
4,12
8,367
6,12
Bulgaria
0,9151 0,9294 7,4282 4,0763 0,1989
2.225 (1990) 0.458 (2002) 0.2174 (2007) 4.852 (1990) 0.417 (1988) 866.3 (1990) 7.494 (1990)
0.141 (1997) 0.71 (2007) 3.769 (2000) 1.178 (2000) 720.9 (2002) 5.69 (2002)
0.743 (2002) 0.3155 (2011) 4.16 (2007) 1.026 (2007) 811.6 (2007) 6.201 (2007)
1.015 (2007) 4.145 (2009) 0.978 (2009) 827.1 (2009) 6.119 (2009)
0.996 (2009)
8,367
6,12
Bulgaria
0,9151 0,9294 7,4282 4,0763 0,1989
Romania
1,3200 1,2200 5,1600 / / 13,135
6,88
9.1 (1990) 0.784 (2002) 0.379 (2007) 9.06 (1990) 2.25 (1990) 884.4 (1990) 20.41 (1990)
2.98 (1997) 0.718 (2007) 7.43 (1997) 2.05 (1997) 554.6 (1997) 12.46 (1997)
1.192 (2002) 0.731 (2009) 6.17 (2002) 1.86 (2002) 419 (2002) 9.222 (2002)
1.099 (2007) 5.64 (2003) 1.69 (2003) 389.4 (2007) 8.429 (2007)
1.171 (2009) 4.2 (2009) 1.505 (2005) 321.5 (2009) 6.876 (2009)
Romania
1,3200 1,2200 5,1600 / / 13,135
6,88
8,155
9,47
Greece
6,8272 0,9316 0,3964 0,2482 0,1482
5.694 (1990) 8.42 (2002) 5.441 (2007) 0.063 (2000) 0.795 (2000) 679.2 (1990) 7.03 (1990)
7.6 (1997) 8.458 (2007) 0.167 (2007) 0.846 (2007) 837.9 (2002) 9.278 (2002)
8.42 (2002) 841.4 (2007) 9.471 (2007)
8.458 (2007)
8,155
9,47
Greece
6,8272 0,9316 0,3964 0,2482 0,1482
Slovenia
0,0036 0,1642 0,6762 / / 0,844
0,94
0.007 (2002) 0.007 (2002) 0.0006 (2010) 0.085 (2000) 0.22 (2000) 456.7 (2002) 0.312 (2002)
0.005 (2007) 0.005 (2007) 0.758 (2005) 0.163 (2005) 460.2 (2007) 0.926 (2007)
0.002 (2009) 0.0016 (2010) 0.775 (2009) 0.165 (2009) 461.8 (2009) 0.942 (2009)
Slovenia
0,0036 0,1642 0,6762 / / 0,844
0,94

77/82 WP3 report v.5 30.6.2014
Figure 2: Attribute list for water demand shape file.
This shape-file was then transformed into several water demand raster layers by ArcGIS
(Feature to raster tool). Total water demand was rasterized into WD_tot layer, agricultural
water demand into AGRWD, domestic water demand into DWD, and industrial water
demand into INDWD layer.
WD maps were produced on NUTS 3 level in vector format, except for Serbia, for which data
was collected on Municipality level. When all WD maps were transformed from vector to
raster, “Feature to Raster (Conversion)” was applied. This tool always uses the cell center to
decide the value of raster pixel. Thus at the country borders empty cells can be observed.

78/82 WP3 report v.5 30.6.2014
ANNEX 2 – The PRELUDE scenarios for land use
changes in the future
Land use changes in the future were adopted form the PRELUDE scenarios (EEA 2007) for
land use changes:
- Scenario 1: Great Escape — Europe of contrast
- Scenario 2: Evolved Society — Europe of harmony
- Scenario 3: Clustered Networks — Europe of structure
- Scenario 4: Lettuce Surprise U — Europe of innovation
- Scenario 5: Big Crisis — Europe of cohesion.
A short description of these five scenarios is listed below. More about this study and
scenarios is written in the EEA Technical report No 9/2007: “Land-use scenarios for Europe:
qualitative and quantitative analysis on a European scale”.
Figure 1 shows the simplified spider diagrams for all five scenarios according to driving
forces: environmental awareness (including renewable energy, environmental awareness,
climate change), solidarity and equity (including social equity, quality of life, human
behavior, health concern), governance and intervention (including policy intervention,
subsidiarity), agricultural optimization (including agricultural intensity, self‑sufficiency,
international trade) and technology and innovation (technological growth). Driving forces
the ones addressing economy and population were not used in this aggregation. They are:
population growth, ageing society, settlement density, internal migration, immigration, daily
mobility, economic growth.
Figure 1: Simplified spider diagrams of key drivers.

79/82 WP3 report v.5 30.6.2014
SHORT DESCRIPTION OF LAND USE SCENARIOS
1. The base year situation
- Rural landscapes present majority of landscapes in EU
- Urban areas are largely concentrated in NW EU
2. The Great Escape – Europe of contrast
- Economic globalization increases global competition pressure
- High technological innovation rates
- Social protection become more individualized
- Social tension causes impoverishment and poor immigrants move to urban
centers
- Total agriculture diminishes (abandoned grasslands and converted into arable
land, nature reserves and farmlands are lost)
Key developments in this scenario concern:
- the increased importance of international trade (economic globalization),
- the decreasing societal solidarity,
- strong reduction of policy interventions.
3. Evolved Society – Europe of harmony
Summary:
- Heavy floods and exploding energy prices reinforce environmental awareness
- People believe that lifestyle and economy must change
- People move away from densely populated areas to rural safe areas (Eastern
EU)
- High-tech farming – increasingly organic
- Agricultural areas remain the same with decreasing farming intensity
- At flooding areas cropland is reduced considerably
Key developments in this scenario concern:
- far reaching energy crisis, which triggers increased support for renewable
energies
- strong increase of environmental awareness
- policy makers  environmentally sustainable regional development
4. Clustered Networks – Europe of structure:
Summary:
- Globalization propels economic growth – environmental and health
conditions get worse
- Local shops and services close down at the countryside
- Needs of aging society lead to the development of coherent spatial planning
policies
- Migration from polluted urban areas is encouraged
- Peripheral areas become economic and social focal points
- Urbanization is concentrated
- Agriculture marginalizes – large scale abandonment

80/82 WP3 report v.5 30.6.2014
- Biodiversity, water, soil and air quality benefits from receding agriculture and
green belts
Key developments in this scenario concern:
- the impacts of the pollution dynamics (affecting society),
- strong marginalization of agriculture occurrence of strong spatial planning
5. Lettuce Surprise U – Europe of innovation
Summary:
- Food security crisis hits EU
- Crisis management fails faith in central government and safety decreases
- Alternatives for food production, control regime and regional self-sufficiency
are strived for
- New communication technologies, development of innovative technologies,
new opportunities
- Migration is limited and urban patterns doesn’t change
- Environmental awareness grows, env. friendly produced food
- New crop inventions (higher yields with lower inputs)
- Agriculture is high tech, clean and relatively small-scale
- Increased productivity decreases croplands
- Reduction of agricultural areas leads to increase in biodiversity,
improvements in soils, water and air quality
Key developments in this scenario concern:
- open source technological breakthrough,
- strong increase of environmental awareness
- increase of far-reaching decentralization of political decision-making, self-
regulation becomes more important.
6. Big Crisis – Europe of cohesion
Summary:
- A series of environmental disasters highlights EUs vulnerability and inability to
effectively adapt
- Widespread support of EU policies and concerns for solidarity and equity
- Sustainable and regionally balanced development is consolidated
- Public transports are strongly prompted due to env. awarness
- Agricultural oversupply diminishes and focuses on landscape stewardship
- Land-use changes are limited, cropland and grassland decreases moderately
- Population in urban areas decreases slightly
- Soil, water and air benefit from agricultural extensification and limited land
abandonment
Key developments in this scenario concern:
- growing the environmental awareness and social solidarity (due to increase of
env. disasters)

81/82 WP3 report v.5 30.6.2014
LANDSCAPE TYPE CHANGES
Urban land use increases slightly in all scenarios. The main difference between the scenarios
is the shift between urban areas and urban landscapes. As an example of this development,
we look at the 'Lettuce Surprise U'‑scenario where a considerable shift towards urban
landscape takes place. This increase is, however, not due to an increase in urban area itself.
It is because landscapes with an urban character, which were dominated by both urban
areas and cropland in 2005, are only dominated by urban land in 2035. On the other hand, in
the case of the 'Clustered Networks'‑scenario, the agricultural area is maintained around
urban areas as landscape buffer zones, so that the ratio of urban to agricultural land is not
changed. Therefore, we see no change in the landscape with urban character.
Agricultural land use decreases in all scenarios. Whereas in 2005, rural landscapes
(particularly those that are cropland‑dominated) present a majority of landscapes in Europe,
in 2035 this is only true for the 'Evolved Society'‑scenario. Due to substantial abandonment
of both cropland and grassland, there is a shift in dominance in 'Great Escape', 'Lettuce
Surprise U', and 'Big Crisis' towards at least one of the three other rural landscape types.
Shifts in land use patterns do not occur homogeneously throughout Europe. Whereas
Scandinavia remains almost unchanged in all five scenarios, changes are particularly large for
Eastern Europe, the Iberian Peninsula and some countries inside the 'Blue Kangaroo',
depending on the particular scenario. Figure 2 shows the landscape type comparison
between 2005 and 2035 for all scenarios.
Great Escape: This scenario is the only scenario where landscapes with agricultural character
are maintained only in cropland areas that are optimal for production. Therefore, we see a
large shift from cropland‑dominated rural landscapes towards rural mosaic landscapes in
central and Eastern Europe whereas in southern Europe, especially in Spain, these turn into
rural landscapes with large fractions of abandoned land.
Evolved Society: In this scenario the landscape patterns of the base-year situation are almost
maintained. Land abandonment remains limited, mainly due to targeted policies. The
reduction of grassland by about 14 % has relatively small overall impacts.

82/82 WP3 report v.5 30.6.2014
Figure 2: Landscape type comparison between 2005 and the scenarios in 2035.