Strategies to reduce water stress in Euro-Mediterranean river basins

STOTEN-17720; No of Pages 13 Science of the Total Environment xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Strategies to reduce water stress in Euro-Mediterranean river basins☆ Luis Garrote a, Alfredo Granados a, Ana Iglesias b,⁎ a b

Department of Civil Engineering, Hydraulics, Energy and Environment, Technical University of Madrid (UPM), Spain Department of Agricultural Economics and Social Sciences, Technical University of Madrid (UPM), Spain

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• We model water scarcity accounting for reservoir operation and water management. • We apply the high resolution model to the Euro-Mediterranean region. • We show that increasing resolution increases the quality of the assessment. • Incorporating water management will define more appropriate policy choices. • A portfolio of management strategies illustrated regional policy choices.

a r t i c l e

i n f o

Article history: Received 30 January 2015 Received in revised form 22 April 2015 Accepted 28 April 2015 Available online xxxx Keywords: Water stress Mediterranean Climate change Adaptation Modelling approach

a b s t r a c t A portfolio of water management strategies now exists to contribute to reach water demand and supply targets. Among them, integrated water resource management has a large potential for reducing water disagreement in water scarcity regions. Many of the strategies are based on well tested choices and technical know-how, with proven benefits for users and environment. This paper considers water management practices that may contribute to reduce disagreement in water scarcity areas, evaluating the management alternatives in the Mediterranean basins of Europe, a region that exemplifies other water scarcity regions in the world. First, we use a model to compute water availability taking into account water management, temporal heterogeneity, spatial heterogeneity and policy options, and then apply this model across 396 river basins. Second, we use a wedge approach to illustrate policy choices for selected river basins: Thrace (Greece), Guadalquivir, Ebro, Tagus and Duero (Spain), Po (Italy) and Rhone (France). At the wide geographical level, the results show the multideterminant complexities of climate change impacts and adaptation measures and the geographic nature of water resources and vulnerability metrics. At the local level, the results show that optimisation of water management is the dominating strategy for defining adaptation pathways. Results also show great sensitivity to ecological flow provision, suggesting that better attention should be paid to defining methods to estimate minimum ecological flows in water scarcity regions. For all scales, average water resource vulnerability computed by traditional vulnerability indicators may not be the most appropriate measure to inform climate change adaptation policy. This has large implications to applied water resource studies aiming to derive policy choices, and it is especially interesting in basins facing water scarcity. Our research aims to contribute to shape realistic water management options at the regional level and therefore provide information to climate change, agricultural and water policies. © 2015 Elsevier B.V. All rights reserved.

☆ Special Issue: CLIWASEC (Climate Change, Water and Security in the Mediterranean, Ralf Ludwig & Roberto Roson, Guest Editors). ⁎ Corresponding author.

http://dx.doi.org/10.1016/j.scitotenv.2015.04.106 0048-9697/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

2

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

1. Introduction Humans are deeply integrated into the water cycle and their actions strongly determine the availability of water resources (Vörösmarty, 2002). This is especially true in Mediterranean Europe. The human influence in water supply in Mediterranean European countries is one of the oldest influences in the world. Water infrastructure to deliver water supply and sanitation was implemented in Greece in IV B.C. and irrigation systems were developed in Spain in IV B.C. At the same time, Mediterranean societies have developed a great culture of water management. As a result, Mediterranean Europe follows a continuous adaptation process. When more water is available the demands rise while in scarcity situations the demands are restricted. I.e. water use is balanced with water availability. The main adaptation process has been to regulate water, so water availability is defined by regulation. From Greece to Spain, the Euro-Mediterranean region has sustained his people for millennia. But rapid changes in population, lifestyle and climate change are turning the region into disagreement over water and land. The last ten years were the hottest on the global record and in many areas of Southern Europe were also the driest, resulting in crop failures and water imbalances that caused instability in many rural areas and, to a lesser extent, in cities. During these years the water resource community has made exceptional efforts in expanding the scope of research to include aspects of vulnerability that could be linked to social choices and has made a great collective effort to address the water vulnerability problem. But some issues remain unsolved. First, little is known about how vulnerability indicators of water scarcity respond to water management choices. Here we use a water adaptation policy model to contribute to the understanding of vulnerability in a region where water is intensively managed. A second unsolved issue is the prioritisation of adaptation strategies. Here we use a wedge approach to understand quantitatively the additive effect of management choices and illustrate the results with local examples. Water resource vulnerability has been evaluated with indicators at the global scale (Alcamo et al., 1997; Wallace, 2000; Sullivan, 2002; Hanasaki et al., 2013), and at smaller scales (Meigh et al., 1999; Vörösmarty et al., 2000; Boithias et al., 2014). Commonly used indicators include the Falkenmark Index (FI, defined as the average per capita water available per year, Falkenmark, 1986) that indicates social water stress, and the Criticality Ratio (CR, is computed as the ratio of mean water use to availability) that indicates technical water stress. These indicators are extremely useful for an overview of the vulnerability levels and they influence the decision-making process in the planning and management of the water resource systems (UN, 2003). Recent works have proposed indicators for assessing water scarcity problems under climate change (Chávez-Jiménez et al., 2013) and have developed methodologies that are based on these indicators to allocate water resources under demand constraint scenarios (Chávez-Jiménez et al., 2015). However, indicators do not analyse the causes behind real water scarcity challenges (Ludwig et al., 2011; Victoria et al., 2005). In practice, however, water shortages often differ due to three major characteristics of water resource systems that are not included in these first order assessments: spatial heterogeneity, climatic variability and regulation. To overcome these limitations we present a study for the entire Euro-Mediterranean region carried out at high spatial and temporal resolution accounting for reservoir operation and water management strategies under different climatic scenarios. Furthermore, practical planning and management of water resources require models with greater local real representation; these models provide improved estimates of the reasons behind vulnerability levels and how these might change as actions are implemented (Schewe et al., 2014). Here we use the WAAPA model (Garrote et al., 2011, 2015) that responds to these three determinants to understand the choices that could mitigate water shortages in the Mediterranean region. In 2004 Pacala and Socolow provided an analysis and clarification of how current mitigation options could contribute to the stabilising

atmospheric CO2, creating the concept of stabilisation wedges (Pacala and Socolow, 2004). This concept has been used widely ever because it provides a clear-cut way to link science to policy. Wada et al. (2014) applied a ‘water wedges’ concept as a framework to examine policies to mitigate the negative effects of water scarcity. In all cases, the wedge concept reflects on the understanding that a suite of management alternatives has to be used since none of the alternatives could provide a unique solution alone. The water management choices necessary to achieve the EU good environmental status target depend on the quantitative details of water saving potential of the strategies and the water policy that influences behaviour of the stakeholders that need to implement them. Water scarcity areas in Southern Europe are approximately one half of the total water scarcity of the entire Mediterranean region, which in turn represents the global region that sustains the largest amount of population with water stress (Iglesias et al., 2007). The environmental targets for this area include those collective for all the European Union and those of individual member States committed to focus his climate change adaptation plans on water resources (Quevauviller et al., 2005). In the global effort to reduce water scarcity, the potential of the Mediterranean region can significantly help to meet adaptation targets that exemplify other regions at risk (UN, 2003; Iglesias et al., 2011). The role of water resource management to provide solutions was recognised since the early 1990s (Gleick, 2003) and has been a major area of research in the last decades (Boithias et al., 2014; Garrote et al., 2015; Quevauviller et al., 2005). Water management has a large, costcompetitive adaptation potential to meet short to medium term targets for reducing water scarcity risks (UN, 2013; Iglesias et al., 2012). The inefficient use of water by poor management is exacerbated by climate variability, which could produce severe impacts on future water resources in Europe. Thus there is a need to increase the implementation of strategies which maintain water availability and, in turn, optimise good ecological status. Many researchers agree on a set of specific strategies, measures and technologies which have potential to increase efficiency, improve the environment, maintain the water services of rural populations, and contribute to adaptation to climate change (Gleick, 2003; Ludwig et al., 2011). This set of strategies can include among others, a more coherent integrated management and the implementation of policies that make urban water systems more efficient. Effective options, and their costs, are the main focus of recent research to facilitate governments' better understanding about the implementation of specific sets of strategies at the regional and local levels (IPCC, 2014). But the information on the cost of strategies is still limited and fragmented and remains a topic of disagreement (IPCC, 2014). The literature on adaptation of water resources includes a diverse array of options and some very large estimates of the global potential (IPCC, 2014). Here we restrict our attention to the strategies that are relevant to semiarid environments and have linkages to climate change adaptation. About two thirds of the volume of water used in EuroMediterranean countries is managed inappropriately, up to one-half of the water is under poor ecological status, primarily due to the interannual climatic variability (Garrote et al., 2015). About half of the agricultural land suffers from water scarcity every year. Strategies such as improved management are not quantified at the local level or tested against other alternatives, and choices are often made without data to support them. By 1995, water reuse was well established in the region, and irrigation with water reused has been adopted in about one tenth of the agricultural land. This trend has a large potential for being implemented in 20% of the irrigated cropland (Garrote et al., 2015). A combination of options could be extended to most of the basins with water stress, accompanied by a verification programme that enforces the adoption of strategies that actually work as advertised, a good case could be made for the European Union Water Framework Directive

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

Water availability under 4 climate scenario sand 2 time periods across 396 Mediterranean subbasins (WAAPA model)

Exploring strategic choices for reducing water stress in Mediterranean basins

Effect of management choices in water availability as a fraction of mean annual flow (FWA)

A wedge approach for prioritising management choices & inform policy, examples in 7 sub-basins

Fig. 1. Structure of the study to estimate the strategies to reduce water stress in Mediterranean basins.

(EUWFD) that an additional half of the water resources in Europe could be restored in this way. With the aim of providing support for the implementation of water management measures, we believe that two questions are particularly relevant: What is the danger in the different areas under current management? How successful are the proposed measures for reducing water scarcity and defined new policies? We address these questions first by evaluating water availability under different climate assumptions; define the effect of implementing different strategies at the regional level. Second, we provide a wedge approach analysis of the strategies as a tool for linking science to policy. In order to provide realism to the analysis we selected seven representative case studies in Southern Europe that exemplify Mediterranean river basins. The paper is structured into 5 sections: Section 1 is the introduction; Section 2 presents the methods and data; Section 3 presents the results and discussion of the adaptation choices; and Section 4 draws some conclusions for the use of the results.

2. Methods 2.1. Framework Our approach to explore strategies to reduce water scarcity is summarised in Fig. 1. The methodology includes three sequential steps: First, an evaluation of water availability under four climate change scenarios and three time periods across the 396 sub-basins of Europe that have Mediterranean climate; second, estimation of the effect of management choices in water availability as a fraction of mean annual flow, these two steps are carried out with the WAAPA

3

model (see below), and third, estimation of the water scarcity mitigation potential of the selected strategies to link the results to policy choices. 2.2. Geographical extent The spatial coverage extends to the major river basins from Europe that have Mediterranean climate, that is the 396 sub-basins shown in Fig. 2 that account for 1,750,000 km2. The location of the case study basins is also indicated in the map. 2.3. Model for the study: WAAPA The analysis is based on the Water Availability and Adaptation Policy Analysis (WAAPA) model, which focuses on the quantitative evaluation of maximum potential water withdrawal for different types of demands (Garrote et al., 2011, 2015). WAAPA is a GIS-based model that is conceived as a tool for the analysis of water resource systems to evaluate the effectiveness of different policies. The model is prepared to be used in systems and situations (as the climate change projections) where limited information is available. Thus, the inputs could be taken from public geographical data-bases. The model inputs are: the system topology and characteristics, the inflow series, the evaporation rates, the restrictions (as the environmental flow) and the demand location, type, value and priorities. The evaluation of the water availability depends on the management strategies. WAAPA offers three different management options to compute water availability, depending on the development of the system: local management, in which reservoirs are operated to serve local demands only; global management of distribution, in which sub-systems of reservoirs are jointly operated to serve the subbasin demands; and global management of supply and distribution, in which all reservoirs are jointly operated to supply all demands. The WAAPA model is used to estimate the water vulnerability. The model computes the maximum water demand that could be provided at a given point in the river network with the available water infrastructure. For each type of demand the WAAPA provides the minimum required reliability and certain seasonal variation; here we used 98% for urban areas and 90% for irrigation. WAAPA performs the simulation of water resource systems at the monthly time scale and allows the estimation of the demand–reliability curve in every sub-basin of the river network. The demand performance analysis was applied to estimate the exposure of the basins to climate change. 2.4. Scenarios Streamflow monthly time series were obtained from the results of the ENSEMBLES project in four climate scenarios. The transient runs (1950–2100) were split into three periods: CTL: 1960–1990 (Oct 1961

Fig. 2. Geographical extent and basins simulated with the WAAPA model. The case study basins are indicated in shaded colour: 1. Duero; 2. Tagus; 3. Guadalquivir; 4. Ebro; 5. Rhone; 6. Po; and 7. Thrace.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

4

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

Table 1 Climate change scenarios. Scenario name in this study

Global model

Regional model

Resolution and time frame

ENSEMBLES file

Socio-economic assumptionsa

KNMI A1B ETHZ A1B CRNM A1B CRNM E1

ECHAM5-r3 HadCM3Q0 ARPEGE ARPEGE

RACMO2 CLM RM5.1 CM4

25 × 25 km, 1950–2100 25 × 25 km, 1950–2100 25 × 25 km, 1950–2100 50 × 50 km, 1950–2100

KNMI-RACMO2_A1B_ECHAM5-r3_MM_25km_mrro.nc ETHZ-CLM_SCN_HadCM3Q0_MM_25km_mrro.nc CNRM-RM5.1_SCN_ARPEGE_MM_25km_1950-2100_mrro.nc ARPEGE4.6_SCN_CNRM-CM4_MM_50km-CRU_1950-2100_mrro.nc

A1B A1B A1B E1, stabilisation scenario

a

See Nakiçenoviç et al. (2000).

to Sep 1991), short term (ST: 2020–2050, Oct 2019 to Sep 2049), and long term (LT: 2070–2100, Oct 2069 to Sep 2099). In 2010 Moss et al. (2010) published an update approach to construct climate change scenarios and introduced the concept of Representative Concentration Pathways. Here we have followed this approach although the scenarios available at the time the research was conducted were forced by the greenhouse gas concentrations described by Nakiçenoviç et al. (2000) (see Table 1), since the climate simulations for the new representative concentration pathways were not available at the time of the research. However, the CO2 equivalent concentration for 2050 for the scenario A1B is very close to one in RCP8.5, representing a business as usual increase in greenhouse gases, and for the scenario E1 is very close to the RCP4.4, representing a stabilisation scenario. 2.5. Strategies with potential to reduce water stress Here we focus exclusively on practices that contribute to reduce water scarcity and have a clear benefit in climate change adaptation, and we consider only measures that produce additive effects, in order to calculate the stabilisation wedges. The choice of strategies is also guided by four additional factors: (a) the strategies have a proven essential role for global adaptation potential, so the results could be extrapolated to other water scarcity regions; (b) the strategies are linked to the good ecological status targets of the EU WFD; (c) the strategies have a clear impact on the water resource services in rural areas and contribute to understand agricultural adaptation, a key component of the EU Adaptation strategy; and (d) the strategies can be used in all basins included in the study. Here we have selected three strategies that reflect the above criteria: improved system management, improved urban use efficiency and modification of environmental flows. Improved system management allows the possibility of combining the regulation of many reservoirs in larger basins, as opposed to local management, where reservoirs are only managed to satisfy local demands. Improved management

requires the enhancement of the water distribution network, connecting many demand and supply nodes that can be managed as a joint system. As a result of system interconnection, deficit in some regions can be compensated by surplus in other regions and overall reliability is improved. In the improved urban efficiency case, a reference gross per capita withdrawal of 300 l/p·d is taken as reference. This figure was obtained by comparing the data on per capita water withdrawal for urban use from three different data sources: the study performed by the Spanish CEDEX on freshwater in Mediterranean Europe (Estrela et al., 2000) and the data contained in the databases published by World Bank (2015) and AQUASTAT (FAO, 2015). Although reporting of urban water abstraction is inconsistent among European countries due to lack of a uniform methodology, the national figures used in all studies are coincident, as is shown in Fig. 3. Population-weighted average values are 355 l/p·d for the CEDEX study, 312 l/p·d for the World Bank data and 322 l/p·d for the AQUASTAT data. The target for improved urban use efficiency is assumed as 200 l/p·d. Since urban use has priority over irrigation, the improved efficiency of urban use will release resources that could be used in irrigation. The effect is non-linear due to the differences in required reliabilities for both uses. Finally the environmental flow strategy will explore the tradeoffs of changing the allocation to environmental flows. Environmental policy may have a strong effect on water resources because there is a trade-off between water allocated for ecosystems and water availability for economic uses. If future environmental policies emphasise a stronger protection of the environment, more water will be allocated to ecological flows and water availability for agriculture will be reduced. Conversely, if future policies emphasise economic activity, water availability for agriculture may be increased by reducing the allocation to ecological flow. Environmental flows are specified as a given percentile of the marginal distribution of monthly flows. In the reference condition, a 10% percentile is adopted according to the Spanish Technical Instruction for Hydrological Planning (MARM, 2008), which recommends the establishment of an environmental flow

D Water Availability Fraction (WAF, water availability as fraction of mean annual flow) Present time

C B

A

Fig. 3. Per capita water withdrawal for urban use for five representative countries (PT: Portugal, ES: Spain, FR: France, IT: Italy and GR: Greece) according to three datasets: CEDEX (Estrela et al., 2000), WORLD BANK (World Bank, 2015) and AQUASTAT (FAO, 2015).

Medium term Long term Future (2020- Future (20702100) 2050)

Fig. 4. Simplified representation of the water wedge strategies based on the concept of Pacala and Socolow (2004). A: water availability under current management; B: increased water availability due to improved system management; C: increased water availability due to improved urban efficiency; D: increased water availability due to changes in the environmental flows.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

5

Fig. 5. Spatial distribution of per unit runoff reduction between long term (2070–2100) and control (1960–1990) periods for the subbasins considered in the analysis. Results correspond to KNMI (upper left), ETHZ (upper right) and CRNM (lower left) models in the emission scenario A1B and CRNM model (lower right) in the emission scenario E1. Values are shown on the colour scale shown at the right side of each figure, ranging from −1 to 1.

Mean Annual Runoff vs Irrigation Water Availability KNMI CTL1960-1990

Mean Annual Runoff vs Irrigation WaterAvailability ETHZ CTL 1960-1990 1.E+05

Availability for irrigation (hm3/yr)

Availability for irrigation (hm3/yr)

1.E+05

1.E+04

1.E+03

1.E+02

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+04

1.E+03

1.E+02

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

Mean annual runoff (hm3/yr)

Mean annual runoff (hm3/yr)

Mean Annual Runoff vs Irrigation Water Availability CRNM CTL 1960-1990 Availability for irrigation (hm3/yr)

1.E+05

1.E+04

1.E+03

1.E+02

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

Mean annual runoff (hm3/yr) Fig. 6. Mean annual availability for irrigation as a function of mean annual flow for the control period (1960–1990) for models KNMI (upper left), ETHZ (upper right) and CRNM (lower centre). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are presented on a logarithmic scale.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

between the 5% and 15% percentiles of the marginal monthly flow distribution when hydrological methods are used. The Tagus River Basin Management Plan (MAGRAMA-CHT, 2014) included the determination of the environmental flows using an habitat-simulation model (RHYHABSIM) in a limited number of river stretches, in the main river course and in large and small tributaries; for 14 cases that were compared to the results of hydrological methods, 9 were within the abovementioned interval, 3 were over the 15% percentile and the other 2 were below the 5% percentile. With reference environmental flow fixed in the middle of the interval (10% percentile), two future environmental policies have been considered: a) a greater protection of economy in detriment of the environment, where ecological flows are reduced to the 5% percentile, and b) a stronger protection of the environment in detriment of economy, where ecological flows are increased to the 15% percentile. These two environmental policies are compared against the intermediate scenario, where the protection of the environment and of the economy is balanced, represented by the 10% percentile. 2.6. Water wedge approach Here we apply the water adaptation wedge concept to rank the strategies in order to inform environmental and climate policy. Fig. 4 has been prepared to explain this approach and its interpretation. The area A is the base of the plot; it represents the projections of water availability under current management for three time periods: current

Water Availability for Irrigation A1B 2070-2100 (hm3/yr)

Effect of climate change on Irrigation Water Availability KNMI A1BLT - CTL 1.E+04

1.E+03

1.00 0.75

1.E+02

0.50 0.25

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

climate, short term climate change, and long term climate change. The policy target is to maintain in future time periods the same water availability as in the current time period. Three different adaptation policies are assessed: improved system management, improved urban water use efficiency and reduction of environmental flows. The effect of each policy is represented as a wedge that overlaps the base area. The effects of the different policies are accumulated until the current conditions are reached, that is when the gap between current and future water availability is fully covered. The area B represents the increased water availability due to improved system management; the area C represents the increased water availability due to improved urban water use efficiency; and the area D represents the increased water availability due to changes in the environmental flows. Given that the effects of each policy vary depending on the reference condition, the sequence in which water wedges are laid out provides different adaptation alternatives. In this work, policies are applied in order and in a cumulative way. Policy B is implemented first and, if needed, then C and finally D. The application of this approach for different emission scenarios, global climate models, and regional climate models provides a range of results that may support policy choices. 2.7. Limitations There are some major limitations of our approach derived from limitations of the models used, the assumptions, the datasets and the lack of empirical evidence to validate the results. First, our scenarios

Effect of climate change on Irrigation Water Availability ETHZ A1BLT - CTL

Water Availability for Irrigation A1B 2070-2100 (hm3/yr)

6

1.E+04

1.E+03

1.00 0.50 0.25

1.E+01

1.E+00 1.E+00

Effect of climate change on Irrigation Water Availability CRNM A1BLT - CTL

1.E+03

1.00 0.75

1.E+02

0.50 0.25

1.E+01

1.E+02

1.E+03

Water Availability for Irrigation CTL1960-1990 (hm3/yr)

1.E+02

1.E+03

1.E+04

Effect of climate change on Irrigation Water Availability CRNM E1LT - CTL

1.E+04

1.E+01

1.E+01

Water Availability for Irrigation CTL1960-1990 (hm3/yr)

1.E+04

Water Availability for Irrigation A1B 2070-2100 (hm3/yr)

Water Availability for Irrigation A1B 2070-2100 (hm3/yr)

Water Availability for Irrigation CTL 1960-1990 (hm3/yr)

1.E+00 1.E+00

0.75

1.E+02

1.E+04

1.E+03

1.00 0.75

1.E+02

0.50 0.25

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

Water Availability for Irrigation CTL 1960-1990 (hm3/yr)

Fig. 7. Changes of mean annual availability for irrigation from control period (1960–1990) to long term period (2070–2100). Three different models were analysed under the same emission scenario: KNMI-A1B (upper left), ETHZ-A1B (upper right) and CRNM-A1B (lower left). Two different emission scenarios where considered for the same model: CRNM-A1B (lower left) and CRNM-E1 (lower right). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are presented on a logarithmic scale. The main diagonal represents no change (legend 1.00); below, reference lines representing reductions of 25% (0.75), 50% (0.50) and 75% (0.25) are also plotted.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

do not consider changes in water demand under climate change. However, an essential contribution of this modelling approach is to include the monthly distribution of demand and supply in the analysis. The assumption is that this seasonality remains unchanged in the future. There is reasonable evidence to assume that the Mediterranean climate will remain with hot and dry summers and cold and wet winters. Given the range of climate change projections used in this study and available globally (IPCC, 2014), the consequence of our assumption is an underestimation of the potential adverse results. Second, groundwater supplied demand is considered unchanged under climate change scenarios. Third, our list of proposed adaptation policies for water management does not capture the full range of possible adaptation policies to be implemented, particularly since it does not propose technological changes, subsidies, or voluntary market solutions. Fourth, our water management model is limited because it is an oversimplification of the complex hydraulic infrastructure that usually exists in Mediterranean basins, which could be an interesting topic to explore in further research. Additionally, in the context of the water policy model the factors considered are likely to be only the most relevant ones and other important future factors are not considered. Future research is needed to further understand the underlying water availability and adaptation. Fifth, the solutions proposed may not be the most adequate ones for providing safe drinking water, an essential element of water policy (Neumann, 2012).

1.0

Finally, there are limitations in the process of policy development: (a) management choices are based on hydrological processes and thus, these choices are location specific and climate, demands and existing infrastructure determine their management potential; (b) since water sustains environmental health, the environmental targets have to be also regionally evaluated and (c) the aggregated regional potential of the combination of measures has to be defined in order to inform policy choices.

3. Results and discussion 3.1. Spatial runoff changes The results obtained in the analysis are presented in this section. Fig. 5 shows an example of the spatial distribution of impact of climate change on mean annual runoff. The figure presents the expected change of mean annual runoff in the long term scenario (2070–2100) with respect to baseline (1960–1990) for the A1B scenario as computed with the KNMI, ETHZ and CRNM models and for the E1 scenario as computed with the CRNM model. The spatial pattern shows a significant runoff reduction across Southern Europe, with larger impacts in the South West. A similar spatial pattern was observed for other models and emission scenarios.

Effect of climate change on irrigation water availability KNMI CTL (1960-1990)-A1B (2070-2100) Change in Irrigation Water Availability

Change in Irrigation Water Availability

Effect of climate change on irrigation water availability KNMI CTL (1960-1990)-A1B (2070-2100) 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8

1.0 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0

-1.0 -1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

-1.0

1.0

-0.8

-0.6

Effect of climate change on irrigation water availability CRNM CTL (1960-1990)-A1B (2070-2100) 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

Change in Mean Annual Runoff

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Effect of climate change on irrigation water availability CRNM CTL (1960-1990)-A1B (2070-2100) Change in Irrigation Water Availability

1.0

-1.0

-0.4

Change in Mean Annual Runoff

Change in Mean Annual Runoff

Change in Irrigation Water Availability

7

0.6

0.8

1.0

1.0 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Change in Mean Annual Runoff

Fig. 8. Changes of mean annual availability for irrigation from control period (1960–1990) to long term period (2070–2100) as a function of changes in runoff. Three different models were analysed under the same emission scenario: KNMI-A1B (upper left), ETHZ-A1B (upper right) and CRNM-A1B (lower left). Two different emission scenarios where considered for the same model: CRNM-A1B (lower left) and CRNM-E1 (lower right). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

8

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

3.2. Quantification of water stress Water availability was computed for each subbasin applying the WAAPA model, as described in the Methods section. The results obtained in the control period (1960–1990) are shown in Fig. 6. Fig. 6 shows the values of mean annual water availability for irrigation as a function of mean annual runoff in all subbasins analysed. Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are shown for the three models used in the analysis: KNMI, ETHZ and CRNM. There is a large variability in the results, which reflects the variability of hydrologic regimes and regulation capacity across the basins of Southern Europe. Overall, the figure shows that the widely adopted assumption that water availability is a given fraction of mean annual runoff (i.e., 40%) is an oversimplification that is usually misleading. Factors like hydrologic variability or regulation capacity should be accounted for to diagnose water scarcity in the region. In subbasins with small hydrologic variability and very large regulation capacity water availability for irrigation is very similar to the mean annual flow. The fact that in some basins it is even larger may be explained because the volumetric reliability required for irrigation is only 90% and thus it is theoretically possible to satisfy a demand larger than the mean annual flow. On the other hand, there are also basins where water availability for irrigation is only a small fraction of the mean annual flow. This may be explained by very large hydrologic variability, small regulation capacity or very large urban demand.

Water scarcity can be diagnosed by comparing water demand and water availability. In the Mediterranean region water availability is usually the result of significant investment in infrastructure that is normally driven by the potential irrigation demand. Therefore, it can be assumed that current water demand is roughly balanced with water availability. This balance is only broken in basins with structural water deficit or basins with very abundant water resources where regulation is only developed for hydropower. 3.3. Projections of water stress Changes in water availability for irrigation under climate change scenario are shown in Fig. 7. It represents water availability for irrigation in the long term period (2070–2100) as a function of water availability in the control period (1960–1990), for the three models (KNMI, ETHZ and CRNM) in emission scenario A1B and also for the CRNM model in emission scenario E1. The comparison of the results of different models (KNMI, ETHZ and CRNM) under the same scenario (A1B) allows the determination of a range for the possible impacts and for the effects of the adaptation policies. The comparison of the results of the same model (CRNM) under different scenarios (A1B and E1) leads to understand the influence that mitigation policies may have on the need for adaptation strategies. As in Fig. 6, small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins. Some basins present extremely large reduction of water availability in the long term period. If results obtained with

Effect of System Management on Irrigation Water Availability ETHZ - A1B LT 2070-2100

1.E+04

1.E+03 1.00 1.50 1.E+02

2.00 5.00

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

Availability for irrigation under improved management (hm3/yr)

Availability for irrigation under improved management (hm3/yr)

Effect of System Management on Irrigation Water Availability KNMI - A1B LT 2070-2100

1.E+04

1.E+04

1.E+03 1.00 1.50 1.E+02

5.00

1.E+01

1.E+00 1.E+00

3

1.E+04

1.E+03 1.00 1.50 2.00 5.00

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

Availability for irrigation under local management (hm3/yr)

1.E+02

1.E+03

1.E+04

Effect of System Management on Irrigation Water Availability CRNM - E1 LT 2070-2100 Availability for irrigation under improved management (hm3/yr)

Availability for irrigation under improved management (hm3/yr)

Effect of System Management on Irrigation Water Availability CRNM - A1B LT 2070-2100

1.E+01

1.E+01

Availability for irrigation under local management (hm3/yr)

Availability for irrigation under local management (hm /yr)

1.E+02

2.00

1.E+04

1.E+03 1.00 1.50

1.E+02

2.00 5.00

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

Availability for irrigation under local management (hm3/yr)

Fig. 9. Effect of system management on irrigation water availability. Availability under improved management as a function of availability under local management for long term period (2070–2100). Three different models were analysed under the same emission scenario: KNMI-A1B (upper left), ETHZ-A1B (upper right) and CRNM-A1B (lower left). Two different emission scenarios where considered for the same model: CRNM-A1B (lower left) and CRNM-E1 (lower right). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are presented on a logarithmic scale. The main diagonal represents no change (legend 1.00); above, reference lines representing improvements of 50% (1.50), 100% (2.00) and 400% (5.00) are also plotted.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

different models are compared, large model uncertainty is apparent. Under the A1B scenario, ETHZ shows the smallest reduction of water availability, CRNM the greatest reduction and KNMI the largest variability. If results obtained with different emission scenarios are compared the A1B scenario produces larger reductions than the E1 scenario. The reductions seen in Fig. 7 are largely due to reductions in runoff and changes in hydrologic variability, but basins are affected differently depending on their regulation capacity. This point is illustrated with Fig. 8, where changes of mean annual availability for irrigation from the control period (1960–1990) to long term period (2070–2100) are shown as a function of changes in runoff. This figure shows clearly how changes in water availability are not the same as changes in runoff suggesting the inadequacy of methodologies that estimate availability as a fraction of mean annual runoff. Basins with the same reduction of runoff experience different reductions in availability as a result of changes in hydrologic variability and their different regulation capacities. For instance, the fact that for KNMI the changes in water availability are generally smaller than changes in runoff suggests that this model predicts less hydrologic variability for the long term scenario. Overall, results presented in Figs. 7 and 8 support the need to include water management in the assessment of climate change impacts on water resources. 3.4. Facing water stress The projected reductions in water availability for future scenarios suggest an intensification of water stress in many regions of Mediterranean

Europe. This will require adaptive management to face water stress and compensate for the reductions. The effect of three management measures to face water stress is quantified in this section: improved system management, improved urban use efficiency and modification of environmental flows. The potential effect of improved system management is shown in Fig. 9. The improvement is quantified for the three models under emission scenario A1B and the CNRM model under emission scenario E1. In addition to the main diagonal (no change), lines representing improvements of 50%, 100% and 400% are shown. Although in many basins the potential effect of improved system management is small, in some basins it can be significant, and in a few cases it may even be dramatic, increasing water availability by twofold. As the case of impact, the effect is largest for model KNMI and for emission scenario A1B. In general, the effect is larger for basin with larger availability, suggesting a factor of scale. The potential effect of improved water use efficiency in urban areas is shown in Fig. 10. The effect is quantified for the three models under emission scenario A1B and the CNRM model under emission scenario E1. The potential effect of improved urban use is small. In general, the effect is larger for basin with smaller availability for irrigation. The results of the modification of the environmental flow under policy a) (flow reduction) are shown in Fig. 11 and the results under policy b) (flow increase) are shown in Fig. 12. Changes in ecological flows have a moderate impact on water availability. The model KNMI shows the greatest sensitivity. The reduction of the percentile from 10% to 5%

Effect of urban efficiency on Irrigation Water Availability A1B (2070-2100) - KNMI - Per capita water use

Effect of urban efficiency on Irrigation Water Availability A1B (2070-2100) - ETHZ - Per capita water use 1.E+04

1.E+03

200l/pd

1.E+02

1.00 1.50 2.00 5.00

1.E+01

1.E+01

1.E+02

1.E+03

Availability with 200 l/pd (hm3/yr)

Availability with 200 l/pd (hm3/yr)

1.E+04

1.E+00 1.E+00

1.E+03

200l/pd

1.E+02

2.00 5.00

1.E+01

1.E+00 1.E+00

1.E+04

1.E+01

1.E+02

1.E+03

1.E+04

Availability with 300 l/pd (hm3/yr)

Availability with 300 l/pd (hm /yr)

Effect of urban efficiency on Irrigation Water Availability E1 (2070-2100) - CRNM - Per capita water use

Effect of urban efficiency on Irrigation Water Availability A1B (2070-2100) - CRNM - Per capita water use

1.E+04

1.E+03

200l/pd

1.E+02

1.00 1.50 2.00 5.00

1.E+01

1.E+01

1.E+02

1.E+03

Availability with 300 l/pd (hm3/yr)

1.E+04

Availability with 200 l/pd (hm3/yr)

1.E+04

Availability with 200 l/pd (hm3/yr)

1.00 1.50

3

1.E+00 1.E+00

9

1.E+03

200l/pd

1.E+02

1.00 1.50 2.00 5.00

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

Availability with 300 l/pd (hm3/yr)

Fig. 10. Effect of urban efficiency on irrigation water availability. Availability under improved urban efficiency as a function of availability under reference urban efficiency for long term period (2070–2100). Three different models were analysed under the same emission scenario: KNMI-A1B (upper left), ETHZ-A1B (upper right) and CRNM-A1B (lower left). Two different emission scenarios where considered for the same model: CRNM-A1B (lower left) and CRNM-E1 (lower right). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are presented on a logarithmic scale. The main diagonal represents no change (legend 1.00); above, reference lines representing improvements of 50% (1.50), 100% (2.00) and 400% (5.00) are also plotted.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

10

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

would increase water availability by 50% in many basins. Results obtained in other models show less sensitivity. Overall, the sensitivity to a reduction of environmental flows is larger than the sensitivity to an increase. This is explained because of its effect on very low flows. Environmental flows have the highest priority and thus they receive all water allocation during very low flows, limiting the amount of water than can be supplied. 3.5. A wedge approach for prioritising strategies The results obtained in the previous analysis are summarised in Fig. 13 for 7 representative river basins: Thrace, Ebro, Duero, Tagus, Guadalquivir, Rhone and Po, using the wedge approach. To facilitate comparison, water availability values have been normalized by mean annual flows. Four graphs are presented for each basin, each corresponding to a model and emission scenario. In each graph, water availability has been represented as a function of time period: control (1960–1990), short term (2020–2050) and long term (2070–2100). The management objective is to maintain in future periods the same water availability as in the control period. Values for the reference management are represented in grey, and they usually imply a reduction of water availability in future periods. This tendency is corrected by applying several policy alternatives that increase water availability in future scenarios. The blue wedge represents increased water availability values that can be obtained through improved system management.

The red wedge represents increased water availability values that can be obtained through improved urban use efficiency. The orange wedge represents increased water availability values that can be obtained through reduced ecological flows. Policies have been applied as necessary, only to reach the fixed target of maintaining water availability with respect to the control period. The results show striking differences between basins. The water availability fraction varies from 0.5 in Thrace to less than 0.05 in Po. These differences are due to the different hydrologic regimes and regulation capacities across basins. The impacts of climate change also vary across basins. The most significant impacts are produced in the Guadalquivir basin and are due mostly to a dramatic reduction in mean annual runoff. The Rhone basin presents the least impacts, due to a very mild reduction in mean annual runoff. There are also striking differences between models for the same emission scenario. ETHZ produces the largest water availability fraction, while CRNM produces the least water availability. These differences are due to the different implementations of the hydrologic cycle across models, which produce different seasonal cycles that impact the hydrologic regime. This is a weakness of current regional climate models that should be improved in the future. In many cases, the policy measures analysed are enough to compensate for the projected reduction of water availability in future periods. This is the case of Po, Rhone and, except for one scenario, Ebro and Duero. These basins have relatively abundant water resources compared to their demands and thus offer a range of policy alternatives to

Effect of ecological flows on Irrigation Water Availability A1B (2070-2100) - KNMI - Percentile

Effect of ecological flows on Irrigation Water Availability A1B (2070-2100) - CRNM - Percentile 1.E+04

1.E+03

1.00

1.E+02

1.50 2.00 5.00

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

Availability with 5% percentile (hm3/yr)

Availability with 5% percentile (hm3/yr)

1.E+04

1.E+03

5.00

1.E+01

3

1.E+04

1.E+03

1.00

1.E+02

1.50 2.00 5.00

1.E+01

1.E+02

1.E+03

Availability with 10% percentile (hm3/yr)

1.E+02

1.E+03

1.E+04

Effect of ecological flows on Irrigation Water Availability E1 (2070-2100) - CRNM - Percentile

1.E+04

Availability with 5% percentile (hm3/yr)

Availability with 5% percentile (hm3/yr)

Effect of ecological flowson Irrigation Water Availability A1B (2070-2100) - ETHZ - Percentile

1.E+01

1.E+01

Availability with 10% percentile (hm3/yr)

Availability with 10% percentile (hm /yr)

1.E+00 1.E+00

1.50 2.00

1.E+00 1.E+00

1.E+04

1.00

1.E+02

1.E+04

1.E+03

1.00

1.E+02

1.50 2.00 5.00

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

Availability with 10% percentile (hm3/yr)

Fig. 11. Effect of ecological flow reduction. Availability under reduced ecological flows as a function of availability under reference ecological flows for long term period (2070–2100). Three different models were analysed under the same emission scenario: KNMI-A1B (upper left), ETHZ-A1B (upper right) and CRNM-A1B (lower left). Two different emission scenarios where considered for the same model: CRNM-A1B (lower left) and CRNM-E1 (lower right). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are presented on a logarithmic scale. The main diagonal represents no change (legend 1.00); above, reference lines representing improvements of 50% (1.50), 100% (2.00) and 400% (5.00) are also plotted.

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

Effect of ecological flows on Irrigation Water Availability A1B (2070-2100) - ETHZ - Percentile

1.E+04

1.E+03

1.00

1.E+02

0.75 0.50 0.25

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

Availability with 15% percentile (hm3/yr)

Availability with 15% percentile (hm3/yr)

Effect of ecological flows on Irrigation Water Availability A1B (2070-2100) - KNMI - Percentile

1.E+04

1.E+04

1.E+03

0.50 0.25

1.E+01

1.E+00 1.E+00

1.00

1.E+02

0.75 0.50 0.25

1.E+01

1.E+04

Availability with 10% percentile (hm3/yr)

Availability with 15% percentile (hm3/yr)

Availability with 15% percentile (hm3/yr)

1.E+03

1.E+03

1.E+01

1.E+02

1.E+03

1.E+04

Effect of ecological flows on Irrigation Water Availability E1 (2070-2100) - CRNM - Percentile

1.E+04

1.E+02

0.75

Availability with 10% percentile (hm3/yr)

Effect of ecological flows on Irrigation Water Availability A1B (2070-2100) - CRNM - Percentile

1.E+01

1.00

1.E+02

Availability with 10% percentile (hm3/yr)

1.E+00 1.E+00

11

1.E+04

1.E+03

1.00

1.E+02

0.75 0.50 0.25

1.E+01

1.E+00 1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

Availability with 10% percentile (hm3/yr)

Fig. 12. Effect of ecological flow increase. Availability under increased ecological flows as a function of availability under reference ecological flows for long term period (2070–2100). Three different models were analysed under the same emission scenario: KNMI-A1B (upper left), ETHZ-A1B (upper right) and CRNM-A1B (lower left). Two different emission scenarios where considered for the same model: CRNM-A1B (lower left) and CRNM-E1 (lower right). Small, lighter dots represent results in intermediate subbasins, while larger, darker dots represent results in the final basins of the river basin districts. Results are presented on a logarithmic scale. The main diagonal represents no change (legend 1.00); below, reference lines representing reductions of 25% (0.75), 50% (0.50) and 75% (0.25) are also plotted.

adapt to climate change. Thrace and Tagus are in an intermediate situation. Policy can compensate for the reduction of water availability in the A1B scenario for models KNMI and ETHZ, but results of the CRNM model show that the climate change impact cannot be mitigated by the suite of analysed policies alone. If predictions of this model are correct, water availability may be limited in the future and irrigation activity will have to be reduced. The Guadalquivir basin shows strong impacts in three of the four scenarios, with very dramatic decreases in water availability and little scope for action. This basin is facing the strongest challenges of climate change adaptation. The key findings can be summarised as the recognition of simple systematic strategies which may be used in wedges. These behaviours may provide guidance towards broader fundamental adaptation to climate change in multi-scale analyses in water resource research and policy studies. 4. Conclusions Our analysis advances our knowledge of differing public support for climate change adaptation policy by providing increased comprehension of water vulnerability, understanding the effect of measures spatially in support of adaptation policies in a geographically and socially diverse region. In our analysis the climate change scenarios are not the main drivers of differential regional choices to reduce vulnerability, as we would

have expected. This reflects the empirical evidence that choice in water management is driven by both cultural and rational approaches (i.e., the existing demand). The water infrastructure has evolved relatively well to adjust to regional characteristics, guided by water demand, technology and change in behaviour. In contrast, policy may evolve radically differently due to tipping points of vulnerability thresholds derived from changes in climate and policy. In addition, flexible adaptation may be guided in the near term by improvements in monitoring and early warning systems (Pozzi et al., 2013). Our study of water vulnerability in the Mediterranean basins of Europe shows that fewer than 20% of the area may have the right policy to face climate change. Therefore, there is considerable potential for improving water management in support of adaptation in the region. Our study has shown that, while reduction of agricultural demand is an important factor to derive new policy in support of adaptation, strategies that focus on urban technical efficiency and improved system management policies play essential roles. Vulnerability in the future is affected by local determinants, which indirectly influence individuals' support for adaptation policies. To this end, further evaluations of public choices seem to be particularly appropriate. According to the results, it could be assumed that in each basin it is necessary to address social motivations and barriers in order to promote the most effective adaptation strategies. Thus, policy can play an essential role in improving the effect of the choices by interacting with science results so some policy recommendations could be proposed to

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx Thrace (ETZHA1B)

2070-2100

1960-1990

0,4 0,3 0,2 0,1 0 2020- 2050

2070-2100

0,4 0,3 0,2 0,1 0 1960-1990

Water availability fraction (WAF)

Water availability fraction (WAF)

0,4 0,3 0,2 0,1 0 2070-2100

0,3 0,2 0,1 0

0,1 0 2070-2100

0,4 0,3 0,2 0,1 0 1960-1990

0,3 0,2 0,1 0 2070-2100

0,3 0,2 0,1 0 1960-1990

Water availability fraction (WAF)

0,2 0,1 0

0,3 0,2 0,1 0

2020-2050

2070-2100

0,5 0,4 0,3 0,2 0,1 0 1960-1990

0,3 0,2 0,1 0 2070-2100

0,4 0,3 0,2 0,1 0 2020-2050

2070-2100

0,4 0,3 0,2 0,1 0

Water availability fraction (WAF)

Po (ETZHA1B)

0,5 0,4 0,3 0,2 0,1 0 2070-2100

0,4 0,3 0,2 0,1 0 2020-2050

2070-2100

0,4 0,3 0,2 0,1 0 2020-2050

0,2 0,1 0 2020-2050

2070-2100

0,5 0,4 0,3 0,2 0,1 0 2020-2050

2070-2100

0,5 0,4 0,3 0,2 0,1 0 2020-2050

2070-2100

0,5 0,4 0,3 0,2 0,1 0 2020-2050

2070-2100

0,5 0,4 0,3 0,2 0,1 0 1960-1990

2020-2050

2070-2100

Po(CRNME1)

0,5

1960-1990

0,3

1960-1990

2070-2100

Po (CRNMA1B)

0,5

1960-1990

2020-2050

0,4

Rhone(CRNME1)

0,5

1960-1990

2070-2100

Guadalquivir (CRNME1)

2070-2100

Rhone (CRNMA1B)

0,5

1960-1990

2020-2050

2020-2050

0,5

1960-1990

2070-2100

Guadalquivir (CRNMA1B)

Water availability fraction (WAF)

0,4

2020-2050

0

Tagus (CRNME1)

0,4

1960-1990

0,1

1960-1990

2070-2100

0,5

Rhone (ETZHA1B) Water availability fraction (WAF)

Water availability fraction (WAF)

2070-2100

0,4

Po (KNMIA1B) Water availability fraction (WAF)

2020-2050

0,5

Rhone (KNMIA1B)

0,5

2020- 2050

0,3

2020-2050

0,2

Duero (CRNME1)

0,4

1960-1990

0,3

1960-1990

2070-2100

0,5

Guadalquivir (ETZHA1B) Water availability fraction (WAF)

Water availability fraction (WAF)

0,4

1960-1990

0

Tagus (CRNMA1B)

0,5

Guadalquivir (KNMIA1B)

0,5

2020- 2050

2070-2100

Water availability fraction (WAF)

0,2

1960-1990

2020-2050

Water availability fraction (WAF)

0,3

2020- 2050

0,1

Duero (CRNMA1B)

0,4

1960-1990

Water availability fraction (WAF)

Water availability fraction (WAF)

0,4

1960-1990

0,2

2020-2050

0,4

Ebro (CRNME1)

0,3

1960-1990

0,5

1960-1990

2070-2100

0,4

Tagus (ETZHA1B)

0,5

2020- 2050

2070-2100

0,5

Tagus (KNMIA1B)

1960-1990

2020-2050

2020-2050

0,5

Duero (ETZHA1B)

0,5

2020- 2050

0

Ebro (CRNMA1B)

0,5

Duero (KNMIA1B)

1960-1990

0,1

1960-1990

Water availability fraction (WAF)

0,5

1960-1990

2070-2100

0,2

Ebro (ETZHA1B) Water availability fraction (WAF)

Water availability fraction (WAF)

Ebro (KNMIA1B)

2020-2050

Water availability fraction (WAF)

2020- 2050

0

Water availability fraction (WAF)

1960- 1990

0,1

0,3

Water availability fraction (WAF)

0

0,2

Water availability fraction (WAF)

0,1

0,3

0,4

Water availability fraction (WAF)

0,2

0,4

0,5

Water availability fraction (WAF)

0,3

0,5

Water availability fraction (WAF)

0,4

Thrace (CRNME1)

Thrace (CRNMA1B) Water availability fraction (WAF)

Water availability fraction (WAF)

Water availability fraction (WAF)

Thrace (KNMIA1B) 0,5

Water availability fraction (WAF)

12

2070-2100

0,5 0,4 0,3 0,2 0,1 0 1960-1990

2020-2050

2070-2100

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106

L. Garrote et al. / Science of the Total Environment xxx (2015) xxx–xxx

facilitate climate change adaptation. The European Environment Agency (2012) and the EU adaptation strategy (EU 2013) recommend education and awareness raising as overarching adaptation measures to climate change. These types of policies aim to enhance the environmental and climate change awareness of society which would positively influence societal support for adaptation policies (Semenza et al., 2008) or benefit from economic choices (Roson and Sartori, 2015). Finally, economic losses are a fundamental barrier to the support of adaptation policies that reduce agricultural water supply. So any irrigation reduction would suppose a personal economic loss for rural communities, which is difficult to be addressed by the Rural Development component of the CAP (Rey et al., 2014). Therefore, it is essential to promote technologies that use and re-use the water more efficiently, to promote crops better adjusted to the adverse effects of climate change such as those which consume less water as well as to attempt to maintain farmer's current wellbeing by other diversified economic activities. Acknowledgements This research was supported by the European Commission WasserMed project (Project reference 244255, funded under FP7ENVIRONMENT) and the European Commission BASE project (Grant Agreement No. 308337, funded under FP7-ENVIRONMENT). References Alcamo, J., Döll, P., Kaspar, F., Siebert, S., 1997. Global Change and Global Scenarios of Water Use and Availability: An Application of Water GAP1.0. Center for Environmental Systems Research (CESR), University of Kassel, Germany. Boithias, L., Acuña, V., Vergoñós, L., Ziv, G., Marcé, R., Sabater, S., 2014. Assessment of the water supply:demand ratios in a Mediterranean basin under different global change scenarios and mitigation alternatives. Sci. Total Environ. 470, 567–577. Chávez-Jiménez, A., Lama, B., Garrote, L., Martín-Carrasco, F., Sordo-Ward, A., Mediero, L., 2013. Characterisation of the sensitivity of water resources systems to climate change. Water Resour. Manag. 27 (12), 4234–4258. Chávez-Jiménez, A., Granados, A., Garrote, L., Martín-Carrasco, F., 2015. Adapting water allocation to irrigation demands to constraints in water availability imposed by climate change. Water Resour. Manag. 29 (5), 1413–1430. Estrela, T., Marcuello, C., Dimas, M., 2000. Las aguas continentales en los países mediterráneos de la Unión Europea. CEDEX (Available at http://hispagua.cedex. es/sites/default/files/aguas_continentales_union_europea.pdf). Falkenmark, M., 1986. Fresh water: time for a modified approach. Ambio 15, 192–200. FAO, 2015. Main AQUASTAT Country Database. Retrieved 2014-04-03 from. http://www. fao.org/nr/water/aquastat/data/query/index.html?lang=en. Garrote, L., Iglesias, A., Martín-Carrasco, F., Mediero, L., 2011. WAAPA: a model for water availability and climate change adaptation policy analysis. Proceedings of the VI EWRA International Symposium — Water Engineering and Management in a Changing Environment. University of Catania, Catania, Italy. Garrote, L., Iglesias, A., Granados, A., Mediero, L., Martin-Carrasco, F., 2015. Quantitative assessment of climate change vulnerability of irrigation demands in Mediterranean Europe. Water Resour. Manag. 29, 325–338. Gleick, P.H., 2003. Global freshwater resources: soft-path solutions for the 21st century. Science 302 (5650), 1524–1528. Hanasaki, N., Fujimori, S., Yamamoto, T., Yoshikawa, S., Masaki, Y., Hijioka, Y., Kainuma, M., Kanamori, Y., Masui, T., Takahashi, K., Kanae, S., 2013. A global water scarcity assessment under Shared Socio-economic Pathways — part 2: water availability and scarcity. Hydrol. Earth Syst. Sci. 17, 2393–2413. Iglesias, A., Garrote, L., Flores, F., Moneo, M., 2007. Challenges to manage the risk of water scarcity and climate change in the Mediterranean. Water Resour. Manag. 21, 775–788. Iglesias, A., Garrote, L., Diz, A., Schlickenrieder, J., Martin-Carrasco, F., 2011. Rethinking water policy priorities in the Mediterranean Region in view of climate change. Environ. Sci. Pol. 14, 744–757. Iglesias, A., Garrote, L., Quiroga, S., Moneo, M., 2012. From climate change impacts to the development of adaptation strategies: challenges for agriculture in Europe. Clim. Chang. 112, 143–168. IPCC, 2014. Climate Change 2014: impacts, adaptation, and vulnerability. Part A: global and sectoral aspects. In: Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J.,

13

Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S., Levy, A.N., MacCracken, S., Mastrandrea, P.R., White, L.L. (Eds.), Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1–32. Ludwig, R., Roson, R., Zografos, C., Kallis, G., 2011. Towards an inter-disciplinary research agenda on climate change, water and security in Southern Europe and Neighbouring countries. Environ. Sci. Pol. 14, 794–803. MAGRAMA-CHT, 2014. Plan Hidrológico de la parte española de la Demarcación Hidrográfica del Tajo (Ciclo de planificación 2009–2015). Ministerio de Agricultura, Alimentación y Medio Ambiente, Confederación hidrográfica del Tajo. MARM, 2008. Instrucción de Planificación Hidrológica. Ministerio de Medio Ambiente y Medio Rural y Marino, Orden ARM/2656/2008. Meigh, J.R., McKenzie, A.A., Sene, K.J., 1999. A grid-based approach to water scarcity estimates for Eastern and Southern Africa. Water Resour. Manag. 13, 85–115. Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., van Vuuren, D.P., et al., 2010. The next generation of scenarios for climate change research and assessment. Nature 463 (7282), 747–756. Nakiçenoviç, N., Alcamo, J., Davis, G., de Fries, B., Fenhann, J., Gaffin, S., Gregory, K., Grübler, A., Jung, T.Y., Kram, T., La Rovere, E.L., Michaelis, L., Mori, S., Morita, T., Pepper, W., Pitcher, H., Price, L., Raihi, K., Roehrl, A., Rogner, H.-H., Sankovski, A., Schlesinger, M., Shukla, P., Smith, S., Swart, R., von Rooijen, S., Victor, N., Dadi, Z. (Eds.), 2000. Emissions Scenarios. A Special Report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge (599 pp.). Neumann, M.B., 2012. Comparison of sensitivity analysis methods for pollutant degradation modelling: a case study from drinking water treatment. Sci. Total Environ. 433, 530–537. Pacala, S., Socolow, R., 2004. Stabilization wedges: solving the climate problem for the next 50 years with current technologies. Science 305 (5686), 968–972. Pozzi, W., Sheffield, J., Stefanski, R., Cripe, D., Pulwarty, R., Vogt, J.V., Heim Jr., R.R., Brewer, M.J., Svoboda, M., Westerhoff, R., Van Dijk, A., Lloyd-Hughes, B., Pappenberger, F., Werner, M., Dutra, E., Wetterhall, F., Wagner, W., Schubert, S., Mo, K., Nicholson, M., Bettio, L., Nunez, L., van Beek, R., Bierkens, M., Goncalves de Goncalves, L.G., Zell, Gerd, de Mattos, J., Lawford, R., 2013. Towards global drought early warning capability: expanding international cooperation for the development of a framework for global drought monitoring and forecasting. Bull. Am. Meteorol. Soc. 94 (6), 776–785. Quevauviller, P., Balabanis, P., Fragakis, C., Weydert, M., Oliver, M., Kaschl, A., Arnold, G., Kroll, A., Galbiati, L., Zaldivar, J.M., Bidoglio, G., 2005. Science-policy integration needs in support of the implementation of the EU Water Framework Directive. Environ. Sci. Pol. 8 (3), 203–211. Rey, D., Garrido, A., Calatrava, J., 2014. The water markets in Spain: moving towards 21st century mechanisms and approaches with 20th century regulations. In: Easter, W., Huang, Q. (Eds.), Water Markets for the 21st. Century: What Have We Learned? Springer, pp. 127–147. Roson, R., Sartori, M., 2015. System-wide implications of changing water availability and agricultural productivity in the Mediterranean economies. Water Econ. Policy 1, 1–30. Schewe, J., Heinke, J., Gerten, D., Haddeland, I., Arnell, N.W., Clarke, D.B., Dankers, R., Eisner, S., Fekete, B.M., Colón-González, F.J., Gosling, S.N., Kim, H., Liu, X., Masaki, Y., Portmann, F.T., Satoh, Y., Stacke, T., Tang, Q., Wada, Y., Wisser, D., Albrecht, T., Frieler, K., Piontek, F., Warszawski, L., Kabatt, P., 2014. Multimodel assessment of water scarcity under climate change. Proc. Natl. Acad. Sci. U. S. A. 111, 3245–3250. Semenza, J.C., Hall, D.E., Wilson, D.J., Bontempo, B.D., Sailor, D.J., George, L.A., 2008. Public Perception of Climate Change Voluntary Mitigation and Barriers to Behavior Change. Am J Prev Med 35 (5), 479–487. Sullivan, C.A., 2002. Calculating a water poverty index. World Dev. 30, 1195–1210. UN, 2003. Water for People, Water for Life: UN World Water Development Report (WWDR). UNESCO/Berghahn Books/UN. UN, 2013. United Nations, Department of Economic and Social Affairs, Population Division world population prospects. The 2012 Revision, Key Findings and Advance Tables Working Paper No. ESA/P/WP.227. Victoria, F.B., ViegasFilho, J.S., Pereira, L.S., Teixeira, J.L., Lanna, A.E., 2005. Multi-scale modeling for water resources planning and management in rural basins. Agric. Water Manag. 77, 4–20. Vörösmarty, C.J., 2002. Global water assessment and potential contributions from earth system science. Aquat. Sci. 64, 328–351. Vörösmarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2000. Global water resources: vulnerability from climate change and population growth. Science 289, 284–288. Wada, Y., Gleeson, T., Esnault, L., 2014. Wedge approach to water stress. Nat. Geosci. 7, 615–617. Wallace, J.S., 2000. Increasing agricultural water use efficiency to meet future food production. Agric. Ecosyst. Environ. 82, 105–119. World Bank, 2015. World Bank World Development Indicators Country Database. Retrieved 2014-04-03 from. http://wdi.worldbank.org/table/3.5.

Fig. 13. Wedge approach to water management for selected river basins. Water availability fraction (as a fraction of mean annual flow) for 7 basins under standard management (grey), improved management (blue), improved urban efficiency (red) and reduced ecological flows (orange) for long term period (2070–2100). Three different models where analysed under the same emission scenario: KNMI-A1B (left), ETHZ-A1B (centre left) and CRNM-A1B (centre right). Two different emission scenarios where considered for the same model: CRNM-A1B (centre right) and CRNM-E1 (right).

Please cite this article as: Garrote, L., et al., Strategies to reduce water stress in Euro-Mediterranean river basins, Sci Total Environ (2015), http:// dx.doi.org/10.1016/j.scitotenv.2015.04.106