Floods and droughts - outlook from the University of Kassel

Assessment versions

Published (reviewed and quality assured)

• No published assessments

Rationale

Justification for indicator selection

Extreme flood events can cause tremendous damage to economy and ecology and, in the worst case, bear enormous risks for life. At the same time, climate change scenarios generally imply an increase in rainfall variability and, on global average, an increase in total precipitation which could lead to even more frequent and severe floods. Modeling the impact of climate change on future river floods, flood frequencies or the risk of flooding thus draws increasing attention both from a scientific and political point of view.

As compared to floods, droughts are often perceived by society to play a less dominant role when thinking of natural hazards. This may be caused by the circumstance that, unlike the effects of a flood which can be immediately seen and felt, droughts build up rather slowly, creeping and steadily growing. Whatever the reason, this perception has led to a relative disregard for droughts, despite the fact that they regularly cause serious damage to economy, society and the environment both in the affected areas and further afield. According to data compiled by the National Drought Mitigation Center (2001), the average annual economic costs and losses through droughts in the United States (US$ 6-8 billion) are more than double the average annual costs for floods (US$ 2.4 billion). In the early 1990s Europe experienced severe droughts resulting in significant economic and environmental costs. The damage in Spain (1992-95), where the drought affected about 500 000 hectares of irrigated land alone in the Guadalquivir river basin, was estimated at several billion Euro (Garrido and Gomez-Ramos, 2000).

Droughts were long considered a hazard affecting mainly developing countries, but public awareness has increased in the past years in the industrial countries especially with respect to the climate change issue predicting more extreme hydrological conditions (Demuth and Stahl, 2001). Since the demand for European water resources has increased in the past decades, future conflicts between human requirements (commercial, social and political) and ecological needs are likely to increase, too. These conflicts are most critical and intensive during severe and extensive droughts.

Scientific references

• No rationale references available

Indicator definition

According to WaterGAP model the indicator 'floods and droughts'  provides the following objects:

• Drought events and deficit volumes are presented in the form of  the drought frequency distributions. Within this indicator the concept of river flow drought (or hydrological drought) is adopted.
• Floods are presented in the form of the flood frequency distributions or flood discharges. Flood is defined strictly in terms of discharge. To answer to what extent a given discharge value is related to a realflooding, in terms of bursting river banks and setting a considerable area under water, in particular a high-resolutions elevation model is required.

Units

No units have been specified

Policy context and targets

Context description

The aforementioned indicator and indicator of water use stress can be used to monitor some policies at global, regional and national levels. It provides, for example, the information on efficiency of water-use management plans.

Global policy context
At the global level problems of fresh water use and water stress are becoming ones of the most actual. The central aims were emphasized within UN "Millennium Development Goals" (7^th goal to ensure environmental sustainability)  and include reduction of proportion people without access to safe drinking water.

Pan-European policy context
In 2002 the EU launched a Water Initiative (EUWI) designed to contribute to the achievements of the Millennium Development Goals (MDGs) and World Summit for Sustainable Development targets for drinking water and sanitation, within the context of an integrated approach to water resources management. The EUWI covers EU region as well as EECCA regions.

The UNECE Convention on the Protection and Use of Transboundary Watercourses and International Lakes   was signed by 34 UNECE countries and the European Community. The Convention establishes main principles and rules for its Parties to develop and promote coordinated measures of sustainable use of water and related resources of transboundary rivers and international lakes, as well as of institutional mechanisms to be created for it. The UNECE Convention on the protection and Use of Transboundary Watercourses andInternationalLakes is an important instrument for the protection of freshwater resources and the development of transboundary water cooperation.

EU policy context
Achieving the objective of the EU's Sixth Environment Action Programme (2001-2010), to ensure that rates of extraction from water resources are sustainable over the long term, requires monitoring of the efficiency of water use in different economic sectors at the national, regional and local level. The WEI is part of the set of water indicators of several international organisations such as UNEP, OECD, EUROSTAT and the Mediterranean Blue Plan. There is an international consensus about the use of this indicator.

The indicator describes how the total water abstractions put pressure on water resources identifying those countries having high abstractions in relation to their resources and therefore prone to suffer water stress. The changes in WEI help to analyse how the changes in abstractions impact on the freshwater resources by adding pressure to them or by making them more sustainable.

There is a number of agreements relate to European river water use management, for example of the oldest one is the International Commission for the Protection of the Rhine (ICPR). (Basel on July 11, 1950).

EECCA policy context
EECCA Environmental Strategy promotes sustainable water use based on long-term projection of available water resources. The EECCA environment strategy has actions on development and implementation of integrated water management programmes based on river basin principles.

Some sub-regional policies aim to stimulate development and implementation of action plans to improve water resource management systems.

A regional Cooperation strategy to promote the rational use and conservation of water resources in Central Asia focus on the sustainable use of freshwater in the Aral SeaWaterBasin. The strategy helps to support achievability of targets set in the Aral Sea Basin Water Vision 2025 developed with support from UNESCO (SABAS vision). The document provides recommendations for water distribution, particularly within agriculture sector, as well as an accent on improving hydro electricity technologies with 'less losses of water' over the 2025 horizon.

Number of transboundary rivers negotiations focus on sustain river's water use and are implemented for such river  basins as Neman (Nemanus) and Western Dvina (Daugava); also for Dniester between Ukraine and Moldova.

LInks to other related policies:

EECCA Environmental Strategy

UNECE Convention on the Protection and Use of Transboundary Watercourses and International Lakes

Water Initiative (EUWI)

Aral Sea Vision Cooperation Strategy

UN 'Millennium Development Goals"

IUCN Water and Nature Initiative

River Basin Commissions for Daugava and Nemunas

Transboundary Cooperation and Sustainable Management of the Dniestr River

International Commission for the Protection of the Rhine (ICPR). (Basel on July 11, 1950)

Targets

Global level:
There is no specific targets, however, some of them could influence indirectly on the indicator's issues in particular, share of domestic water withdrawal:

• to halve by 2015 the proportion of people who are unable to reach or afford safe drinking water (MDGs)
• to launch action plans to achieve aims in accordance with MDGs
• to improve management of river basin's water use (IUCN)

Pan-European level:
There is no qualitative targets. However several document put some action oriented targets, for instance, EU Water Initiative aims to establish water resource management plans by 2005 and theUNECE Convention on the Protection and Use of Transboundary Watercourses and International Lakes aims to implement rational and sustain water use within all Europe river basins and lakes for all countries (UNECE). Also, UNECE Convention proclaims to establish and implement joint programmes for monitoring the conditions of transboundary waters, including floods and ice drifts, as well as transboundary impact.

EU level
There are no specific agreements as to the quantitative targets related to this indicator. The EU requires all countries to promote sustainable water use based on long-term projection of available water resources and to ensure a balance between abstraction and recharge of groundwater. Both the Water Framework Directive and the 6th Environment Action Programme set out the goal to achieve a 'good status' (ecological, chemical and quantitative) for all EU water bodies by 2015. More generally, a warning threshold of at 20 % water exploitation index is widely used to indicate a river basin is water stressed, while sever water stress is indicated by values above 40 %. While this may indicate strong competition for water resources, this may (but does not necessarily have to) trigger frequent water crises, depending on the socio-economic and environmental context within river basins.

EECCA level

• to sustain water use within river basins (Baltic region, Black sea's region andCentral Asia)
• to implement integrated management systems for water resource use (EECCA environmental strategy)'
• to implement new technologies for irrigation and hydro power plants (Aral Sea Vision, Water and Energy strategy forCentral Asia)
• to contribute more than 20 km³ water for ecological services within Aral region by 2025 (Aral Sea Vision);

Related policy documents

• Sixth Environment Action Programme (decision No 1600/2002/EC)
  DECISION No 1600/2002/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 22 July 2002 laying down the Sixth Community Environment Action Programme
• Water Framework Directive (WFD) 2000/60/EC
  Water Framework Directive (WFD) 2000/60/EC: Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the field of water policy.
• World Summit on Sustainable Development Plan of Implementation
  UN 2002: 'Plan of Implementation of the World Summit on Sustainable Development'

Methodology

Methodology for indicator calculation

Indicators of flood and drought frequency distribution are calculated using the WaterGAP model (Water: Global Assessment and Prognosis; version 2.1). This is a global model that computes both water availability and water use on the river basin scale.

The model, developed at the University of Kassel, Germany, has two main components: A global hydrology model and a global water use model.

WaterGAP's global hydrology model simulates the characteristic macro-scale behaviour of the terrestrial water cycle to estimate water availability. The model uses both land use and climate data at a 0.5 x 0.5 degree latitude-longitude grid. Thus it can compute water availability for both past and present temperature and precipitation regimes, as well as using output from climate models for expected future conditions. A drainage direction map then allows the analysis of the water resources situation (including water stress) in all larger river basins. This methodology allows calculating water related indicators both on the country level and on the river basin scale, depending on what is more relevant to address specific policy questions.

A more detailed version of the model exists for EEA member states (except Iceland). Compared with the global version, the European model sees (i) improved country-level calibration for domestic water use, based on better abstration data available in this region; (ii) the use of a data on the geographical explicit location of power station and their cooling water requirements; and (iii) estimates of water use for manufacturing presented seperately for six water intensive industrial activities.

In order to derive today's and future drought  and flood frequency distributions, the following procedure is applied in the same manner to all cells of the WaterGAP grid, as well as, for evaluation purposes, to the data of selected gauging stations:

For drought frequency distribution:

• Monthly discharge values are applied. This temporal resolution is used as

a) the month is the usual time unit for river flow drought studies;

b) WaterGAP is based on monthly climate data.

• A drought event is defined to start when the discharge falls below the threshold value and to end when the discharge exceeds the threshold. The deficit volume (or severity) of a such identified drought event is calculated by accumulating the monthly differences between threshold and actual discharge values over time.
• The frequently used median of monthly discharges, here calculated from the time series 1961-90, is applied as a constant threshold value for all data over time (both for today's calculations and for the future).
• The annual maximum series of drought deficit volumes is chosen.
• As drought calculations generally require long discharge series, the 30-year series 1961-90 is applied to calculate today's droughts (data before 1961 is considered increasingly uncertain). For the future scenarios, 30-year projections are applied (i.e. 2011-40 for the 2020s, 2061-91 for the 2070s; for more details on deriving the climate scenarios see here).
• In order to finally derive drought frequency probabilities, the commonly used Log-Pearson Type III distribution is fitted to the ranked annual maximum series. This leads to a statistical distribution function which can be inter- and extrapolated.

For flood frequency distribution:

• Daily discharge values (WaterGAP provides daily output) are applied. This temporal resolution is used as

a) a longer time step is considered not appropriate for flood calculations,

b) the day is the highest temporal resolution for which WaterGAP calculations are conceptually designed (e.g. pseudo-daily rainfall and temperature values are derived from given monthly averages), and

c) no higher-resolution measurement data is available on a global scale for evaluation purposes.

• The annual maximum discharge series is chosen.
• As flood calculations generally require long discharge series, the 30-year series 1961-90 is applied to calculate today's floods (data before 1961 is considered increasingly uncertain). For the future scenarios, 30-year projections are applied (i.e. 2011-40 for the 2020s, 2061-91 for the 2070s; for more details on deriving the climate scenarios see here).
• In order to finally derive flood frequency probabilities, the commonly used Log-Pearson Type III distribution is fitted to the ranked annual maximum series. This leads to a statistical distribution function which can be inter- and extrapolated.

For more description of the WaterGAP model and the indicator see: http://www.usf.uni-kassel.de/usf/archiv/dokumente/kwws/kwws.5.en.htm 

Methodology for gap filling

The water use models for domestic, industrial and electricity-related water use have been calibrated against observed past trends on the country level, where reliable data to do so is available. Where this is not the case, parameters derived from regional averages have been used. For more detail see Floerke & Alcamo (2004) and Alcamo et al. (2003).

Methodology references

No methodology references available.

Data specifications

EEA data references

• No datasets have been specified here.

External data references

• Input data to WaterGAP model - GDP growth - output from IMAGE 2.1 data
• Input data to WaterGAP model - climate projections - output from ECHAM4/OPYC3 model data
• Input data to WaterGAP model - climate projections - output from HadCM3 model data
• Input data to WaterGAP model - electricity production - output from IMAGE 2.1 data
• Input data to WaterGAP model - irrigated area - output from digital map provided by Siebert and Döll (2001)
• Input data to WaterGAP model - population growth - output from UNSTAT data, medium scenario
• Input data to WaterGAP model - population distribution - output from CIESIN data
• Output data from WaterGAP - Flood and drought frequency distribution

Data sources in latest figures

Uncertainties

Methodology uncertainty

List of uncertainties is presented below and particularly can be found in 'Data uncertainty'.

• Baseflow. The accuracy of the baseflow modeled by WaterGAP, however, is not fully evaluated yet as WaterGAP was originally developed for estimating long-term averages where the temporally explicit calculation of the baseflow component was only of secondary interest.

• Within this study only extrapolations up to 200-year droughts are analyzed as, when looking at the model and data uncertainties described above, any statements on more extreme events are not considered justified.

• For the drought frequency calculations the annual maximum series of drought deficit volumes is chosen. Thus, for every year the highest occurring deficit volume is selected. With this simple approach, however, multi-year droughts might be picked more than once.

Except aforementioned uncertainties Floerke and Alcamo (2004) presented a list of some of the main factors determining water use that are particularly uncertain in the European version of WaterGAP. In general, these also hold true for the global version.

Domestic – In most European countries the relationship between future income and water use seems to be well defined. However, in a countries undergoing a major economic transition, it is not possible to define a reliable relationship between income and water use. Another source of uncertainty in estimating future water use in the domestic sector is the future population of water users.

Manufacturing – The water use intensity of different industries is a major uncertainty in most countries. But perhaps more important is the water use of industries that are not now important but will become important over the next 30 years. Key questions are, what will these industries be and how much water will they use?

Electricity Production – Major uncertainties in this sector are the use lifetime of power stations, the percentage of new power stations having tower versus once-through cooling, and their future geographic location. Also important is the uncertainty of future thermal electricity production, and general electricity production trends.

Agriculture – Major unknowns in the agriculture sector are the future extent of irrigated crops, the types of crops to be irrigated, and future climate conditions.

Additional to the above, the uncertainty of the model's estimates on future water availability depend much on the reliability of the land use and climate data used.

Data sets uncertainty

Land cover.

Although in principle WaterGAP is able to take into account the impact of changing land cover on runoff generation via its direct or indirect effect on root depth, albedo, soil moisture and interception, all following low flow calculations are performed without a change in land cover or land use. This is mainly due to the absence of realistic, reliable macroscale land use change scenarios, which are expected to be available at a later stage. For the interpretation of the results, this simplification has to be considered.

Wetlands, lakes and reservoirs.

Although WaterGAP 2.1 distinguishes between lakes, reservoirs and wetlands, at present a rather simple non-linear storage approach is applied for all freshwater storage as no further data on reservoir control or retention behavior is available. As a consequence, WaterGAP will locally underestimate the possible human influence of drought mitigation. Also, there is no information on existing or planned canals for flow diversion implemented in the model. As a consequence, WaterGAP will locally underestimate the possible human influence of flood control.

Additionally, data on current and past water use need to be considered with reservation due to the lack of common European definitions and procedures for calculating water abstraction and freshwater resources. For some countries in the European, Caucasus and Central Asia no reliable time series on water use by sector exist.
These data uncertainties affect model calibration and are propagated through to the modeled results.

Rationale uncertainty

It should be noted, that in this indicator definition a flood is defined strictly in terms of discharge. To what extent a given discharge value is related to a real flooding, in terms of bursting river banks and setting a considerable area under water, is a complex question. To answer it, additional information would be required, in particular a high-resolution elevation model. These data are currently not available on a continental scale. However, knowing the frequencies and volumes of extreme flows is a step towards assessing the risks of river floods.

Within this indicator's definition the concept of river flow drought (or hydrological drought) is adopted. Still, for any drought study a consistent definition is important. The following categories of droughts are frequently used (Tate and Gustard, 2000): a) climatological drought (deficit in precipitation); b) agro-meteorological drought (deficit in soil water); c)river flow drought (deficit in river discharge); d) groundwater drought (deficit in groundwater storage); e) operational drought (conflict of water shortage and water management demands). However, there is no universally accepted definition of drought (Tate and Gustard, 2000).