Final Report Summary - DROUGHT-R&SPI (Fostering European Drought Research and Science-Policy Interfacing)
Executive Summary:
Drought Dialogue Fora at the pan-European (pan-EU DDF) and the Case Study scale (CS DDF) were held from the start and served as excellent platforms through facilitated discussions among stakeholders, policy-makers and the consortium: (i) to collect information on factors that govern vulnerability and risk to drought, (ii) to interpret options proposed in EC policy documents, (iii) to test and further shape developed methods (e.g. vulnerability, risk reduction) to identify drought-sensitive in Europe, and (iv) to communicate on past and future natural hazard.
A European Drought Reference database (EDR) was established that displays historical drought as a natural hazard (1958–2009). Trend studies using recent instrumental data confirmed the wetting trend in northern Europe and a drying trend in southern Europe. Time series of historic drought beyond the instrumental record (1500-1950) derived from proxy data (e.g. archives) reveal that frequency of dry episodes has not increased in several European regions. Understanding has improved about to what degree observed climatic trends can be attributed to changes in atmospheric circulation changes (i.e. trends in circulation type frequencies, anthropogenic forcing). Historic drought information (e.g. EDR) was used to test performance of large-scale hydrological models that reveal that ensembles of multi hydrological models are required for impact studies (climate change). Several modelling experiments using different IPCC climate output assessed climate change impact on drought. Large spreads over all time frames are observed, but projected droughts are more robust by the end of 21st century in multiple European regions. Different research lines were followed to improve drought forecasting, essential for early warning. Links of antecedent ocean-atmosphere variation to large-scale drought characteristics show that atmospheric bridge connecting North Atlantic SSTs to drought development is too complex to be described solely by NAO. Improved understanding of drought generating processes led to a comprehensive drought typology that is essential for credible multi-monthly to seasonal drought forecasting. Catchment storage is essential to explain the connection between lagged cumulative precipitation deficits and streamflow drought.
Policy recommendations for future drought risk reduction were identified for the six selected Case Study areas (CS) based upon: (i) collated past drought hazards and responses, (ii) anticipated future drought impacts and vulnerability, (iii) assessment of alternative options, (iii) recommendations of CS DDFs, and (iv) EU and international guidance documents. Factors that either reduce or increase drought vulnerability and risk are given, which due to context-specific conditions differ per CS. The CS studies conclude that an 'auto-pilot' approach to drought risk mitigation is impossible: even in a particular area, because drought intensity and impacts are not the same and particularities of potential measures may change from one event to another.
The European Drought Impact Report Inventory (EDII, publicly available through the EDC website) was used for impact analysis to identify drought sensitive areas in Europe. EDII contains categorized data on drought impacts derived from a variety of text sources (about 5000 reports). Impact reports on agriculture, water supply, water quality and freshwater ecosystems dominate, although region-specific as well as event-specific differences and temporal changes occur. Some sectors (e.g. rainfed agriculture) respond to short-term drought, whereas other sectors (water supply, freshwater ecosystems, energy and industry) relying on groundwater and river flow are more sensitive to long-duration droughts. Some regions (e.g. the Mediterranean) are more adapted through reservoir storage and irrigated agriculture. Thresholds in drought indicators were found by combining EDII with the natural hazard database (EDR). Based on these threshold and the assumption that drought impacts can serve as a proxy for vulnerability, the Likelihood of Impact Occurrence (LIO, varies between 0 and 1) was introduced to generate pan-European drought risk maps for different sectors and severity of the natural hazard. Spatial variations of drought risk occur. For moderate drought (hazard), risk is highest in the Mediterranean, in particular for agriculture, and in the densely populated areas of central Europe, for public water supply. In addition to the LIO approach, pan-European vulnerability to drought has been elaborated through a detailed factor-based vulnerability assessment. The three vulnerability components (exposure, sensitivity, adaptive capacity) were broken down in 19 individual factors. Maps per factor were stepwise aggregated and eventually to a pan-European map showing drought vulnerability (NUTS-2 scale. Assessment of drought vulnerability and associated risk at pan-European level involves compromises because these are different for individuals, sectors and regions, depending on several biophysical and socio-economic factors.
Because of the complex nature of the drought phenomenon, one single quantification of drought (indicator) does not exist that satisfies all monitoring and early warning requirements. Best predictive indicators for particular drought impacts depend on sector and on sectorial management practices and need to be defined with stakeholders. At the detailed, operational scale use of tailored indicator systems in a combined monitoring-modeling-decision support system framework are a prerequisite to communicate with stakeholders about ongoing and forecasted hazards, usually manifold impacts, potential measures to be taken, real time drought management and pro-active drought planning.
Project Context and Objectives:
The DROUGHT-R&SPI project aimed to reduce future Europe’s vulnerability and risk to drought by innovative in-depth studies that combine drought investigations in six case study areas in water-stressed regions (river basin and national scale) with drought analyses at the pan-European scale. Knowledge transfer across these scales is paramount because vulnerability is context-specific (e.g. physical, environmental, socio-economic, cultural, legal, institutional), which requires analyses on detailed scales, whereas international policies and drought-generating climate drivers and land surface processes are operating on large scales. The project planned to adopt Science-Policy Interfacing at the various scales, by establishing Case Study Dialogue Fora and a pan-Europe Dialogue Forum, which should ensure that the research was well integrated into the policy-making from the start of the project onwards. The study intended to foster a better understanding of past droughts (e.g. underlying processes, occurrences, environmental and socio-economic impacts, past responses), which then is anticipated to contribute to the assessment of drought hazards and potential vulnerabilities in the 21th C. Methodologies for early drought warning at the pan-European scale were proposed to be developed, which aim to improve forecasting of a suite of interlinked physical and drought impact indicators. This intends to increase drought preparedness, and to identify and implement appropriate Disaster Risk Reduction measures. The project meant to result through combined drought studies at different scales to the identification of drought-sensitive regions and sectors across Europe, which likely will lead to a more thorough implementation of the EU Water Framework Directive, particularly by further developing of methodologies for Drought Management Plans at different scales (incl. EU level). The work was anticipated to be linked with the European Drought Centre (EDC) ensuring that the outcome has been consolidated beyond the project’ lifetime.
The main objectives of Drought-R&SPI were:
1. Foster a better understanding of past droughts (underlying processes, occurrences, incl. frequencies, severities and scales) and in particular to investigate driving factors and characteristics of the most extreme historic events at the pan-European scale, which will be validated against observed natural hazards at the small scale (case studies);
2. Provide an in-depth understanding of the range of pertinent environmental and socio-economic impacts triggered by past drought hazards in different contexts, taking into account economic, political, and social factors at a small scale in water-stressed regions (case studies) and considering the large-scale drivers (transfer of knowledge and information across scales);
3. Evaluate past responses to drought events at the local, river basin and national scale (case studies), and identify best practice examples and lessons learnt in alleviating drought impacts and reducing associated risks from different local environments, which are crucial for the development of methodologies for drought management planning;
4. Develop and test an innovative suite of drought indicators that integrate physical, impact and vulnerability indices, which will address different spatio-temporal scales, interaction across scales and water-related sectors, and that will build upon: (i) the comprehensive knowledge obtained in different contexts at the detailed scale (case studies), and (ii) exploration of the link between impact records (socio-economic, environmental), management factors and vulnerability thresholds across Europe. Development and testing will be linked and shared with research done through the European Drought Observatory (EDO) at the EC Joint Research Centre (JRC) and Waterbase, WSDIS and WISE-RTD (European Environmental Agency, EEA);
5. Develop an innovative methodology for early warning (monitoring and forecasting) of drought at the pan-European scale using the developed European Drought Impact Inventory (EDII database) and suite of indicators, which will help to: (i) to increase drought preparedness, and (ii) to identify and implement appropriate Disaster Risk Reduction (DRR) measures. Drought-R&SPI will share the obtained knowledge and information with the JRC, which are designing and developing an operational version;
6. Establish Drought Dialogue Fora at different scales, i.e. Science-Policy Interfacing, (i) to identify and evaluate, in close collaboration with the Case Study Drought Dialogue Fora, potential responses for drought risk mitigation, taking into account the drought hazard, its impacts, vulnerabilities and desired development goals, strategies and relevant trade-offs, and (ii) to recognize and assess, in close dialogue with the pan-European Drought Dialogue Forum through an iteration process with feedback loops, how EU and other international policies affect potential responses at the small scale (case studies). Finally, how these together with large-scale impact maps and vulnerabilities will support design, development or implementation of drought-related international policies and planning, incl. Drought Management Plans at different scale (river basin to EU level);
7. Assess the drought hazard at the pan-European scale in the 21st Century (future climate) and identify drought sensitive regions through combining improved knowledge on past events (through the developed European Drought Reference Database, EDR) and related long-term climate variability with projected future changes, incl. the analyses of the outcome from multi-models and multi-scenarios;
8. Assess potential vulnerabilities of the studied systems at the local, river basin and national scale in (potentially) water stressed areas (case studies), taking into account the assessment of the large-scale 21st Century drought hazard, socio-economic drivers of change and anticipated impacts of relevant EU policies, which is essential for development of drought management planning;
9. Assess changes in the modelled impacts for the 21st Century and resulting changes in risk of passing a vulnerability threshold at the pan-European scale in co-operation with the pan-European Drought Dialogue Forum to identify drought sensitive sectors, incl. their geographical location; 10. Share knowledge on drought with experts working on the 2nd cycle of WFD River Basin Management Plans and the general public through a web-based information and discussion platform hosted by the European Drought Centre (EDC)5, and to disseminate knowledge through a summer school and on the proposed 2nd Drought Conference as follow-up of the 1st Drought conference held in Brussels, February 2010.
The DROUGHT-R&SPI project consisted of five work packages. The objectives of these work packages (excl. management package) are given below.
DROUGHT AS A NATURAL HAZARD
The work package Drought as a Natural Hazard elaborated past and future drought and had the following objectives:
• Foster a better understanding of drought as a natural hazard (i.e. controlling processes and occurrence), and in particular to investigate the main climate drivers and characteristics (onset, severity, persistence and recovery) of the most extreme events in Europe;
• Identify drought sensitive regions by combining new knowledge on past events, natural long-term variability and projected future change;
• Advance system knowledge (persistence characteristics) and methodology for drought early warning (i.e. monitoring and forecasting), considering different time scales and geo-climatic regions;
• Develop and test a range of drought indicators suitable for early warning and impact studies at the pan-European scale;
• Facilitate the transfer of knowledge and information across scales through interaction between the pan-European scale and the case studies (local, river basin and national scale).
DROUGHT VULNERABILITY ASSESSMENT AND RISK REDUCTION AT DIFFERENT SCALES; DEVELOPMENT OF CASE STUDIES
This work package elaborated drought impacts, vulnerabilities, risk reduction and response options at the smaller scale (case studies) and had the following objectives:
• Provide an in-depth understanding of past drought hazards and of the range of pertinent environmental and socio-economic impacts in different contexts;
• Evaluate past responses to drought events, and identify best practice examples and pitfalls in alleviating drought impacts and reducing relevant risks;
• Assess future drought hazards and potential vulnerabilities of the studied systems, on the basis of climate projections, socio-economic drivers of change, environmental constraints and anticipated impacts of relevant EU policies;
• Identify and evaluate, in collaboration with the local Case Study Dialogue Platforms, potential responses for drought risk reduction, in line with the Hyogo Framework for Action;
• Assess the applicability of drought monitoring and early warning indicators at different contexts (local, river basin and national), and develop recommendations for their refinement to support decision-making.
• Examine the link between local-level research results to the pan-European assessments of drought hazards and vulnerabilities, to reveal whether drought events at local/regional scale, their impacts and root causes are relevant to broader contexts and vice -versa.
DROUGHT SENSITIVE AREAS IN EUROPE; IMPACTS, VULNERABILITIES & RISKS
This work package elaborated drought impacts, vulnerabilities and risks at the pan-European scale and had the following objectives:
• Improve the knowledge on links between drought indicators that can be monitored and predicted and a range of drought impacts (incl. crop loss, water supply shortage, low flow, energy production loss, forest fire, heat waves, user conflicts etc.) in different geo-climatic regions in Europe.
• Evaluate and compare vulnerability thresholds and explore the link between the magnitude of impacts and socio-economic, environmental and management factors influencing drought resilience.
• Develop in collaboration with the pan-European Dialogue Platform mapping procedures and map prototypes that allow the identification of sensitive areas in Europe in terms of drought risk from the physical exposure to drought, future changes, and the identified vulnerability thresholds.
• Provide recommendations for impact-related drought indicators and indices that will complement and expand existing Pan European hazard assessment, vulnerability monitoring, and early warning suitable to inform risk reduction measures in EU policies.
SCIENCE-POLICY INTERFACTING; DIALOGUE FORA AND DISSEMINATION AND TRAINING
The work package Science-Policy Interfacing: Dialogue Fora and dissemination and training had the following objectives:
• Foster live dialogue, an iteration and feedback process between researchers and water actors at different levels (river basin, national, pan-European) on drought research and policy agendas through establishing Dialogue Platform Fora;
• Support information-sharing and dissemination of knowledge among the DROUGHT-R&SPI consortium across scales (local, national, pan-European), and with non-European experts (USA, Canada, Asia), particularly with regard to identifying drought-sensitive regions and sectors, and efficient solutions for reducing drought vulnerabilities.
• Develop policy recommendations, stemming from the DROUGHT-R&SPI research and Science-Policy Interaction, to support the formulation of post-2015 WFD River Basin, incl. Drought Management Plans, and other international policies (e.g. UN/ISDR HFA);
• Effectively disseminate and communicate the DROUGHT-R&SPI results to wider audiences, incl. science, policy making and practitioner’s communities at different levels (river basin, national.
Project Results:
DROUGHT AS A LARGE-SCALE NATURAL HAZARD
The European Drought Reference database (EDR) has been developed during the project and was used to disseminate a detailed summary of for major European drought events. The database was made publicly available through the website of the European Drought Centre (www.geo.uio.no/edc). It contains a tool to display drought conditions on any day in the historical period 1958 - 2009, including the recommended meteorological indicators (Standardised Precipitation Index, SPI, Standardised Precipitation and Evaporation Index, SPEI) and hydrological drought indicators.
Studies using historical meteorological drought indicators confirmed the wetting trend in northern Europe and a drying trend in southern Europe. Understanding has improved about to what degree observed climatic trends can be attributed to changes in atmospheric circulation changes (i.e. trends in circulation type frequencies). Circulation changes from January to March were to found to have relatively high influence, contributing to wetting in the North and drying in the South. In general, circulation influence affects climate trends in north-western Europe stronger than the South-East. Several studies explored whether these trends are associated with natural forcing or anthropogenic climate change. Only the change pattern from simulations with anthropogenic forcing could resemble the observed changes.
Time series of historic drought beyond the instrumental record, i.e. 1500-1950 also have been obtained using proxy data (e.g. archives). The cases of England, France, Rhine and Syros series do not prove any increase in frequency of dry episodes since the 16th century, but outcome requires careful interpretation because drought reports are driven by negative impacts, which are associated with the vulnerability at that time. Likely vulnerability has decreased over time.
Historic drought information was also used to test performance of a suite of large-scale hydrological models (GHGs). The choice of model has a significant effect on areal statistics and persistence of historic drought and there is an overall tendency to overestimate the number of drought events and hence, underestimate drought duration. The studies learnt that ensembles of multi hydrological models are required for impact studies, incl. adaptation to climate change.
Impact of climate change on future meteorological drought, soil water and hydrological drought (severity, frequency, duration in different geo-climatic regions across Europe) was assessed through several modelling experiments using SREX or CMIP5 climate output. These experiments display large spreads over all time frames, but projected changes in the frequencies of future drought events show more robust signal-to-noise ratios towards more frequent meteorological drought and soil water drought before the end of the 21st century, particularly in the Mediterranean region. A high-resolution study focussing only at the European continent confirmed significant increases in meteorological drought severity for the Mediterranean region along with increases for areas along the Atlantic coast and in south-eastern Europe. Additional studies in several river basins across the globe, incl. Europe showed that not all combinations GHM-GCM were able to sufficiently capture historic drought in runoff. The constrained multi-model ensemble projects for the cold climates a regime shift and increase in low flows between the control period and future period (increased temperature and precipitation). Dry climates were found to become even drier in the future. One of the modelling studies demonstrated for the Case Study areas that the signal-to-noise ratios clearly are higher for climate change impact on drought characteristics (e.g. duration, severity) than on low flows. A model experiment showed that the drought identification approach has a large influence on projected hydrological droughts. Only in about 25% of the world both the drought duration and the deficit volume are expected to increase when applying an identification approach that considers gradual adaptation to the altered hydrological regime as compared to over 60% when the conventional approach (reference: 1971-2000 hydrological regime) was applied.
Improved understanding of hydroclimatological and catchment processes generating drought development is essential for credible multi-monthly to seasonal drought forecasting, which is a major scientific imperative of direct practical relevance. Different research lines were followed to improve drought forecasting. The first line is a large-scale approach based on antecedent atmospheric circulation or sea surface temperature. Processes were studied that link antecedent ocean-atmosphere variation to drought characteristics in Great Britain (streamflow) and Europe (SPI and SPEI). Patterns in geopotential height, wind, moisture vapour flux and precipitation prior to streamflow drought onset in Great Britain support the influence of the NAO, but also demonstrate that the atmospheric bridge linking North Atlantic SST to drought development is too complex to be described solely by indices of the NAO. Correlation of monthly time series of the percentage of European area in drought with geopotential height indicates that a weakening of the prevailing westerly circulation is associated with drought onset. Such conditions are linked to variation in the East Atlantic/Western Russia (EA/WR) and North Atlantic Oscillation (NAO) atmospheric circulation patterns. In the second research line, the potential to forecast streamflow drought from SPI and SPEI using different time accumulation periods was explored for a large set of catchments across Europe. The analysis illustrates that lagged cumulative precipitation deficits may be useful proxies to estimate hydrological drought, but that there is a large variability of relevant lag times and time accumulation periods from place to place due to catchment storage processes. A more comprehensive approach to understand drought-generating processes relevant for forecasting of hydrological drought involved the development of a drought typology, which describes how meteorological drought propagates into hydrological drought. A generic framework explains for 8 different drought types how precipitation and temperature anomalies control the conversion of a meteorological drought into a hydrological drought. In the last research line the potential predictability of soil moisture and streamflow drought was investigated using a hydrological model. The two drought types are predictable until lead times of approximately 7 days and 2–3 days, respectively, when using initial soil moisture information and climatological atmospheric forcing. Using also initial snow information and seasonal weather forecasts as forcing, the predictable lead time doubles in case of soil moisture (2 weeks) and triples for streamflow (1 week). Furthermore, the study demonstrates that soil moisture and streamflow forecast skills increase with increasing initial soil moisture anomalies, which implies an increased predictability for very dry soil moisture conditions.
VULNERABILITY AND RISK REDUCTION AT CASE STUDY SCALE
A detailed inventory and an ex-post evaluation of past practices to cope with drought were performed in the six selected Case Studies areas to support identification and selection of options for policy development and drought risk reduction. Stakeholders still perceive significant gaps in the establishment of preventive measures, structured drought planning, and a poor tradition of ex-post evaluation. In a few Case Study areas a detailed economic impact analysis was performed, which generated interesting outcome, in particular for the agriculture and energy sectors (producers and consumers are differently affected, i.e. winners and losers).
Policy recommendations for future drought risk reduction were formulated for the Case Study areas, on the basis of: (i) anticipated future drought impacts and vulnerability, (ii) assessment of alternative options, (iii) recommendations made by the members of the Case Study Dialogue Fora, and (iv) provisions in EU and international guidance documents. A common methodological framework was formulated for each research activity, however, in the implementation phase differentiation among Case Studies was introduced, to account for differences in spatial scales (regional, river basin, national), data availability, policy priorities in each of the areas, and stakeholder involvement. Each Case Study concluded on strategy-related actions for reducing drought risk and, as a further step, on recommendations/outcomes that were of value for the more generic Pan-EU drought assessments.
The analysis of future vulnerability to drought and of potential impacts as a result of future drought exposure and socio-economic development scenarios focused on three sectors: agriculture, urban water supply and energy production. Qualitative (surveys, stakeholder consultation through Case Study Drought Dialogue Fora) and quantitative (modelling) approaches were adopted. Agriculture remains the sector most affected by drought, whereas the risk of forest wildfires is expected to increase also in the future. Factors that may reduce future vulnerability are: (i) improvement of institutional framework on drought, (ii) overcoming policy gaps, (iii) establishment of monitoring & early warning systems, (iv) development of Drought Management Plans, (v) increasing water-use efficiency, and (vi) use of alternative water sources, and increasing user awareness. Factors that may increase future vulnerability are: (i) conflicts due to water scarcity, (ii) status of water sources (quality and quantity), (iii) increased variability of water availability due to climate change, and (iv) economic development patterns. Water scarcity was mentioned as an important exposure-related factor of vulnerability to drought. Any effort to cope with water scarcity will thus contribute to reducing vulnerability to drought. Stakeholder anticipation is that the socio-economic development pattern is the most influencing factor on future drought-related risks, even more than climate change.
Following the analysis of future impacts and vulnerability, options for drought risk reduction were analysed (quantitatively and qualitatively) using also input from stakeholders (next round Case Study Drought Dialogue Fora)). The identification of options was also based on recommendations from EU (e.g. Blueprint to safeguard water resources) and International (e.g. Hyogo Framework for Action) guidance documents. The study learnt that there a wide range of measures that can be implemented for drought risk mitigation and improved drought management. Their applicability is context-specific as susceptibility to drought and coping capacity are local in nature and defined by local/ regional characteristics. Currently already drought-prone and affected areas will more easily develop and implement options adapted to the local context, compared to areas that do not experience droughts frequently. The selection and implementation of drought risk mitigation options depends on the characteristics of the ongoing or upcoming drought. An 'auto-pilot' approach to drought risk mitigation is not possible: even in a particular area, drought intensity and impacts are not the same and particularities of measures may change from one event to another. All Case Studies concluded on proposals for enhancing drought management and summarised the guiding principles for their implementation, focusing on legislative requirements, technical capacity (tools, data requirements), financial issues, environmental considerations, and any other constraints for the strategy development or its improvement.
DROUGHT SENSITIVE AREAS IN EUROPE; IMPACTS, VULNERABILITIES & RISKS
The European Drought Impact Report Inventory (EDII) was developed during the project and extensively used for analysis of impacts to identify drought sensitive areas in Europe. EDII contains categorized data on drought impacts derived from a variety of text sources. The EDII was made publicly available through the website of the European Drought Centre (www.geo.uio.no/edc). Early 2015, the European Drought Impact Report Inventory (EDII) database contained about 5000 entries. The impact reports span the period 1900 to date, but most entries relate to impacts that occurred since the 1970s. Impact reports on agriculture, water supply, water quality and freshwater ecosystems were found to have dominated in most regions during most events. We found and illustrated that at the pan-European scale, the relative percentage of reported impacts by country suggests slightly higher relative importance of agriculture in the South and East, and of public water supplies in the South and West. The importance of energy and industry impacts, for example, but also other impact types, appears to be more country or region-specific as well as event-specific, e.g. dependent on whether an event is short and accompanied by a heat wave, such as the event of 2003 in Central Europe or characterized by a long multi-year deficit in surface water supplies such as the event of 2004-2008 in the Iberian Peninsula.
The EDII was combined with the natural hazard database (EDR) showing that the major large-scale drought events that occurred in Europe over the past few decades had SPI values below -2 for extended regions. However, a variety of drought impacts have also occurred for values below about -1 or -1.5. Regional studies found that impacts on water supplies, freshwater ecosystems, and energy and industry, which all rely on surface or groundwater resources, are more sensitive to long-duration droughts. These are best indicated by SPI-12 to -24. There are regional differences. Generally, the Mediterranean, which has adapted to lower precipitation and longer dry periods with reservoir storage and irrigated agriculture is less sensitive to short precipitation anomalies and is affected most by long accumulated precipitation deficits. The remainder of Europe responds to more seasonal droughts.
In addition to the impact analysis approach using EDII, quantitative drought impact data (crop yields, and area-burned by forest fire reported by member states) were used. In a pan-European study crop yield anomalies for five major crops were correlated with time series of SPI and SPEI. Correlations are low, but correlation patterns for wheat and barley show stronger links with SPI accumulated over three months in the flowering period. The potential to predict above normal wildfire activity was assessed using meteorological drought indicators (SPI). The study illustrates significant predictability of wildfire activity in southern Europe several months in advance.
Multi-model ensembles of GHM-GCM were used to investigate the impact of a changing climate on crop yields across Europe. For both the near future and the far future, a substantial increase in mean yield reduction was found for the mid-latitudes in Europe due to climate change (SRES A2 scenario). Because the southern regions already suffered from a high yield reduction under the current climate, changes were smaller. Northern regions in Europe are projected to have a lower yield reduction. However, the spread in individual model results was high indicating the high uncertainty in the changes towards the future.
Assessment of vulnerability to drought on a pan-European scale has been elaborated through evaluating and comparing vulnerability thresholds and explore the link between the magnitude of impacts and socio-economic, environmental and management factors influencing drought resilience through a detailed factor-based vulnerability assessment and mapping exercise focusing at the pan-European scale. For this purpose, a wealth of data was collected and evaluated to quantify vulnerability factors. The three vulnerability components (exposure, sensitivity, adaptive capacity) were broken down in 19 individual factors, such as physical drought characteristics, water stress, water body status (WFD), population, legal/institutional aspects, socio-cultural circumstances, water infrastructure, financial and economic conditions. After mapping individual factors these were aggregated to produce maps by component, and eventually the map of vulnerability to drought at pan-European level was produced. The results of this work lead to a number of key findings that show variation and relevance of factors vary for spatial scales around the NUTS-2 level, for situations and particular sectors. The maps resulting from this vulnerability assessment indicate trends of the level of vulnerability across Europe that can contribute to inform the design of future drought management policies and guidelines.
The vulnerability maps, which typically were based on a combination of relevant, more or less subjectively weighted vulnerability factors, served to complement the modelling and drought risk mapping. For the risk assessment the DROUGHT-R&SPI project took advantage of reported drought impacts from the EDII database and empirically estimated risk as the Likelihood of Impact Occurrence (LIO). The approach assumes that drought impacts are symptoms of vulnerability and can therefore serve as a proxy for part of the vulnerability. Three generations of risk model development and risk map construction were developed within the project. The EDII version (i.e. reflect number of impact reports) and the number of explaining variables (single variable versus multi variable) are the main differences. To map the risk of drought, LIOs were displayed for four impact categories “Agriculture & Livestock Farming”, “Public Water Supply”, ”Energy & Industry” and “Water Quality” for given hazard levels (index of SPEI), which correspond to particular drought return periods. The maps show interesting spatial variations of drought risk. For moderate drought, risk is highest in the Mediterranean, in particular for Agriculture and Livestock Farming, and in the densely populated areas of central Europe, in particular for Water Quality and Public Water Supply, a pattern that generally dominates the maps. However, for more severe drought hazard (SPEI values of -2.5 to -2), risk increases everywhere, but with the most differences spatially and for different considered categories. LIO of impacts on Energy and Industry only increase for the most severe drought hazard. For the most severe drought, risk is high almost everywhere.
The impact reports in the EDII evidenced that there are European regions where awareness of drought hazard has increased in recent years, and resilience building likely took place. The dynamics of drought vulnerability to be expected was analysed based on the reported drought impacts over time. Under the assumption that the composition of impacts represents a particular vulnerability profile at the time of a drought event, the study revealed that vulnerability is dynamic, i.e. changes from event to event. The distributions of impact types suggest at gradual changes following a certain trend over time but also following particular legislative changes that may have influenced this pattern. Overall, national water management plans and the European Framework directive may have decreased the vulnerability for impacts related to water resources management, but, may also have increased the awareness and recognition of environmental impacts (and with that seemingly the vulnerability) as a result of better monitoring and public interaction. Untangling these aspects for the past will be a key towards understanding the combination of future hazard predictions and vulnerability changes that lead to changes in drought risk in the future.
The studies also have revealed important issues complicating analyses. For instance, the spatial differences of impact reporting at administrative regional levels is often too coarse to associate it to other relevant units such as river basins, thus limiting the use of hydrological indicators. Hydrological indicators are also more difficult to obtain, as no EU wide observation based data product exist that is freely accessible to support a suitable analysis at the pan-European scale. Commonly, large-scale models do not account for important human alterations to river flow in the necessary detail.
Assessment of vulnerability to drought and associated risk at a pan-European level is a comprehensive task because it is different for individuals, sectors and nations, depending on several biophysical and socio-economic factors. Thus all indices, models, and consequently the maps are based on a number of compromises due to the aim for pan-European comparability.
DROUGHT INDICATORS
The complex nature of drought related phenomena also implies that there cannot be one single quantification of drought, satisfying all monitoring and early warning requirements. Hence, it is unlikely that one single drought monitoring and early warning system will satisfy the needs of all potential users. Therefore, a prerequisite for an efficient monitoring system is to identify the relevant stakeholders and to formulate well defined monitoring and early warning objectives, incl. indicators. Drought indicators are usually associated with so-called triggers, i.e. indicator values at which negative drought impacts are expected. Typical triggers can be classified into: (i) probabilistic triggers, which relate to the occurrence frequency of drought events; (ii) deterministic triggers, which refer to indicator values that should not be reached (e.g. environmental flow regulations); and (iii) empirical triggers which are based on empirically linking past drought impacts to drought indicators using statistical techniques.
The pan-European analyses and modelling of drought impacts and vulnerability served as a basis to provide recommendations for impact-related drought indicators and indices that will complement and expand existing Pan European hazard assessment, vulnerability monitoring, and early warning suitable to inform risk reduction measures in EU policies. Main observations are:
• Drought indicators showing the best link to drought impacts varied regionally at spatial scales smaller than the national or continental scale. Scale differences between impacts reported for administrative units (NUTS) and river basins used for hydrological monitoring complicate analyses.
• SPEI showed closer links to impacts than SPI, confirming other studies on the correlation between SPI (or SPEI) and different hydrological, agricultural, and ecological response variables. In the case of the quantitative crop yield anomalies SPEI links were not substantially stronger than the SPI.
• Best predictive indicators for particular drought impacts depend on sector and on sectorial management practices, e.g.:
- For rainfed agriculture, predominantly short to intermediate accumulation periods of SPI or SPEI (about 2-6 months) are better linked to impact occurrence.
- For irrigated agriculture, water supply, energy and industry, generally longer accumulation times of SPI and SPEI or combination of short and long accumulation periods are better linked to impacts, with optimal choices varying.
- For water-borne transport on the large rivers rather short accumulation periods of SPI or SPEI (about 1 2 month) likely will apply, although not specifically investigated in this project.
To fully account for these practices, more information and complex models and/or hydrological indices will be necessary (see below, detailed operational scale).
• Different drought impacts occur during different seasons, highlighting the need for a season-specific analysis, indicator or to otherwise account for this. Unfortunately, many impact reports only allow determining the year of occurrence, not the month or season.
• The link between drought indicators and impacts is often non-linear and affected by competing influences (e.g. fuel availability for wildfires; positive effects of drought-related weather during particular phases of crop growth or for particular crops) that can be revealed through analysis with well-chosen indicators.
• Little information is available for the onset, duration and end of impacts, making it difficult to analyse impacts over prolonged droughts as common in the Mediterranean and in large groundwater bodies in other parts of Europe (e.g. UK), and thus to select specific indicators that take into account the severity and duration of these long droughts.
However, at the detailed, operational scale (river basin scale, or in some cases national scale, like the Netherlands) use of tailored indicator systems is more efficient to communicate with stakeholders about the ongoing and forecasted hazard, the usually manifold impacts, the potential measures to be taken and the real time drought management, to better link with impacts and also for real time management of events. The physical meaning is hard to communicate to stakeholders. For these purposes, the use of a Decision Support System for drought management is almost indispensable to accomplish appropriate planning for drought and also an effective real time decision making. The involvement of stakeholders is essential, not only to support decision making during the drought (incl. the onset and recovery phases), but also for the planning of structural measures to reduce future drought vulnerability, hence to lower the risk (pro-active).
SCIENCE-POLICY INTERFACING; DROUGHT DIALOGUE FORA
Drought Dialogue Fora (DDFs) were further been developed. In most Case Study areas several rounds of fora (CS DDFs) were organised. These fora were excellent platforms through facilitated discussions to collect information on factors that governed vulnerability and risk of past drought or will determine these in the future. The dialogue was also fundamental in testing the developed methods (see Vulnerability and risk reduction at Case Study scale, above). Stakeholders in the CS DDFs prefer regular contact and cooperation with scientists in an open discussion atmosphere, rather than simply participating in annual events/workshops with single-way presentations and short discussion time in usually considerably crowded places The outcome of the CS DDFs very much contributed to defining methodologies to assess drought sensitive areas in Europe: through pan-European analyses of impacts, vulnerabilities and risks (see above). The methodologies for large-scale analyses were further shaped and tested in three pan-European Drought Dialogue Fora (pan-EU DDF). On the other hand the pan-EU DDFs helped to interpret options proposed in EC policy documents (e.g. Blueprint) to become meaningful at the Case Study scale.
While the 1st pan-EU DDF (Nicosia, Cyprus) served to explore requirements of policy makers and coordinators of other large national and international projects in terms of methodologies for drought vulnerability and risk assessment (see 1st Periodic Report), the 2nd pan-EU DDF (Brussels) focussed on discussion and feedback to four issues related to drought vulnerability assessment and potential risk mapping at the pan-European scale. The four issues were: (i) the scale and use of risk maps, (ii) specific impacts (and hence sectors) of drought that are relevant in different geo-climatic regions in Europe, (iii) factors that affect vulnerability to drought, and (iv) potential models to assess this vulnerability. Options for the size of spatial units were indicated, as well as factors that either increase vulnerability (e.g. high population density in urban areas, high dependence of energy production on freshwater) or decrease vulnerability (prominent stakeholders involvement, well-developed regulatory framework related to drought, high degree of social awareness about drought risk). On the 3rd pan-EU DDF (Brussels) preliminary project’ outcome was presented and discussed. This involved: (i) European Drought Impact Database (EDII); (ii) first pan-EU maps of drought vulnerability factors, (iii) drought: monitoring, forecasting and indicators; and (iv) how to upscale context-specific CS experiences to the pan-European scale. The 4th pan-EU DDF had a more scientific nature and was linked to the final DROUGHT-R&SPI Conference in Valencia. Final vulnerability and risk maps were presented there. Associated developments were presented and discussed, such as the guidelines for ecological flows (e-flows) and implementation of pan-EU knowledge and experiences in the Member States (Portugal, UK and Jucar).
Potential Impact:
POTENTIAL IMPACT
The adoption of a trans-disciplinary approach and the incorporation of stakeholders and policy makers across scales right from the beginning led to DROUGHT-R&SPI having already impact on Drought Management Planning, i.e. helping to formulate reduce drought vulnerability and policy options in the Case Study areas. In particular, in the Mediterranean cases (Portugal, Jucar River basin, Po River basin and the Island of Syros) were in close co-cooperation happened with stakeholders and policy makers to formulate options. In the less-water stresses cases (Switzerland, the Netherlands), the DROUGHT-R&SPI project was not steering, but contributed to planning to make the countries less vulnerable to drought through associations with ongoing national drought initiatives (Switzerland: DROUGHT.CH project; and the Netherlands: Delta programme, securing fresh water supply programme).
The DROUGHT-R&SPI project has provided scientifically sound and new knowledge that is ready to be used to reduce drought vulnerability and associated risk in Europe through:
• Development of a methodology to assess and map pan-European drought vulnerability at the level of the NUTS2 scale, which is based on 19 vulnerability factors that address exposure, sensitivity and adaptive capacity as key vulnerability components
• Development of a methodology to assess and map pan-European drought risk taking advantage of a wealth of collated historic impact reports (European Drought Impact Report Inventory, EDII). Risk is estimated as Likelihood of Impact Occurrence (LIO). LIO maps were displayed for four impact categories “Agriculture & Livestock Farming”, “Public Water Supply”, ”Energy & Industry” and “Water Quality” for given hazard levels (expressed as Standardized Precipitation Evaporation Index), which correspond to particular drought return periods.
The maps resulting from the vulnerability assessment indicate the spatial distribution of the level of vulnerability across Europe, incl. causing factors. The generated vulnerability maps can be combined with the impact-based risk maps to identify drought sensitive areas and impacted sectors for different return periods. The outcome of this analysis can contribute to fine tune the design of future drought management policies and guidelines.
The pan-European vulnerability and associated risk maps have a generic nature because of their pan-European inter-comparability. These large-scale maps can be complemented with outcome of: (i) the analysis of the future vulnerability to drought and of potential impacts at the Case Study scale as a result of future drought exposure and socio-economic development scenarios, and (ii) policy recommendations for future drought risk reduction.
In addition to the above-mentioned vulnerability and risk maps, the project’ results can assist further development of drought management plans, incl. adequate drought monitoring and early warning (DMEW) as part of River Basin Management Plans, EU Water Framework Directive (WFD). Drought indicators are an important DMEW component. Guidance is provided on drought indicators, which can have a different nature (probabilistic, deterministic, empirical). Awareness indicators are recommended for large-scale studies and for informing the general public. These indicators may have a composite nature combining different drought types, provided that these are well communicated. More specific, impact-based indicators are required when decisions need to be taken. The studies using the EDII prove that these indicators should be specific to the vulnerability of sector)s) and potential impacts of the region or river basin. At the pan-European scale, DMEW mainly benefits from transboundary communication in its current state. However, it could become an important tool for future policy choices (e.g. drought declaration, aid and compensation) and risk management decisions at the national and multi-national scale, if it was linked more specifically to drought impacts Comprehensive, real-time drought management at the river basin scale often uses a framework of monitoring-modeling-Decision Support Systems driven by operational drought indicators. Monitored and forecasted spatial-distributed states of hydrometeorological variables and fluxes are translated in indicators tailored to the impacted sectors. Indicators are also simulated to explore effects of potential short-term measures that reduce impacts. Knowledge obtained on drought impacts is already used to explore if information provided by the European Drought Observatory can be enhanced.
The project will provide support to adaption to climate change, in particular through:
• The European Drought Reference database (EDR) that can help to test the next generation of global hydrological models (e.g. better subsurface storage representation, higher spatial resolution) driven by recent climate projections. DROUGHT-R&SPI has provided a number of methods on how drought characteristics identified with multi-model ensembles can be tested against observations and documented events
• Projections of different drought types (meteorological, soil water, runoff, river flow) using different future climate forcings. Different uncertainty assessment methods (e.g. multi-model, signal-to-noise) have been developed
• Different drought identification approaches (e.g. standardized variables, threshold approaches). Moreover, the project has illustrated the substantial influence of the identification method on hydrological drought whether it considers adaptation to a gradually changing hydrological regime or not. The latter usually is applied, with e.g. the period 1971-2000) as reference
• Quantitative Impacts on future crop yield, incl. ways to express uncertainty
• The EDR that can assist further development of adaptation measures by using exposure data form historic and projected droughts. The EDII can help with reported manifold impacts as response to drought and the links between the impacts and meteorological indicators (Standardized Precipitation Index, SPI; Standardized Precipitation Evaporation Index, SPEI.
The DROUGHT-R&SPI communicated with several policy makers in Brussels, i.e. DG Environment, DG Climate and DG Enterprise, as well as with Commission Services (JRC, EEA) dealing with hydrological extremes. This was done through the pan-European Drought Dialogue Fora, targeted meetings, the WFD-CIS Expert Group on Water Scarcity and Drought, the WFD-CIS Working Group on Groundwater and a recent EEA meeting on climate and weather-induced extremes.
The European Drought Impact Report Inventory (EDII) is a powerful database to be used in the European Research Area (ERA). It can be extremely suitable, for example in Water and Climate studies, disaster risk studies that aim to further develop or to introduce new measures and technologies to enhance effective response capacity to extreme weather and climate events affecting the security of people and assets.
The project also impacted the implementation of the WMO/Global Water Partnership Integrated Drought Management Programme (IDMP) in ten Central and Eastern European countries (CEE). The 7-step approach to develop a national drought management policy and plan was reviewed and concrete suggestions were given to improve. The IDMP-CEE approach will be the basis for the drought chapter in the next cycle of River Basin Management Plans that the EU countries compile as response to the EU Water Framework Directive (WFD). The non-EU countries in CEE are encouraged to adopt the IDMP approach in their national and regional water management and policy.
The DROUGHT-R&SPI project can also assist implementation of several priorities of the Hyogo Framework for Action, in particular to: (i) identify, assess and monitor disaster risks and enhance early warning, (ii) reduce the underlying risk factors, (iii) strengthen disaster preparedness for effective response at all levels, and (iv) use knowledge, innovation and education to build a culture of safety and resilience at all levels. Further development of these priorities is essential as concluded at the recent World Water Forum in Korea (April 2015), where it was discussed that globally, water-related disasters account for 90% of all natural disasters, and impacts from water-related disasters can in economic terms be up to 15% of annual GDP for certain countries.
The project already has significant outreach during the project life time due to Drought Dialogues in the Case Study areas and four pan-European Drought Dialogue Fora (pan-EU DDFs). In some of the pan-EU DDFs there was also interaction with several drought-related EU projects, initiatives of International Organisations (e.g. UNESCO-IHP, WMO/GWP) and large national projects (e.g. SCARCE). The project further trained about 10 PhD and postdocs. The Drought Summer School in Syros (Greece) increased knowledge of 25 PhD students, postdocs and junior scientists and allowed them to build a network. Preliminary Project’ outcome was already presented and discussed at tens of conferences and workshops. The DROUGHT-R&SPI Conference in Valencia (March 2015) allowed the consortium to present the final outcome and to discuss it with over 150 people working on drought. Project’s results will be available beyond its lifetime through the Science-Policy Briefs, 35 two-page flyers, about 50 peer-reviewed papers, and 35 DROUGHT-R&SPI Technical Reports. The website will be accessible until at least 2020. The European Drought Reference database (EDR) and the European Drought Impact Report Inventory (EDII) and the R-software package to calculate Standardized Climate Indices (SCI) will also help to remain the DROUGHT-R&SPI project still visible for a number of years.
DISSEMINATION ACTIVITIES AND EXPLOITATION RESULTS
In the 2nd Science-Policy Brief (SPB), the DROUGHT-R&SPI consortium decided for two perspectives, both starting from the urgent need to reduce drought impacts, through increased knowledge, drought management plans and an improved science-policy interfacing that will lead to better preparedness and reduced vulnerability to future drought and the risks they pose for Europe. One perspective is that of the “Tools and Instruments” that contribute to Improved Drought Management and Preparedness. The second perspective resulted from realising that “Science-Policy Interfacing” alone is not enough in the chain to implement procedures and measures in relation to drought. There is a need to move beyond SPI towards SPI-Implementation. After ‘science’ and ‘policy’ have ‘interfaced’ to enable formulation and implementation of policies, guidelines, procedures and the like, the Member States’ water managers have to ensure the correct ’Implementation’ of the measures, in cooperation with stakeholders.
The drought-related messages to different audiences started from the view that drought is a multi-stakeholder and multi-dimensional phenomenon for which uniform measures are difficult to provide and implement, and therefore management plans need to include local demands and make use of tailored advice to deal adequately with the uncertainties involved in dealing with drought. An on-line questionnaire survey was designed to collect views from different people dealing with drought. Eventually messages were developed for three major stakeholder groups: i.e. Public Authorities, Businesses and Citizens. The desire for water saving technologies, techniques and drought management strategies and plans, in other words pro-active measures, were topics to be addressed in all messages. The same applies to communication on drought, but the means are different for the categories. Businesses and Public Authorities appear to have similar preferences in way they would like to obtain information (e.g. meetings, information sessions/conferences and educational programmes), whilst Citizens have a preference towards the media and smart phone applications.
The “International Conference on Drought Research and Science-Policy Interfacing” was the final event of DROUGHT-R&SPI project. The conference was held from 10-13 March 2015 in Valencia, Spain. With a total number of about 150 participants, it included 61 oral presentations and 36 posters. A total of 73 of these contributions were published in a book (CRC/ Balkema). The conference incorporated the 4th pan EU DDF. Drought related research were discussed, incl. progress on response policies to foster development of Drought Policies and Plans to reduce risk and vulnerability, and to enhance preparedness and resiliency. The conference addressed drought identification and characterization, drought assessment (impacts) and drought policies to develop drought management plans at different scales (river basin, national and international). The conference was an outstanding platform to convey the main results from the project and it served as a forum for discussion among scientists researching drought, stakeholders, water managers, experts and representatives of authorities on experiences, about the needs and challenges for drought management.
The 1-week Summer School on ‘Drought Hazard and Management: Challenges in a Changing World’ was held in June 2014 in Hermoupolis Syros, Greece. It brought together 25 PhD candidates and young researchers (14 women; 11 men; 6 participants had a non-European nationality and came from China, India, Algeria, Brazil, Iran, German/Uruguayan). The School included lectures, workshops, a poster session, where participants presented their on-going research activities, a field trip and a closing session, where participants presented the outcomes from the Workshops. In addition to obtaining knowledge about drought and skills to handle drought, the school offered a perfect platform for networking, which should not be underestimated. Previous school with a similar setup have proven to be the basis for long-lasting working contacts.
In the project about 10 PhD students and postdocs were further educated in the various aspects related to drought. About 35 DROUGHT-R&SPI Technical Reports were written, which explain in detail the methodologies that have developed and the results that were obtained. Around 35 two page flyers have been made available, which provide specific project’ outcome in two pages. Two Science Policy Briefs were written. The reports, flyers and Science Policy Briefs are freely available on the internet. About 50 peer-reviewed papers are published in scientific journals dealing with natural hazards, water management, hydrology, climate, environment. Almost all are open access. In addition to the Drought Dialogue Fora, project results have been presented and shared with colleagues on almost 100 conferences, workshops, assemblies, symposia, seminars. The content of many of these presentations can be found in abstracts or conference proceedings (see 2nd Periodic Report, Section 4.5 and 1st Periodic Report, Section 4.2).
The project website (www.eu-drought.org/) provides in addition to information about the DROUGHT-R&SPI project, links to the DROUGHT-R&SPI Technical reports, flyers, Science Policy Briefs and peer-reviewed papers. The website will remain accessible at least until 2020. The DROUGHT-R&SPI website is closely linked to the website of the European Drought Centre (EDC, www.geo.uio.no/edc). The EDC website houses both the European Drought Reference (EDR) database, which provides detailed meteorological and hydrological drought statistics, and the European Drought Impact Report Inventory (EDII), which stores user-provided information about drought impacts. The EDC platform has been used in the project for an interactive web based discussion forum to further develop the EDII. This was particularly linked to the Drought Summer School in Syros.
A software package in R (Repository CRAN) was developed. The R package “SCI” calculates Standardized Climate Indices, such as SPI, SRI or SPEI. SCI is a transformation of (smoothed) climate (or environmental) time series that removes seasonality and forces the data to take values of the standard normal distribution. SCI was originally developed for precipitation. In that case it is known as the Standardized Precipitation Index (SPI).
List of Websites:
The public project website can be found at: http://www.eu-drought.org/
Contact details:
dr.ir. Henny A.J. van Lanen
Coordinator EU-FP7 DROUGHT-R&SPI
Wageningen University
P.O. Box 47
6700 AA Wageningen
The Netherlands
Visiting address:
LUMEN (building 100), room 0.C29
Droevendaalsesteeg 3a
6708 PB Wageningen
T: +31 317 48 2418 (office)
T: +31 317 48 2778 (secretary)
M: +31 6 835 96 767
F: +31 317 41 9000
European Drought Centre: www.geo.uio.no/edc
WATCH (WATer and global Change): http://www.eu-watch.org/
DROUGHT-R&SPI SPI (Fostering European Drought Research and Science-Policy Interfacing): http://www.eu-drought.org/
Drought Dialogue Fora at the pan-European (pan-EU DDF) and the Case Study scale (CS DDF) were held from the start and served as excellent platforms through facilitated discussions among stakeholders, policy-makers and the consortium: (i) to collect information on factors that govern vulnerability and risk to drought, (ii) to interpret options proposed in EC policy documents, (iii) to test and further shape developed methods (e.g. vulnerability, risk reduction) to identify drought-sensitive in Europe, and (iv) to communicate on past and future natural hazard.
A European Drought Reference database (EDR) was established that displays historical drought as a natural hazard (1958–2009). Trend studies using recent instrumental data confirmed the wetting trend in northern Europe and a drying trend in southern Europe. Time series of historic drought beyond the instrumental record (1500-1950) derived from proxy data (e.g. archives) reveal that frequency of dry episodes has not increased in several European regions. Understanding has improved about to what degree observed climatic trends can be attributed to changes in atmospheric circulation changes (i.e. trends in circulation type frequencies, anthropogenic forcing). Historic drought information (e.g. EDR) was used to test performance of large-scale hydrological models that reveal that ensembles of multi hydrological models are required for impact studies (climate change). Several modelling experiments using different IPCC climate output assessed climate change impact on drought. Large spreads over all time frames are observed, but projected droughts are more robust by the end of 21st century in multiple European regions. Different research lines were followed to improve drought forecasting, essential for early warning. Links of antecedent ocean-atmosphere variation to large-scale drought characteristics show that atmospheric bridge connecting North Atlantic SSTs to drought development is too complex to be described solely by NAO. Improved understanding of drought generating processes led to a comprehensive drought typology that is essential for credible multi-monthly to seasonal drought forecasting. Catchment storage is essential to explain the connection between lagged cumulative precipitation deficits and streamflow drought.
Policy recommendations for future drought risk reduction were identified for the six selected Case Study areas (CS) based upon: (i) collated past drought hazards and responses, (ii) anticipated future drought impacts and vulnerability, (iii) assessment of alternative options, (iii) recommendations of CS DDFs, and (iv) EU and international guidance documents. Factors that either reduce or increase drought vulnerability and risk are given, which due to context-specific conditions differ per CS. The CS studies conclude that an 'auto-pilot' approach to drought risk mitigation is impossible: even in a particular area, because drought intensity and impacts are not the same and particularities of potential measures may change from one event to another.
The European Drought Impact Report Inventory (EDII, publicly available through the EDC website) was used for impact analysis to identify drought sensitive areas in Europe. EDII contains categorized data on drought impacts derived from a variety of text sources (about 5000 reports). Impact reports on agriculture, water supply, water quality and freshwater ecosystems dominate, although region-specific as well as event-specific differences and temporal changes occur. Some sectors (e.g. rainfed agriculture) respond to short-term drought, whereas other sectors (water supply, freshwater ecosystems, energy and industry) relying on groundwater and river flow are more sensitive to long-duration droughts. Some regions (e.g. the Mediterranean) are more adapted through reservoir storage and irrigated agriculture. Thresholds in drought indicators were found by combining EDII with the natural hazard database (EDR). Based on these threshold and the assumption that drought impacts can serve as a proxy for vulnerability, the Likelihood of Impact Occurrence (LIO, varies between 0 and 1) was introduced to generate pan-European drought risk maps for different sectors and severity of the natural hazard. Spatial variations of drought risk occur. For moderate drought (hazard), risk is highest in the Mediterranean, in particular for agriculture, and in the densely populated areas of central Europe, for public water supply. In addition to the LIO approach, pan-European vulnerability to drought has been elaborated through a detailed factor-based vulnerability assessment. The three vulnerability components (exposure, sensitivity, adaptive capacity) were broken down in 19 individual factors. Maps per factor were stepwise aggregated and eventually to a pan-European map showing drought vulnerability (NUTS-2 scale. Assessment of drought vulnerability and associated risk at pan-European level involves compromises because these are different for individuals, sectors and regions, depending on several biophysical and socio-economic factors.
Because of the complex nature of the drought phenomenon, one single quantification of drought (indicator) does not exist that satisfies all monitoring and early warning requirements. Best predictive indicators for particular drought impacts depend on sector and on sectorial management practices and need to be defined with stakeholders. At the detailed, operational scale use of tailored indicator systems in a combined monitoring-modeling-decision support system framework are a prerequisite to communicate with stakeholders about ongoing and forecasted hazards, usually manifold impacts, potential measures to be taken, real time drought management and pro-active drought planning.
Project Context and Objectives:
The DROUGHT-R&SPI project aimed to reduce future Europe’s vulnerability and risk to drought by innovative in-depth studies that combine drought investigations in six case study areas in water-stressed regions (river basin and national scale) with drought analyses at the pan-European scale. Knowledge transfer across these scales is paramount because vulnerability is context-specific (e.g. physical, environmental, socio-economic, cultural, legal, institutional), which requires analyses on detailed scales, whereas international policies and drought-generating climate drivers and land surface processes are operating on large scales. The project planned to adopt Science-Policy Interfacing at the various scales, by establishing Case Study Dialogue Fora and a pan-Europe Dialogue Forum, which should ensure that the research was well integrated into the policy-making from the start of the project onwards. The study intended to foster a better understanding of past droughts (e.g. underlying processes, occurrences, environmental and socio-economic impacts, past responses), which then is anticipated to contribute to the assessment of drought hazards and potential vulnerabilities in the 21th C. Methodologies for early drought warning at the pan-European scale were proposed to be developed, which aim to improve forecasting of a suite of interlinked physical and drought impact indicators. This intends to increase drought preparedness, and to identify and implement appropriate Disaster Risk Reduction measures. The project meant to result through combined drought studies at different scales to the identification of drought-sensitive regions and sectors across Europe, which likely will lead to a more thorough implementation of the EU Water Framework Directive, particularly by further developing of methodologies for Drought Management Plans at different scales (incl. EU level). The work was anticipated to be linked with the European Drought Centre (EDC) ensuring that the outcome has been consolidated beyond the project’ lifetime.
The main objectives of Drought-R&SPI were:
1. Foster a better understanding of past droughts (underlying processes, occurrences, incl. frequencies, severities and scales) and in particular to investigate driving factors and characteristics of the most extreme historic events at the pan-European scale, which will be validated against observed natural hazards at the small scale (case studies);
2. Provide an in-depth understanding of the range of pertinent environmental and socio-economic impacts triggered by past drought hazards in different contexts, taking into account economic, political, and social factors at a small scale in water-stressed regions (case studies) and considering the large-scale drivers (transfer of knowledge and information across scales);
3. Evaluate past responses to drought events at the local, river basin and national scale (case studies), and identify best practice examples and lessons learnt in alleviating drought impacts and reducing associated risks from different local environments, which are crucial for the development of methodologies for drought management planning;
4. Develop and test an innovative suite of drought indicators that integrate physical, impact and vulnerability indices, which will address different spatio-temporal scales, interaction across scales and water-related sectors, and that will build upon: (i) the comprehensive knowledge obtained in different contexts at the detailed scale (case studies), and (ii) exploration of the link between impact records (socio-economic, environmental), management factors and vulnerability thresholds across Europe. Development and testing will be linked and shared with research done through the European Drought Observatory (EDO) at the EC Joint Research Centre (JRC) and Waterbase, WSDIS and WISE-RTD (European Environmental Agency, EEA);
5. Develop an innovative methodology for early warning (monitoring and forecasting) of drought at the pan-European scale using the developed European Drought Impact Inventory (EDII database) and suite of indicators, which will help to: (i) to increase drought preparedness, and (ii) to identify and implement appropriate Disaster Risk Reduction (DRR) measures. Drought-R&SPI will share the obtained knowledge and information with the JRC, which are designing and developing an operational version;
6. Establish Drought Dialogue Fora at different scales, i.e. Science-Policy Interfacing, (i) to identify and evaluate, in close collaboration with the Case Study Drought Dialogue Fora, potential responses for drought risk mitigation, taking into account the drought hazard, its impacts, vulnerabilities and desired development goals, strategies and relevant trade-offs, and (ii) to recognize and assess, in close dialogue with the pan-European Drought Dialogue Forum through an iteration process with feedback loops, how EU and other international policies affect potential responses at the small scale (case studies). Finally, how these together with large-scale impact maps and vulnerabilities will support design, development or implementation of drought-related international policies and planning, incl. Drought Management Plans at different scale (river basin to EU level);
7. Assess the drought hazard at the pan-European scale in the 21st Century (future climate) and identify drought sensitive regions through combining improved knowledge on past events (through the developed European Drought Reference Database, EDR) and related long-term climate variability with projected future changes, incl. the analyses of the outcome from multi-models and multi-scenarios;
8. Assess potential vulnerabilities of the studied systems at the local, river basin and national scale in (potentially) water stressed areas (case studies), taking into account the assessment of the large-scale 21st Century drought hazard, socio-economic drivers of change and anticipated impacts of relevant EU policies, which is essential for development of drought management planning;
9. Assess changes in the modelled impacts for the 21st Century and resulting changes in risk of passing a vulnerability threshold at the pan-European scale in co-operation with the pan-European Drought Dialogue Forum to identify drought sensitive sectors, incl. their geographical location; 10. Share knowledge on drought with experts working on the 2nd cycle of WFD River Basin Management Plans and the general public through a web-based information and discussion platform hosted by the European Drought Centre (EDC)5, and to disseminate knowledge through a summer school and on the proposed 2nd Drought Conference as follow-up of the 1st Drought conference held in Brussels, February 2010.
The DROUGHT-R&SPI project consisted of five work packages. The objectives of these work packages (excl. management package) are given below.
DROUGHT AS A NATURAL HAZARD
The work package Drought as a Natural Hazard elaborated past and future drought and had the following objectives:
• Foster a better understanding of drought as a natural hazard (i.e. controlling processes and occurrence), and in particular to investigate the main climate drivers and characteristics (onset, severity, persistence and recovery) of the most extreme events in Europe;
• Identify drought sensitive regions by combining new knowledge on past events, natural long-term variability and projected future change;
• Advance system knowledge (persistence characteristics) and methodology for drought early warning (i.e. monitoring and forecasting), considering different time scales and geo-climatic regions;
• Develop and test a range of drought indicators suitable for early warning and impact studies at the pan-European scale;
• Facilitate the transfer of knowledge and information across scales through interaction between the pan-European scale and the case studies (local, river basin and national scale).
DROUGHT VULNERABILITY ASSESSMENT AND RISK REDUCTION AT DIFFERENT SCALES; DEVELOPMENT OF CASE STUDIES
This work package elaborated drought impacts, vulnerabilities, risk reduction and response options at the smaller scale (case studies) and had the following objectives:
• Provide an in-depth understanding of past drought hazards and of the range of pertinent environmental and socio-economic impacts in different contexts;
• Evaluate past responses to drought events, and identify best practice examples and pitfalls in alleviating drought impacts and reducing relevant risks;
• Assess future drought hazards and potential vulnerabilities of the studied systems, on the basis of climate projections, socio-economic drivers of change, environmental constraints and anticipated impacts of relevant EU policies;
• Identify and evaluate, in collaboration with the local Case Study Dialogue Platforms, potential responses for drought risk reduction, in line with the Hyogo Framework for Action;
• Assess the applicability of drought monitoring and early warning indicators at different contexts (local, river basin and national), and develop recommendations for their refinement to support decision-making.
• Examine the link between local-level research results to the pan-European assessments of drought hazards and vulnerabilities, to reveal whether drought events at local/regional scale, their impacts and root causes are relevant to broader contexts and vice -versa.
DROUGHT SENSITIVE AREAS IN EUROPE; IMPACTS, VULNERABILITIES & RISKS
This work package elaborated drought impacts, vulnerabilities and risks at the pan-European scale and had the following objectives:
• Improve the knowledge on links between drought indicators that can be monitored and predicted and a range of drought impacts (incl. crop loss, water supply shortage, low flow, energy production loss, forest fire, heat waves, user conflicts etc.) in different geo-climatic regions in Europe.
• Evaluate and compare vulnerability thresholds and explore the link between the magnitude of impacts and socio-economic, environmental and management factors influencing drought resilience.
• Develop in collaboration with the pan-European Dialogue Platform mapping procedures and map prototypes that allow the identification of sensitive areas in Europe in terms of drought risk from the physical exposure to drought, future changes, and the identified vulnerability thresholds.
• Provide recommendations for impact-related drought indicators and indices that will complement and expand existing Pan European hazard assessment, vulnerability monitoring, and early warning suitable to inform risk reduction measures in EU policies.
SCIENCE-POLICY INTERFACTING; DIALOGUE FORA AND DISSEMINATION AND TRAINING
The work package Science-Policy Interfacing: Dialogue Fora and dissemination and training had the following objectives:
• Foster live dialogue, an iteration and feedback process between researchers and water actors at different levels (river basin, national, pan-European) on drought research and policy agendas through establishing Dialogue Platform Fora;
• Support information-sharing and dissemination of knowledge among the DROUGHT-R&SPI consortium across scales (local, national, pan-European), and with non-European experts (USA, Canada, Asia), particularly with regard to identifying drought-sensitive regions and sectors, and efficient solutions for reducing drought vulnerabilities.
• Develop policy recommendations, stemming from the DROUGHT-R&SPI research and Science-Policy Interaction, to support the formulation of post-2015 WFD River Basin, incl. Drought Management Plans, and other international policies (e.g. UN/ISDR HFA);
• Effectively disseminate and communicate the DROUGHT-R&SPI results to wider audiences, incl. science, policy making and practitioner’s communities at different levels (river basin, national.
Project Results:
DROUGHT AS A LARGE-SCALE NATURAL HAZARD
The European Drought Reference database (EDR) has been developed during the project and was used to disseminate a detailed summary of for major European drought events. The database was made publicly available through the website of the European Drought Centre (www.geo.uio.no/edc). It contains a tool to display drought conditions on any day in the historical period 1958 - 2009, including the recommended meteorological indicators (Standardised Precipitation Index, SPI, Standardised Precipitation and Evaporation Index, SPEI) and hydrological drought indicators.
Studies using historical meteorological drought indicators confirmed the wetting trend in northern Europe and a drying trend in southern Europe. Understanding has improved about to what degree observed climatic trends can be attributed to changes in atmospheric circulation changes (i.e. trends in circulation type frequencies). Circulation changes from January to March were to found to have relatively high influence, contributing to wetting in the North and drying in the South. In general, circulation influence affects climate trends in north-western Europe stronger than the South-East. Several studies explored whether these trends are associated with natural forcing or anthropogenic climate change. Only the change pattern from simulations with anthropogenic forcing could resemble the observed changes.
Time series of historic drought beyond the instrumental record, i.e. 1500-1950 also have been obtained using proxy data (e.g. archives). The cases of England, France, Rhine and Syros series do not prove any increase in frequency of dry episodes since the 16th century, but outcome requires careful interpretation because drought reports are driven by negative impacts, which are associated with the vulnerability at that time. Likely vulnerability has decreased over time.
Historic drought information was also used to test performance of a suite of large-scale hydrological models (GHGs). The choice of model has a significant effect on areal statistics and persistence of historic drought and there is an overall tendency to overestimate the number of drought events and hence, underestimate drought duration. The studies learnt that ensembles of multi hydrological models are required for impact studies, incl. adaptation to climate change.
Impact of climate change on future meteorological drought, soil water and hydrological drought (severity, frequency, duration in different geo-climatic regions across Europe) was assessed through several modelling experiments using SREX or CMIP5 climate output. These experiments display large spreads over all time frames, but projected changes in the frequencies of future drought events show more robust signal-to-noise ratios towards more frequent meteorological drought and soil water drought before the end of the 21st century, particularly in the Mediterranean region. A high-resolution study focussing only at the European continent confirmed significant increases in meteorological drought severity for the Mediterranean region along with increases for areas along the Atlantic coast and in south-eastern Europe. Additional studies in several river basins across the globe, incl. Europe showed that not all combinations GHM-GCM were able to sufficiently capture historic drought in runoff. The constrained multi-model ensemble projects for the cold climates a regime shift and increase in low flows between the control period and future period (increased temperature and precipitation). Dry climates were found to become even drier in the future. One of the modelling studies demonstrated for the Case Study areas that the signal-to-noise ratios clearly are higher for climate change impact on drought characteristics (e.g. duration, severity) than on low flows. A model experiment showed that the drought identification approach has a large influence on projected hydrological droughts. Only in about 25% of the world both the drought duration and the deficit volume are expected to increase when applying an identification approach that considers gradual adaptation to the altered hydrological regime as compared to over 60% when the conventional approach (reference: 1971-2000 hydrological regime) was applied.
Improved understanding of hydroclimatological and catchment processes generating drought development is essential for credible multi-monthly to seasonal drought forecasting, which is a major scientific imperative of direct practical relevance. Different research lines were followed to improve drought forecasting. The first line is a large-scale approach based on antecedent atmospheric circulation or sea surface temperature. Processes were studied that link antecedent ocean-atmosphere variation to drought characteristics in Great Britain (streamflow) and Europe (SPI and SPEI). Patterns in geopotential height, wind, moisture vapour flux and precipitation prior to streamflow drought onset in Great Britain support the influence of the NAO, but also demonstrate that the atmospheric bridge linking North Atlantic SST to drought development is too complex to be described solely by indices of the NAO. Correlation of monthly time series of the percentage of European area in drought with geopotential height indicates that a weakening of the prevailing westerly circulation is associated with drought onset. Such conditions are linked to variation in the East Atlantic/Western Russia (EA/WR) and North Atlantic Oscillation (NAO) atmospheric circulation patterns. In the second research line, the potential to forecast streamflow drought from SPI and SPEI using different time accumulation periods was explored for a large set of catchments across Europe. The analysis illustrates that lagged cumulative precipitation deficits may be useful proxies to estimate hydrological drought, but that there is a large variability of relevant lag times and time accumulation periods from place to place due to catchment storage processes. A more comprehensive approach to understand drought-generating processes relevant for forecasting of hydrological drought involved the development of a drought typology, which describes how meteorological drought propagates into hydrological drought. A generic framework explains for 8 different drought types how precipitation and temperature anomalies control the conversion of a meteorological drought into a hydrological drought. In the last research line the potential predictability of soil moisture and streamflow drought was investigated using a hydrological model. The two drought types are predictable until lead times of approximately 7 days and 2–3 days, respectively, when using initial soil moisture information and climatological atmospheric forcing. Using also initial snow information and seasonal weather forecasts as forcing, the predictable lead time doubles in case of soil moisture (2 weeks) and triples for streamflow (1 week). Furthermore, the study demonstrates that soil moisture and streamflow forecast skills increase with increasing initial soil moisture anomalies, which implies an increased predictability for very dry soil moisture conditions.
VULNERABILITY AND RISK REDUCTION AT CASE STUDY SCALE
A detailed inventory and an ex-post evaluation of past practices to cope with drought were performed in the six selected Case Studies areas to support identification and selection of options for policy development and drought risk reduction. Stakeholders still perceive significant gaps in the establishment of preventive measures, structured drought planning, and a poor tradition of ex-post evaluation. In a few Case Study areas a detailed economic impact analysis was performed, which generated interesting outcome, in particular for the agriculture and energy sectors (producers and consumers are differently affected, i.e. winners and losers).
Policy recommendations for future drought risk reduction were formulated for the Case Study areas, on the basis of: (i) anticipated future drought impacts and vulnerability, (ii) assessment of alternative options, (iii) recommendations made by the members of the Case Study Dialogue Fora, and (iv) provisions in EU and international guidance documents. A common methodological framework was formulated for each research activity, however, in the implementation phase differentiation among Case Studies was introduced, to account for differences in spatial scales (regional, river basin, national), data availability, policy priorities in each of the areas, and stakeholder involvement. Each Case Study concluded on strategy-related actions for reducing drought risk and, as a further step, on recommendations/outcomes that were of value for the more generic Pan-EU drought assessments.
The analysis of future vulnerability to drought and of potential impacts as a result of future drought exposure and socio-economic development scenarios focused on three sectors: agriculture, urban water supply and energy production. Qualitative (surveys, stakeholder consultation through Case Study Drought Dialogue Fora) and quantitative (modelling) approaches were adopted. Agriculture remains the sector most affected by drought, whereas the risk of forest wildfires is expected to increase also in the future. Factors that may reduce future vulnerability are: (i) improvement of institutional framework on drought, (ii) overcoming policy gaps, (iii) establishment of monitoring & early warning systems, (iv) development of Drought Management Plans, (v) increasing water-use efficiency, and (vi) use of alternative water sources, and increasing user awareness. Factors that may increase future vulnerability are: (i) conflicts due to water scarcity, (ii) status of water sources (quality and quantity), (iii) increased variability of water availability due to climate change, and (iv) economic development patterns. Water scarcity was mentioned as an important exposure-related factor of vulnerability to drought. Any effort to cope with water scarcity will thus contribute to reducing vulnerability to drought. Stakeholder anticipation is that the socio-economic development pattern is the most influencing factor on future drought-related risks, even more than climate change.
Following the analysis of future impacts and vulnerability, options for drought risk reduction were analysed (quantitatively and qualitatively) using also input from stakeholders (next round Case Study Drought Dialogue Fora)). The identification of options was also based on recommendations from EU (e.g. Blueprint to safeguard water resources) and International (e.g. Hyogo Framework for Action) guidance documents. The study learnt that there a wide range of measures that can be implemented for drought risk mitigation and improved drought management. Their applicability is context-specific as susceptibility to drought and coping capacity are local in nature and defined by local/ regional characteristics. Currently already drought-prone and affected areas will more easily develop and implement options adapted to the local context, compared to areas that do not experience droughts frequently. The selection and implementation of drought risk mitigation options depends on the characteristics of the ongoing or upcoming drought. An 'auto-pilot' approach to drought risk mitigation is not possible: even in a particular area, drought intensity and impacts are not the same and particularities of measures may change from one event to another. All Case Studies concluded on proposals for enhancing drought management and summarised the guiding principles for their implementation, focusing on legislative requirements, technical capacity (tools, data requirements), financial issues, environmental considerations, and any other constraints for the strategy development or its improvement.
DROUGHT SENSITIVE AREAS IN EUROPE; IMPACTS, VULNERABILITIES & RISKS
The European Drought Impact Report Inventory (EDII) was developed during the project and extensively used for analysis of impacts to identify drought sensitive areas in Europe. EDII contains categorized data on drought impacts derived from a variety of text sources. The EDII was made publicly available through the website of the European Drought Centre (www.geo.uio.no/edc). Early 2015, the European Drought Impact Report Inventory (EDII) database contained about 5000 entries. The impact reports span the period 1900 to date, but most entries relate to impacts that occurred since the 1970s. Impact reports on agriculture, water supply, water quality and freshwater ecosystems were found to have dominated in most regions during most events. We found and illustrated that at the pan-European scale, the relative percentage of reported impacts by country suggests slightly higher relative importance of agriculture in the South and East, and of public water supplies in the South and West. The importance of energy and industry impacts, for example, but also other impact types, appears to be more country or region-specific as well as event-specific, e.g. dependent on whether an event is short and accompanied by a heat wave, such as the event of 2003 in Central Europe or characterized by a long multi-year deficit in surface water supplies such as the event of 2004-2008 in the Iberian Peninsula.
The EDII was combined with the natural hazard database (EDR) showing that the major large-scale drought events that occurred in Europe over the past few decades had SPI values below -2 for extended regions. However, a variety of drought impacts have also occurred for values below about -1 or -1.5. Regional studies found that impacts on water supplies, freshwater ecosystems, and energy and industry, which all rely on surface or groundwater resources, are more sensitive to long-duration droughts. These are best indicated by SPI-12 to -24. There are regional differences. Generally, the Mediterranean, which has adapted to lower precipitation and longer dry periods with reservoir storage and irrigated agriculture is less sensitive to short precipitation anomalies and is affected most by long accumulated precipitation deficits. The remainder of Europe responds to more seasonal droughts.
In addition to the impact analysis approach using EDII, quantitative drought impact data (crop yields, and area-burned by forest fire reported by member states) were used. In a pan-European study crop yield anomalies for five major crops were correlated with time series of SPI and SPEI. Correlations are low, but correlation patterns for wheat and barley show stronger links with SPI accumulated over three months in the flowering period. The potential to predict above normal wildfire activity was assessed using meteorological drought indicators (SPI). The study illustrates significant predictability of wildfire activity in southern Europe several months in advance.
Multi-model ensembles of GHM-GCM were used to investigate the impact of a changing climate on crop yields across Europe. For both the near future and the far future, a substantial increase in mean yield reduction was found for the mid-latitudes in Europe due to climate change (SRES A2 scenario). Because the southern regions already suffered from a high yield reduction under the current climate, changes were smaller. Northern regions in Europe are projected to have a lower yield reduction. However, the spread in individual model results was high indicating the high uncertainty in the changes towards the future.
Assessment of vulnerability to drought on a pan-European scale has been elaborated through evaluating and comparing vulnerability thresholds and explore the link between the magnitude of impacts and socio-economic, environmental and management factors influencing drought resilience through a detailed factor-based vulnerability assessment and mapping exercise focusing at the pan-European scale. For this purpose, a wealth of data was collected and evaluated to quantify vulnerability factors. The three vulnerability components (exposure, sensitivity, adaptive capacity) were broken down in 19 individual factors, such as physical drought characteristics, water stress, water body status (WFD), population, legal/institutional aspects, socio-cultural circumstances, water infrastructure, financial and economic conditions. After mapping individual factors these were aggregated to produce maps by component, and eventually the map of vulnerability to drought at pan-European level was produced. The results of this work lead to a number of key findings that show variation and relevance of factors vary for spatial scales around the NUTS-2 level, for situations and particular sectors. The maps resulting from this vulnerability assessment indicate trends of the level of vulnerability across Europe that can contribute to inform the design of future drought management policies and guidelines.
The vulnerability maps, which typically were based on a combination of relevant, more or less subjectively weighted vulnerability factors, served to complement the modelling and drought risk mapping. For the risk assessment the DROUGHT-R&SPI project took advantage of reported drought impacts from the EDII database and empirically estimated risk as the Likelihood of Impact Occurrence (LIO). The approach assumes that drought impacts are symptoms of vulnerability and can therefore serve as a proxy for part of the vulnerability. Three generations of risk model development and risk map construction were developed within the project. The EDII version (i.e. reflect number of impact reports) and the number of explaining variables (single variable versus multi variable) are the main differences. To map the risk of drought, LIOs were displayed for four impact categories “Agriculture & Livestock Farming”, “Public Water Supply”, ”Energy & Industry” and “Water Quality” for given hazard levels (index of SPEI), which correspond to particular drought return periods. The maps show interesting spatial variations of drought risk. For moderate drought, risk is highest in the Mediterranean, in particular for Agriculture and Livestock Farming, and in the densely populated areas of central Europe, in particular for Water Quality and Public Water Supply, a pattern that generally dominates the maps. However, for more severe drought hazard (SPEI values of -2.5 to -2), risk increases everywhere, but with the most differences spatially and for different considered categories. LIO of impacts on Energy and Industry only increase for the most severe drought hazard. For the most severe drought, risk is high almost everywhere.
The impact reports in the EDII evidenced that there are European regions where awareness of drought hazard has increased in recent years, and resilience building likely took place. The dynamics of drought vulnerability to be expected was analysed based on the reported drought impacts over time. Under the assumption that the composition of impacts represents a particular vulnerability profile at the time of a drought event, the study revealed that vulnerability is dynamic, i.e. changes from event to event. The distributions of impact types suggest at gradual changes following a certain trend over time but also following particular legislative changes that may have influenced this pattern. Overall, national water management plans and the European Framework directive may have decreased the vulnerability for impacts related to water resources management, but, may also have increased the awareness and recognition of environmental impacts (and with that seemingly the vulnerability) as a result of better monitoring and public interaction. Untangling these aspects for the past will be a key towards understanding the combination of future hazard predictions and vulnerability changes that lead to changes in drought risk in the future.
The studies also have revealed important issues complicating analyses. For instance, the spatial differences of impact reporting at administrative regional levels is often too coarse to associate it to other relevant units such as river basins, thus limiting the use of hydrological indicators. Hydrological indicators are also more difficult to obtain, as no EU wide observation based data product exist that is freely accessible to support a suitable analysis at the pan-European scale. Commonly, large-scale models do not account for important human alterations to river flow in the necessary detail.
Assessment of vulnerability to drought and associated risk at a pan-European level is a comprehensive task because it is different for individuals, sectors and nations, depending on several biophysical and socio-economic factors. Thus all indices, models, and consequently the maps are based on a number of compromises due to the aim for pan-European comparability.
DROUGHT INDICATORS
The complex nature of drought related phenomena also implies that there cannot be one single quantification of drought, satisfying all monitoring and early warning requirements. Hence, it is unlikely that one single drought monitoring and early warning system will satisfy the needs of all potential users. Therefore, a prerequisite for an efficient monitoring system is to identify the relevant stakeholders and to formulate well defined monitoring and early warning objectives, incl. indicators. Drought indicators are usually associated with so-called triggers, i.e. indicator values at which negative drought impacts are expected. Typical triggers can be classified into: (i) probabilistic triggers, which relate to the occurrence frequency of drought events; (ii) deterministic triggers, which refer to indicator values that should not be reached (e.g. environmental flow regulations); and (iii) empirical triggers which are based on empirically linking past drought impacts to drought indicators using statistical techniques.
The pan-European analyses and modelling of drought impacts and vulnerability served as a basis to provide recommendations for impact-related drought indicators and indices that will complement and expand existing Pan European hazard assessment, vulnerability monitoring, and early warning suitable to inform risk reduction measures in EU policies. Main observations are:
• Drought indicators showing the best link to drought impacts varied regionally at spatial scales smaller than the national or continental scale. Scale differences between impacts reported for administrative units (NUTS) and river basins used for hydrological monitoring complicate analyses.
• SPEI showed closer links to impacts than SPI, confirming other studies on the correlation between SPI (or SPEI) and different hydrological, agricultural, and ecological response variables. In the case of the quantitative crop yield anomalies SPEI links were not substantially stronger than the SPI.
• Best predictive indicators for particular drought impacts depend on sector and on sectorial management practices, e.g.:
- For rainfed agriculture, predominantly short to intermediate accumulation periods of SPI or SPEI (about 2-6 months) are better linked to impact occurrence.
- For irrigated agriculture, water supply, energy and industry, generally longer accumulation times of SPI and SPEI or combination of short and long accumulation periods are better linked to impacts, with optimal choices varying.
- For water-borne transport on the large rivers rather short accumulation periods of SPI or SPEI (about 1 2 month) likely will apply, although not specifically investigated in this project.
To fully account for these practices, more information and complex models and/or hydrological indices will be necessary (see below, detailed operational scale).
• Different drought impacts occur during different seasons, highlighting the need for a season-specific analysis, indicator or to otherwise account for this. Unfortunately, many impact reports only allow determining the year of occurrence, not the month or season.
• The link between drought indicators and impacts is often non-linear and affected by competing influences (e.g. fuel availability for wildfires; positive effects of drought-related weather during particular phases of crop growth or for particular crops) that can be revealed through analysis with well-chosen indicators.
• Little information is available for the onset, duration and end of impacts, making it difficult to analyse impacts over prolonged droughts as common in the Mediterranean and in large groundwater bodies in other parts of Europe (e.g. UK), and thus to select specific indicators that take into account the severity and duration of these long droughts.
However, at the detailed, operational scale (river basin scale, or in some cases national scale, like the Netherlands) use of tailored indicator systems is more efficient to communicate with stakeholders about the ongoing and forecasted hazard, the usually manifold impacts, the potential measures to be taken and the real time drought management, to better link with impacts and also for real time management of events. The physical meaning is hard to communicate to stakeholders. For these purposes, the use of a Decision Support System for drought management is almost indispensable to accomplish appropriate planning for drought and also an effective real time decision making. The involvement of stakeholders is essential, not only to support decision making during the drought (incl. the onset and recovery phases), but also for the planning of structural measures to reduce future drought vulnerability, hence to lower the risk (pro-active).
SCIENCE-POLICY INTERFACING; DROUGHT DIALOGUE FORA
Drought Dialogue Fora (DDFs) were further been developed. In most Case Study areas several rounds of fora (CS DDFs) were organised. These fora were excellent platforms through facilitated discussions to collect information on factors that governed vulnerability and risk of past drought or will determine these in the future. The dialogue was also fundamental in testing the developed methods (see Vulnerability and risk reduction at Case Study scale, above). Stakeholders in the CS DDFs prefer regular contact and cooperation with scientists in an open discussion atmosphere, rather than simply participating in annual events/workshops with single-way presentations and short discussion time in usually considerably crowded places The outcome of the CS DDFs very much contributed to defining methodologies to assess drought sensitive areas in Europe: through pan-European analyses of impacts, vulnerabilities and risks (see above). The methodologies for large-scale analyses were further shaped and tested in three pan-European Drought Dialogue Fora (pan-EU DDF). On the other hand the pan-EU DDFs helped to interpret options proposed in EC policy documents (e.g. Blueprint) to become meaningful at the Case Study scale.
While the 1st pan-EU DDF (Nicosia, Cyprus) served to explore requirements of policy makers and coordinators of other large national and international projects in terms of methodologies for drought vulnerability and risk assessment (see 1st Periodic Report), the 2nd pan-EU DDF (Brussels) focussed on discussion and feedback to four issues related to drought vulnerability assessment and potential risk mapping at the pan-European scale. The four issues were: (i) the scale and use of risk maps, (ii) specific impacts (and hence sectors) of drought that are relevant in different geo-climatic regions in Europe, (iii) factors that affect vulnerability to drought, and (iv) potential models to assess this vulnerability. Options for the size of spatial units were indicated, as well as factors that either increase vulnerability (e.g. high population density in urban areas, high dependence of energy production on freshwater) or decrease vulnerability (prominent stakeholders involvement, well-developed regulatory framework related to drought, high degree of social awareness about drought risk). On the 3rd pan-EU DDF (Brussels) preliminary project’ outcome was presented and discussed. This involved: (i) European Drought Impact Database (EDII); (ii) first pan-EU maps of drought vulnerability factors, (iii) drought: monitoring, forecasting and indicators; and (iv) how to upscale context-specific CS experiences to the pan-European scale. The 4th pan-EU DDF had a more scientific nature and was linked to the final DROUGHT-R&SPI Conference in Valencia. Final vulnerability and risk maps were presented there. Associated developments were presented and discussed, such as the guidelines for ecological flows (e-flows) and implementation of pan-EU knowledge and experiences in the Member States (Portugal, UK and Jucar).
Potential Impact:
POTENTIAL IMPACT
The adoption of a trans-disciplinary approach and the incorporation of stakeholders and policy makers across scales right from the beginning led to DROUGHT-R&SPI having already impact on Drought Management Planning, i.e. helping to formulate reduce drought vulnerability and policy options in the Case Study areas. In particular, in the Mediterranean cases (Portugal, Jucar River basin, Po River basin and the Island of Syros) were in close co-cooperation happened with stakeholders and policy makers to formulate options. In the less-water stresses cases (Switzerland, the Netherlands), the DROUGHT-R&SPI project was not steering, but contributed to planning to make the countries less vulnerable to drought through associations with ongoing national drought initiatives (Switzerland: DROUGHT.CH project; and the Netherlands: Delta programme, securing fresh water supply programme).
The DROUGHT-R&SPI project has provided scientifically sound and new knowledge that is ready to be used to reduce drought vulnerability and associated risk in Europe through:
• Development of a methodology to assess and map pan-European drought vulnerability at the level of the NUTS2 scale, which is based on 19 vulnerability factors that address exposure, sensitivity and adaptive capacity as key vulnerability components
• Development of a methodology to assess and map pan-European drought risk taking advantage of a wealth of collated historic impact reports (European Drought Impact Report Inventory, EDII). Risk is estimated as Likelihood of Impact Occurrence (LIO). LIO maps were displayed for four impact categories “Agriculture & Livestock Farming”, “Public Water Supply”, ”Energy & Industry” and “Water Quality” for given hazard levels (expressed as Standardized Precipitation Evaporation Index), which correspond to particular drought return periods.
The maps resulting from the vulnerability assessment indicate the spatial distribution of the level of vulnerability across Europe, incl. causing factors. The generated vulnerability maps can be combined with the impact-based risk maps to identify drought sensitive areas and impacted sectors for different return periods. The outcome of this analysis can contribute to fine tune the design of future drought management policies and guidelines.
The pan-European vulnerability and associated risk maps have a generic nature because of their pan-European inter-comparability. These large-scale maps can be complemented with outcome of: (i) the analysis of the future vulnerability to drought and of potential impacts at the Case Study scale as a result of future drought exposure and socio-economic development scenarios, and (ii) policy recommendations for future drought risk reduction.
In addition to the above-mentioned vulnerability and risk maps, the project’ results can assist further development of drought management plans, incl. adequate drought monitoring and early warning (DMEW) as part of River Basin Management Plans, EU Water Framework Directive (WFD). Drought indicators are an important DMEW component. Guidance is provided on drought indicators, which can have a different nature (probabilistic, deterministic, empirical). Awareness indicators are recommended for large-scale studies and for informing the general public. These indicators may have a composite nature combining different drought types, provided that these are well communicated. More specific, impact-based indicators are required when decisions need to be taken. The studies using the EDII prove that these indicators should be specific to the vulnerability of sector)s) and potential impacts of the region or river basin. At the pan-European scale, DMEW mainly benefits from transboundary communication in its current state. However, it could become an important tool for future policy choices (e.g. drought declaration, aid and compensation) and risk management decisions at the national and multi-national scale, if it was linked more specifically to drought impacts Comprehensive, real-time drought management at the river basin scale often uses a framework of monitoring-modeling-Decision Support Systems driven by operational drought indicators. Monitored and forecasted spatial-distributed states of hydrometeorological variables and fluxes are translated in indicators tailored to the impacted sectors. Indicators are also simulated to explore effects of potential short-term measures that reduce impacts. Knowledge obtained on drought impacts is already used to explore if information provided by the European Drought Observatory can be enhanced.
The project will provide support to adaption to climate change, in particular through:
• The European Drought Reference database (EDR) that can help to test the next generation of global hydrological models (e.g. better subsurface storage representation, higher spatial resolution) driven by recent climate projections. DROUGHT-R&SPI has provided a number of methods on how drought characteristics identified with multi-model ensembles can be tested against observations and documented events
• Projections of different drought types (meteorological, soil water, runoff, river flow) using different future climate forcings. Different uncertainty assessment methods (e.g. multi-model, signal-to-noise) have been developed
• Different drought identification approaches (e.g. standardized variables, threshold approaches). Moreover, the project has illustrated the substantial influence of the identification method on hydrological drought whether it considers adaptation to a gradually changing hydrological regime or not. The latter usually is applied, with e.g. the period 1971-2000) as reference
• Quantitative Impacts on future crop yield, incl. ways to express uncertainty
• The EDR that can assist further development of adaptation measures by using exposure data form historic and projected droughts. The EDII can help with reported manifold impacts as response to drought and the links between the impacts and meteorological indicators (Standardized Precipitation Index, SPI; Standardized Precipitation Evaporation Index, SPEI.
The DROUGHT-R&SPI communicated with several policy makers in Brussels, i.e. DG Environment, DG Climate and DG Enterprise, as well as with Commission Services (JRC, EEA) dealing with hydrological extremes. This was done through the pan-European Drought Dialogue Fora, targeted meetings, the WFD-CIS Expert Group on Water Scarcity and Drought, the WFD-CIS Working Group on Groundwater and a recent EEA meeting on climate and weather-induced extremes.
The European Drought Impact Report Inventory (EDII) is a powerful database to be used in the European Research Area (ERA). It can be extremely suitable, for example in Water and Climate studies, disaster risk studies that aim to further develop or to introduce new measures and technologies to enhance effective response capacity to extreme weather and climate events affecting the security of people and assets.
The project also impacted the implementation of the WMO/Global Water Partnership Integrated Drought Management Programme (IDMP) in ten Central and Eastern European countries (CEE). The 7-step approach to develop a national drought management policy and plan was reviewed and concrete suggestions were given to improve. The IDMP-CEE approach will be the basis for the drought chapter in the next cycle of River Basin Management Plans that the EU countries compile as response to the EU Water Framework Directive (WFD). The non-EU countries in CEE are encouraged to adopt the IDMP approach in their national and regional water management and policy.
The DROUGHT-R&SPI project can also assist implementation of several priorities of the Hyogo Framework for Action, in particular to: (i) identify, assess and monitor disaster risks and enhance early warning, (ii) reduce the underlying risk factors, (iii) strengthen disaster preparedness for effective response at all levels, and (iv) use knowledge, innovation and education to build a culture of safety and resilience at all levels. Further development of these priorities is essential as concluded at the recent World Water Forum in Korea (April 2015), where it was discussed that globally, water-related disasters account for 90% of all natural disasters, and impacts from water-related disasters can in economic terms be up to 15% of annual GDP for certain countries.
The project already has significant outreach during the project life time due to Drought Dialogues in the Case Study areas and four pan-European Drought Dialogue Fora (pan-EU DDFs). In some of the pan-EU DDFs there was also interaction with several drought-related EU projects, initiatives of International Organisations (e.g. UNESCO-IHP, WMO/GWP) and large national projects (e.g. SCARCE). The project further trained about 10 PhD and postdocs. The Drought Summer School in Syros (Greece) increased knowledge of 25 PhD students, postdocs and junior scientists and allowed them to build a network. Preliminary Project’ outcome was already presented and discussed at tens of conferences and workshops. The DROUGHT-R&SPI Conference in Valencia (March 2015) allowed the consortium to present the final outcome and to discuss it with over 150 people working on drought. Project’s results will be available beyond its lifetime through the Science-Policy Briefs, 35 two-page flyers, about 50 peer-reviewed papers, and 35 DROUGHT-R&SPI Technical Reports. The website will be accessible until at least 2020. The European Drought Reference database (EDR) and the European Drought Impact Report Inventory (EDII) and the R-software package to calculate Standardized Climate Indices (SCI) will also help to remain the DROUGHT-R&SPI project still visible for a number of years.
DISSEMINATION ACTIVITIES AND EXPLOITATION RESULTS
In the 2nd Science-Policy Brief (SPB), the DROUGHT-R&SPI consortium decided for two perspectives, both starting from the urgent need to reduce drought impacts, through increased knowledge, drought management plans and an improved science-policy interfacing that will lead to better preparedness and reduced vulnerability to future drought and the risks they pose for Europe. One perspective is that of the “Tools and Instruments” that contribute to Improved Drought Management and Preparedness. The second perspective resulted from realising that “Science-Policy Interfacing” alone is not enough in the chain to implement procedures and measures in relation to drought. There is a need to move beyond SPI towards SPI-Implementation. After ‘science’ and ‘policy’ have ‘interfaced’ to enable formulation and implementation of policies, guidelines, procedures and the like, the Member States’ water managers have to ensure the correct ’Implementation’ of the measures, in cooperation with stakeholders.
The drought-related messages to different audiences started from the view that drought is a multi-stakeholder and multi-dimensional phenomenon for which uniform measures are difficult to provide and implement, and therefore management plans need to include local demands and make use of tailored advice to deal adequately with the uncertainties involved in dealing with drought. An on-line questionnaire survey was designed to collect views from different people dealing with drought. Eventually messages were developed for three major stakeholder groups: i.e. Public Authorities, Businesses and Citizens. The desire for water saving technologies, techniques and drought management strategies and plans, in other words pro-active measures, were topics to be addressed in all messages. The same applies to communication on drought, but the means are different for the categories. Businesses and Public Authorities appear to have similar preferences in way they would like to obtain information (e.g. meetings, information sessions/conferences and educational programmes), whilst Citizens have a preference towards the media and smart phone applications.
The “International Conference on Drought Research and Science-Policy Interfacing” was the final event of DROUGHT-R&SPI project. The conference was held from 10-13 March 2015 in Valencia, Spain. With a total number of about 150 participants, it included 61 oral presentations and 36 posters. A total of 73 of these contributions were published in a book (CRC/ Balkema). The conference incorporated the 4th pan EU DDF. Drought related research were discussed, incl. progress on response policies to foster development of Drought Policies and Plans to reduce risk and vulnerability, and to enhance preparedness and resiliency. The conference addressed drought identification and characterization, drought assessment (impacts) and drought policies to develop drought management plans at different scales (river basin, national and international). The conference was an outstanding platform to convey the main results from the project and it served as a forum for discussion among scientists researching drought, stakeholders, water managers, experts and representatives of authorities on experiences, about the needs and challenges for drought management.
The 1-week Summer School on ‘Drought Hazard and Management: Challenges in a Changing World’ was held in June 2014 in Hermoupolis Syros, Greece. It brought together 25 PhD candidates and young researchers (14 women; 11 men; 6 participants had a non-European nationality and came from China, India, Algeria, Brazil, Iran, German/Uruguayan). The School included lectures, workshops, a poster session, where participants presented their on-going research activities, a field trip and a closing session, where participants presented the outcomes from the Workshops. In addition to obtaining knowledge about drought and skills to handle drought, the school offered a perfect platform for networking, which should not be underestimated. Previous school with a similar setup have proven to be the basis for long-lasting working contacts.
In the project about 10 PhD students and postdocs were further educated in the various aspects related to drought. About 35 DROUGHT-R&SPI Technical Reports were written, which explain in detail the methodologies that have developed and the results that were obtained. Around 35 two page flyers have been made available, which provide specific project’ outcome in two pages. Two Science Policy Briefs were written. The reports, flyers and Science Policy Briefs are freely available on the internet. About 50 peer-reviewed papers are published in scientific journals dealing with natural hazards, water management, hydrology, climate, environment. Almost all are open access. In addition to the Drought Dialogue Fora, project results have been presented and shared with colleagues on almost 100 conferences, workshops, assemblies, symposia, seminars. The content of many of these presentations can be found in abstracts or conference proceedings (see 2nd Periodic Report, Section 4.5 and 1st Periodic Report, Section 4.2).
The project website (www.eu-drought.org/) provides in addition to information about the DROUGHT-R&SPI project, links to the DROUGHT-R&SPI Technical reports, flyers, Science Policy Briefs and peer-reviewed papers. The website will remain accessible at least until 2020. The DROUGHT-R&SPI website is closely linked to the website of the European Drought Centre (EDC, www.geo.uio.no/edc). The EDC website houses both the European Drought Reference (EDR) database, which provides detailed meteorological and hydrological drought statistics, and the European Drought Impact Report Inventory (EDII), which stores user-provided information about drought impacts. The EDC platform has been used in the project for an interactive web based discussion forum to further develop the EDII. This was particularly linked to the Drought Summer School in Syros.
A software package in R (Repository CRAN) was developed. The R package “SCI” calculates Standardized Climate Indices, such as SPI, SRI or SPEI. SCI is a transformation of (smoothed) climate (or environmental) time series that removes seasonality and forces the data to take values of the standard normal distribution. SCI was originally developed for precipitation. In that case it is known as the Standardized Precipitation Index (SPI).
List of Websites:
The public project website can be found at: http://www.eu-drought.org/
Contact details:
dr.ir. Henny A.J. van Lanen
Coordinator EU-FP7 DROUGHT-R&SPI
Wageningen University
P.O. Box 47
6700 AA Wageningen
The Netherlands
Visiting address:
LUMEN (building 100), room 0.C29
Droevendaalsesteeg 3a
6708 PB Wageningen
T: +31 317 48 2418 (office)
T: +31 317 48 2778 (secretary)
M: +31 6 835 96 767
F: +31 317 41 9000
European Drought Centre: www.geo.uio.no/edc
WATCH (WATer and global Change): http://www.eu-watch.org/
DROUGHT-R&SPI SPI (Fostering European Drought Research and Science-Policy Interfacing): http://www.eu-drought.org/