IWMI
RESEARCH
R E P O R T
Related Publications Mapping Drought Patterns 
Ahmad, S.; Hussain, Z.; Qureshi, A. S.; Majeed, R.; Saleem, M. 2004. Drought mitigation in Pakistan: 
current status and options for future strategies. Colombo, Sri Lanka: International Water 
Management Institute (IWMI). 54p. (IWMI Working Paper 85: Drought Series Paper 3). 133 and Impacts: 
www.iwmi.org/Publications/Working_Papers/working/WOR85.pdf
IWMI. 2004. Assessment and mitigation of droughts in South-West Asia: issues and prospects. A Global Perspective  
Background Document for the Regional Workshop on Drought Assessment and Mitigation. 
Colombo, Sri Lanka: International Water Management Institute (IWMI). 20p.
www.iwmi.org/droughtassessment/files/pdf/workshop%20docs/Background.pdf 
Samra, J. S. 2004. Review and analysis of drought monitoring, declaration and management in 
India. Colombo, Sri Lanka: International Water Management Institute. 38p. (IWMI Working Paper 
84: Drought Series Paper 2).
www.iwmi.org/Publications/Working_Papers/working/WOR84.pdf
Smakhtin, V. U.; Hughes, D. A. 2004. Review, automated estimation and analyses of drought Nishadi Eriyagama, Vladimir Smakhtin and Nilantha Gamage
indices in South Asia. Colombo, Sri Lanka: International Water Management Institute. 29p. (IWMI 
Working Paper 83: Drought Series Paper 1).
www.iwmi.org/Publications/Working_Papers/working/WOR83.pdf
Biophysical
Thenkabail, P. S.; Gamage, M. S. D. N.; Smakhtin, V. U. 2004. The use of remote sensing data for Vulnerability
Index
drought assessment and monitoring in southwest Asia. Colombo, Sri Lanka: International Water 0 - 20
20 - 30
Management Institute (IWMI). 30p. (IWMI Research Report 85). 30 - 40
40 - 50
www.iwmi.org/Publications/IWMI_Research_Reports/PDF/pub085/RR85.pdf 50 - 60
Annual River 60 - 70
Discharge 70 - 80
(1,000 m3 80 - 99
per person) water bodies
0 - 0.001 no data
0.001 - 1 no agriculture
1 - 10
Postal Address 10 - 100
P O Box 2075 100 - 1,000
1,000 - 10,000
Colombo 10,000-100,000
Sri Lanka >100,000
low population
Location no data
127, Sunil Mawatha
Pelawatta
Battaramulla
Sri Lanka
Telephone
+94-11-2880000
Fax
+94-11-2786854
E-mail
iwmi@cgiar.org
Website
www.iwmi.org
I n t e r n a t i o n a l
I n t e r n a t i o n a l
Water Management ISSN: 1026-0862 Water  Management
I n s t i t u t e ISBN: 978-92-9090-711-4 I n s t i t u t e
Research Reports
The publications in this series cover a wide range of subjects—from computer
modeling to experience with water user associations—and vary in content from
directly applicable research to more basic studies, on which applied work ultimately
depends. Some research reports are narrowly focused, analytical and detailed
empirical studies; others are wide-ranging and synthetic overviews of generic
problems.
Although most of the reports are published by IWMI staff and their collaborators,
we welcome contributions from others. Each report is reviewed internally by IWMI
staff, and by external reviewers. The reports are published and distributed both in
hard copy and electronically (www.iwmi.org) and where possible all data and analyses
will be available as separate downloadable files. Reports may be copied freely and
cited with due acknowledgment.
About IWMI
IWMI’s mission is to improve the management of land and water resources for food,
livelihoods and the environment. In serving this mission, IWMI concentrates on the
integration of policies, technologies and management systems to achieve workable
solutions to real problems—practical, relevant results in the field of irrigation and
water and land resources.
IWMI Research Report 133
Mapping Drought Patterns and Impacts:
A Global Perspective
Nishadi Eriyagama, Vladimir Smakhtin and Nilantha
Gamage
International Water Management Institute
P O Box 2075, Colombo, Sri Lanka
i
The authors: Nishadi Eriyagama is a Water Resources Engineer (email:
n.eriyagama@cgiar.org); Vladimir Smakhtin is Theme Leader – Water Availability and
Access (email: v.smakhtin@cgiar.org); and Nilantha Gamage is a Remote Sensing/GIS
Specialist (email: n.gamage@cgiar.org), all at the International Water Management Institute
(IWMI) in Colombo, Sri Lanka.
Eriyagama, N.; Smakhtin, V.; Gamage, N. 2009. Mapping drought patterns and impacts:
a global perspective. Colombo, Sri Lanka: International Water Management Institute. 31p.
(IWMI Research Report 133)
/ drought / impact assessment / indicators / mapping / climate change / river basins /
dams / water scarcity / disasters / risks / precipitation / runoff / soil degradation /
ISSN 1026-0862
ISBN  978-92-9090-711-4
Copyright © 2009, by IWMI. All rights reserved. IWMI encourages the use of its material
provided that the organization is acknowledged and kept informed in all such instances.
Cover photograph shows women walking to collect water in India, 2004 (Source: Ms.
Mamta Borgoyary, Winrock International, India).
Please send inquiries and comments to: iwmi@cgiar.org
A free copy of this publication can be downloaded at
www.iwmi.org/Publications/IWMI_Research_Reports/index.aspx
ii
Acknowledgements
This study was supported by IWMI core funds as part of the broader drought-related research. Mrs.
Prasanna Sambandamurthy (Head of the Library, IWMI, Colombo) conducted a literature search for
this study. We are grateful to Dr. Robyn Johnston (IWMI) and an anonymous external reviewer for
their constructive comments on this manuscript.
Project
This research study was conducted as part of the IWMI Drought Assessment Project, which was
carried out from January 2006 to December 2008.
Donors
This project was funded from the core funds of IWMI during 2007-2008, which consisted of
contributions from the following countries and organizations:
Australia Japan
Canada Netherlands
China Norway
Denmark South Africa
France Sweden
Germany Switzerland
India UK (DFID)
Ireland USA (USAID)
Israel World Bank
iii
iv
iv
Contents
Acronyms and Abbreviations vi
Summary vii
Introduction 1
Data and Methods 4
Discussion 16
Conclusions 19
References 20
v
v
Acronyms and Abbreviations
ATEAM Advanced Terrestrial Ecosystem Analysis and Modelling
FAO Food and Agriculture Organization of the United Nations
GDP Gross Domestic Product
GIS Geographic Information System
IIASA International Institute for Applied Systems Analysis
ILRI International Livestock Research Institute
MCM Million cubic meters
START global change SysTem for Analysis, Research, and Training
UNDP United Nations Development Programme
UNEP United Nations Environment Programme
WB World Bank
vi
Summary
This study examines the global patterns and loss of river flow occurs in areas that do not
impacts of droughts through the mapping of several normally experience climate–driven water scarcity.
drought-related characteristics – either at a country It also illustrates that the African continent is
level or at regular grid scales. Characteristics cover lagging behind the rest of the world on many
various aspects of droughts – from global indicators related to drought preparedness and that
distribution of meteorological and hydrological agricultural economies, overall, are much more
drought risks to social vulnerability and indices vulnerable to adverse societal impacts of
related to water infrastructure. The maps are meteorological droughts. Regions with an unreliable
produced by integrating a number of publicly and vulnerable nature of river discharge, and having
available global datasets. The subsequent the largest drought deficits and durations are
discussion of maps allows a number of policy- highlighted, pointing to the danger of focusing on
relevant messages to be extracted. It appears that drought mitigation measures on river flows alone.
arid and semi-arid areas also tend to have a higher The ability of various countries to satisfy their water
probability of drought occurrence. The report points needs during drought conditions is examined using
out that in drought years, the highest per capita storage-related indices.
vii
viii
Mapping Drought Patterns and Impacts: A Global
Perspective
Nishadi Eriyagama, Vladimir Smakhtin and Nilantha Gamage
Introduction
Drought can be generally defined as a temporary from a global development perspective, to
meteorological event, which stems from a understand the pattern of various drought-related
deficiency of precipitation over an extended period characteristics and impacts worldwide. Such
of time compared to some long-term average characteristics should reflect multiple aspects of
conditions. Drought always starts with a shortage drought, ranging from quantification of drought
of precipitation (compared to normal or average hazard and vulnerability of water resources
amounts), but may (or may not, depending on how systems - to measures of preparedness to face
long and severe it is) affect streams, soil moisture, future droughts. One good way of presenting
groundwater, etc. It is a recurring natural event and diverse materials related to droughts is through
a normal part of the climate of all world regions, mapping, whereby various drought-related indicators
regardless of how arid or humid they are. Droughts can be plotted at a country resolution, river basin
develop slowly, are difficult to detect and have or a regular grid – depending on the type of
many facets in any single region. It is, thus, one indicator and information available.
of the most complex natural phenomena, that is Despite significant drought research, studies
hard to quantify and manage, and has multiple and that deal with the global picture of drought patterns
severe social and economic impacts. The and impacts are limited. Even fewer studies deal
magnitude of these impacts is determined by the with global mapping of drought-related indicators.
level of development, population density and Peel et al. (2004, 2005) conducted an analysis of
structure, demands on water and other natural precipitation and runoff periods (runs) of
resources, government policies and institutional consecutive years below the median for 3,863
capacity, technology, and the political system. precipitation and 1,236 runoff stations worldwide.
These points of departure set the scene and scope Run lengths were found to be similar across all
for this study. continents and climates except North Africa, which
Droughts continue to have significant impacts showed a tendency towards longer run lengths.
in both developed and developing countries. The Run lengths for precipitation and runoff at the same
latter still suffer from droughts the most. Ever- location were found to be similar. Run magnitude
increasing exploitation of water resources and for both precipitation and runoff was found to be
associated water scarcity coupled with the growing related to inter-annual variability, and run magnitude
concern that future climate change will exacerbate of runoff was larger than that of precipitation due to
the frequency, severity, and duration of drought a higher coefficient of variability of runoff. Severity
events and associated impacts explains the of drought events (a total of negative deviations
increasing attention that individual countries are from the median for a run) was found to be
paying to drought-related issues (Wilhite 2005). independent of run length but strongly related to
Since drought is a global phenomenon, it is useful, magnitude. These studies, thus, highlighted the
1
importance of accurately reproducing the inter- Drought (ElectrA), produced by the ARIDE
annual variability in global climate models if future (Assessment of the Regional Impact of Droughts in
long-term droughts affected by climate change are Europe) project, which is capable of displaying on-
to be adequately predicted. Fleig et al. (2006) screen images of streamflow conditions over
carried out similar research using daily flow time Western and Central Europe for several drought
series data from 16 selected river basins worldwide. events that occurred within the last 30 years
The above studies were conducted using observed (Zaidman et al. 2000).
data, which is useful in examining geographical Other relevant mapping projects are carried out
differences in the statistical nature of droughts but primarily by a few international organizations/
are constrained by limited observation points. projects, although they are not normally focusing
Sheffield and Wood (2007a) used a monthly on droughts per se. UNEP’s World Atlas of
drought index based on simulated soil moisture Desertification shows the global extent and severity
data for the period 1950-2000 to identify the of desertification (Middleton 1997; UNEP 1992). It
locations most prone to short, medium and long- includes several maps derived from the Global
term droughts and to examine severe past drought Assessment of Human-induced Soil Degradation
events on a regional basis. Dai et al. (2004) have (GLASOD - described elsewhere in this paper;
developed a global monthly dataset of the well- ISRIC 1990) as well as other maps and information
known Palmer Drought Severity Index (PDSI) for related to global climate and vegetation such as
1870-2002 using historical data on precipitation and Global Humidity Index, Mean Annual Precipitation,
temperature on a 2.5o x 2.5o grid and established Change in Humidity Index and Mean Annual
that, globally, very dry areas had more than Temperature (between two 30-year time periods),
doubled since the 1970s. Below et al. (2007) have and Mean Annual Potential Evapotranspiration
undertaken a comprehensive review of 807 drought (PET). Also mapped are the socioeconomics
and 76 famine entries from 1900 to 2004 in the (population estimates, impact on migration and
EM-DAT database (Emergency Events Database: refugees) of the areas at risk of desertification. The
www.emdat.be/) and revised estimates of global study on drylands, people, and ecosystem goods
drought-related deaths. Dettinger and Diaz (2000) and services by the World Resources Institute
have used monthly streamflow series from 1,345 (WRI) examines, through the use of maps, the
sites around the world to characterize and map world’s drylands from a human livelihoods
geographic differences in the seasonality and perspective and how these livelihoods are
annual variability of streamflow. The Climate Impact integrated with dryland ecosystem goods and
on Agriculture (CLIMPAG) project of the FAO has services (White and Nackoney 2003).
carried out an analysis of rainfall variability and The UNDP’s Bureau for Crisis Prevention and
drought for the period 1961-2002 and presented Recovery (BCPR) developed an individual Disaster
results through time series maps, which are Risk Index (DRI) for four types of natural disasters
available at www.fao.org/nr/climpag/nri/ (earthquakes, tropical cyclones, floods and
nrilist_en.asp. Regionally, Lloyd-Hughes and droughts) as well as a multi-hazard DRI (UNDP
Saunders (2002) have developed a grid-based (0.5o 2004). The risk was expressed in terms of the
resolution) climatology for Europe, which provides number of people killed and was viewed as a
- for a given location or region in Europe - the time function of physical exposure and vulnerability. The
series of drought strength, the number, the mean exercise was based on global data from 1980 to
duration and the maximum duration of droughts of 2000. Global maps depicting physical exposure
a given intensity and the trend in drought (people exposed per year) and relative vulnerability
incidence in the twentieth century, based on the to each type of disaster (people killed per million
well-known Standardized Precipitation Index (SPI). exposed per year) were also produced. However,
Another regional example is the software, the BCPR acknowledges that the drought DRI that
Electronic Atlas  for Visualisation of Streamflow was produced may not necessarily represent
2
actual drought risk given the uncertainties response of terrestrial ecosystem processes to
associated with the risk model itself as well as the climate change using dynamic global vegetation
indirect association of death with drought. models. The recent ‘Africa: Atlas of Our Changing
Similarly, The Natural Disaster Hotspots Environment’ (UNEP 2008) and the ‘Impacts of
project of the World Bank has assessed the global Europe’s Changing Climate: An indicator-based
risks of two disaster-related outcomes - mortality assessment’ (European Environment Agency 2004)
and economic losses - on a 2.5' x 2.5' grid by are examples of regional climate change mapping
considering physical exposure and historical loss projects. A few other projects by IIASA (Fischer et
rates (Dilley et al. 2005). This project also produced al. 2002a, 2002b), START (Adejuwon 2006;
global maps of disaster-related mortality risk, risk Snidvongs 2006), ATEAM (Schröter et al. 2004)
of total economic losses and risk of economic and ILRI (Thornton et al. 2002) have focused
losses expressed as a proportion of the GDP (per specifically on climate change impacts on
grid cell) for six major natural hazards: agriculture and did not explicitly highlight droughts.
earthquakes, volcanoes, landslides, floods, drought, Studies to predict future development of drought
and cyclones as well as for all hazards combined. and changes in the occurrence and intensity of
The Americas program (led by the Institute of drought have been carried out by Sheffield (2008),
Environmental Studies (IDEA) of the National Sheffield and Wood (2007b) and Wood et al. (2003)
University of Colombia, for the Inter–American using climate models and future projections of soil
Development Bank (IDB)) (Cardona 2007), and the moisture. Burke et al. (2006) found that at present
European Environment Agency (2003) have climate conditions, on average, 20% of the land
undertaken two regional mapping projects related to surface is in drought at any given time while the
various aspects of disaster risk for Latin America proportion of land surface in extreme drought is
and Europe, respectively. predicted to increase from 1% at present to 30%
The Global Water System Project (GWSP) by the end of the twenty-first century. A few
examines global water assessment indicators with regional studies spell out the impact of climate
links to poverty and food security, such as the change on European droughts with accompanying
Water Wealth Index (WWI) (Sullivan et al. 2006). maps (Kilsby 2001; Lehner et al. 2001; Lehner et
Global Rapid Indicator Mapping System for Water al. 2006).
Cycle and Water Resource Assessment (Global– The above review suggests that while the
RIMS), a web-based integrated monitoring tool research and mapping of disaster risks, water
developed by the Water Systems Analysis Group scarcity, climate change and related subjects has
of the University of New Hampshire (with 130 been significant, there has been little, if any,
global datasets facilitating indicator calculation and attempt to date to comprehensively describe and
mapping) has been used for mapping WWI and map various aspects and impacts of a drought as
other indicators. an individual natural disaster and as a global multi-
Most of the attention in the recent mapping faceted phenomenon. The aim of this study is,
exercises was paid to various social and therefore, to start filling this niche. It is important
environmental impacts of climate change. These to emphasize the word ‘start’, because the number
studies are relevant to understanding and mapping of drought-related characteristics, as well as
global drought patterns and impacts because associated maps, is potentially quite large. This
climate change is likely to exacerbate drought study shall not, therefore, be seen as exhaustive,
severity in many parts of the world. The Atlas of but rather as a starting point for global mapping of
Climate Change by the Stockholm Environment drought patterns. A limited set of maps which is
Institute (Dow 2006) examines possible global designed and analyzed in this study may, with
impacts of climate change including warning subsequent contributions from other research
signals, vulnerable populations, and health impacts. groups, develop into a comprehensive global
Cramer et al. (2001) have studied the global drought indicators’ ‘atlas’ in the future.
3
Data and Methods
Datasets 2004). The monthly gridded precipitation
dataset (CRU TS 2.0) (Mitchell et al. 2004) is
The study used a number of publicly available based on a set of observational databases held
datasets ranging from demographics and at the University of East Anglia, UK.
socioeconomics to natural resources and climate. Annual runoff (mm/year per grid cell) – a 0.5o
The datasets are briefly described below. resolution global gridded dataset of long-term
Gridded Population of the World, version 3 average (1950-2000) annual runoff per grid cell
(GPWv3) - produced by the Center for computed by Water Balance Model (WBM)
International Earth Science Information Network (Vorosmarty et al. 1998) using CRU TS 2.0
(CIESIN) of the Earth Institute at Columbia (Mitchell et al. 2004) as precipitation input.
University, USA (sedac.ciesin.columbia.edu/gpw/).
This dataset depicts population (in absolute Annual river discharge (blended, km3/year per
numbers) and density estimates by 2.5 arc grid cell) - computed as long-term average
minute geo-referenced quadrilateral grids for 232 (1950-2000) flow accumulated runoff along a
countries. Data is available for every fifth year from 0.5o resolution digital river network (STN-30p)
1995 to 2015. The product is also available in developed at the UNH. Blended river flow
other resolutions - 0.25, 0.5 and 1 degree. represents a composite of observed (from
Population estimates for each grid cell are based Global Runoff Data Centre (GRDC) data
on national or sub-national population data for a archive) and (WBM) modeled river flow.
range of reference years. The reference years Dams, lakes and reservoirs database contains
span the period from 1979 to 2003, depending on both vector as well as raster (0.5o) GIS data on
data availability for each country. Population dams, lakes and reservoirs of the world. The
density estimates and population (in absolute dams and reservoirs point dataset, which is
numbers) for 2000 at 0.5 degree resolution were part of this database, is a global data bank of
used in this study. 668 large impoundments with attributes such
World Water Development Report II (WWDRII) as geographic location, dam name and type,
database (wwdrii.sr.unh.edu/) is part of the reservoir capacity and so forth. This dataset
compendium of databases developed by the Water was used in the study along with other dam
Systems Analysis Group of the University of New datasets held by AQUASTAT (FAO’s global
Hampshire (UNH), USA, describing the current information system on water and agriculture)
status of global water resources and associated and the International Commission on Large
human interactions and pressures. The themes Dams (ICOLD) (see the sections on
covered include major water balance components AQUASTAT and the World Register of Dams
(precipitation, evapotranspiration, runoff, etc.), below).
dams, lakes and reservoirs, population, major UNH Monthly Runoff and River Discharge Time
wetlands and floodplains, irrigated lands and Series grids - represent the output (runoff) of the
irrigation water withdrawals, water pollution above WBM (spatial resolution - 0.5o) along with
indicators, digitized river networks and climate blended river flow – both for a standard period of
moisture indices. Four datasets from this database 1901 to 2000 (100 years).
were used in the present study: World Development Indicators (WDI)
Annual precipitation (mm/year per grid cell) - a (web.worldbank.org/) is the World Bank’s premier
global gridded dataset (0.5o spatial resolution) annual compilation of data about development. It
of long-term average (1950-2000) annual includes some 800 indicators (in 2008) on
precipitation per grid cell computed from economic output, welfare, status of the environment
monthly precipitation fields (Mitchell et al. and the quality of governance - for some 209 (in
4
2008) countries in the world. WDI online data on maps such as main soils of the world, soil production
rural access and access to improved water index, soil drainage class, effective soil depth, etc. The
sources, for the most recent available year, were global map of effective soil depth, which has a spatial
used for construction of infrastructure maps in this resolution of 0.5o, was used in this study.
study. AQUASTAT (www.fao.org/nr/water/aquastat/data/
Global Assessment of Human-induced Soil query/index.html) is a global database on water and
Degradation (GLASOD) (www.isric.org/UK/ agriculture developed by the Land and Water
A b o u t + I S R I C / P r o j e c t s / T r a c k + R e c o r d / Division of the FAO which holds data on the global
GLASOD.htm) project of the International Soil status of land and water resources on a country
Reference and Information Centre (ISRIC), basis. This study used country datasets on annual
Wageningen, the Netherlands (commissioned by renewable water quantities, annual water
the UNEP), produced a global map of human- withdrawals and dam capacities.
induced soil degradation in 1990 at an average ProdSTAT (faostat.fao.org/site/526/default.
scale of 1:10,000,000. The initial GLASOD map aspx) maintained by the FAO contains detailed
had loosely defined physiographic units agricultural production data, area/stock and yield
(polygons), and the degradation status (type, data on a country basis starting from 1961.
extent, degree, rate and cause) for individual Cropped area data for seven types of crops were
polygons was mapped based on qualitative expert used in calculating the Socioeconomic Drought
judgment of a large number of soil scientists Vulnerability index in this study.
throughout the world. GLASOD has paved the World Register of Dams compiled by the
way for more detailed assessments of soil International Commission on Large Dams (ICOLD)
degradation, such as the Assessment of Soil (www.icold-cigb.net/) is a reference to large dams
Degradation in South and Southeast Asia (height - greater than or equal to 15 meters (m)) of
(ASSOD) (1:5,000,000). The results of these the world providing information such as dam type,
assessments have been used to update the height, capacity and purpose. The 1998 version of
GLASOD regional coverage, and additional the World Register of Dams, which contains data
updates are also in progress. The data are on 25,410 large dams of the world, was used in
available for download in digital format from the mapping storage related indices in this study with
ISRIC website (ISRIC 1990). supplemental information from AQUASTAT and the
Global Land Use Dataset (www.sage.wisc.edu/ Dams and Reservoirs dataset of the WWDRII
iamdata/) held by the Center for Sustainability and database.
the Global Environment (SAGE), University of In addition to the above, other data sources on
Wisconsin-Madison, USA, describe the geographic freshwater resources of the world such as the
patterns of the world’s croplands, grazing lands, Earthtrends Searchable Database
urban areas, and natural vegetation. Data are (earthtrends.wri.org/index.php) maintained by the
available in both tabular format (for countries, World Resources Institute (WRI), USA, the World’s
states, etc., of the world) as well as in map form. Water database (www.worldwater.org/) maintained
The 0.5o resolution grid dataset illustrating global by the Pacific Institute, USA, the State of Water
cropland area in 1992 (as fraction of grid cell) database (www.wepa-db.net/policies/top.htm)
(Ramankutty and Foley 1998) was used in mapping maintained by the Water Environment Partnership in
the Agricultural Water Crowding Indices. Asia (WEPA), Japan, as well as Malik et al. (2000)
FAO Digital Media Series (www.fao.org/ and White (2005) were also used to inform mapping
landandwater/lwdms.stm) includes global thematic of storage-related indices.
5
Drought Characteristics and Indices drought studies. Some others are existing indices,
which although designed for a different purpose
This section briefly describes the indices and originally, carry useful drought-related information if
characteristics presented and mapped in this study used either as is or with certain modifications.
– primarily focusing on the origin of indices and Long-term Mean Annual Precipitation (MAP)
rationale for mapping. Some of them are drought- (Figure 1(a)), its Coefficient of Variation (CV)
related indices, which were either used locally (Figure 1(b)), and Probability (%) of Annual
rather than globally, or used out of the context with Precipitation in any year being less than 75% of its
(a)
(b)
(c)
FIGURE 1. (a) Global distribution of long-term Mean Annual Precipitation, (b) its Coefficient of Variation, and
(c) Probability (%) of annual precipitation in any year being less than 75% of its long-term mean.
6
Long-term Mean (Figure 1(c)) were calculated and found to be broadly similar and, hence, only the
mapped globally on a 0.5o grid. Since drought is first is presented here. The map was produced
generally defined in relation to a long-term average using population data from CIESIN
condition, it is worth identifying a global pattern of (sedac.ciesin.columbia.edu/gpw) and annual river
such conditions - in this case MAP (Figure 1(a)) discharge data from UNH (wwdrii.sr.unh.edu/
and its inter-annual variability (Figure 1(b)). The download.html).
probability of annual precipitation in any year being A more ‘agriculture-focused’ index is
less than 75% of MAP (Figure 1(c)) can point to Agricultural Water Crowding (Sullivan et al. 2006).
regional differences in the frequency of occurrence It was designed to measure population numbers
of annual droughts and links the pattern of these per one volumetric unit of precipitation falling on
droughts with MAP and CV. A threshold of 75% of croplands, but has not been applied globally. The
MAP, while somewhat arbitrary, is often accepted original index measured crowding with respect to
as an identifier of a meteorological drought. These precipitation only. In this study, we made use of
maps were produced using annual precipitation two variations of this index (AW1 and AW2) by
data from the University of East Anglia considering both MAP and mean annual river
(www.cru.uea.ac.uk/cru/data/hrg.htm). discharge MAR:
Per capita Mean Annual River Discharge
(Figure 2) reflects population pressure on river flow P
within a 0.5° grid cell, which is exacerbated in times AW 1 =   (1)
MAP ⋅ CF
of drought. A similar indicator, per capita Mean
Annual Surface Runoff, only considers runoff
generated internally within a grid cell. Since many P
AW 2 = (2)
rivers in the world are transboundary in nature, the MAR ⋅ CF
second indicator pictures a hypothetical situation
when every country has to rely on its own runoff where: P = population (number of people) per grid
alone. The pattern of both indicators was, however, cell; MAP = mean annual precipitation per grid cell
FIGURE 2. Per capita Mean Annual River Discharge by 0.50 grid cell.
7
(km3); MAR = mean annual river discharge per grid from the two distinct sources. Agricultural water
cell (km3); and CF = crop area as a fraction of cell crowding maps were produced using population
area. Water (either precipitation or river discharge) data from CIESIN (sedac.ciesin.columbia.edu/gpw),
available within a 0.5o grid cell may be split into MAR and MAP data from UNH (wwdrii.sr.unh.edu/
agricultural water and non–agricultural water in download.html), and percentage cropped area from
proportion to the cropped and non-cropped areas in SAGE (www.sage.wisc.edu/iamdata/).
that unit. Agricultural Water Crowding is a measure An Infrastructure Vulnerability Index, similar to
of the number of people who have to share the one developed by the World Travel and Tourism
agricultural water in a grid cell. By mapping it Council (www.wttc.org/eng/Tourism_News/
globally (Figures 3(a) and 4), it is possible to Press_Re leases /Press_Re leases_2004 /
identify “agricultural water-stressed” areas, which New_Statistics_launched/), was used to reflect
are becoming even more stressed in times of a adaptive capacity of a country to a drought. Similar
drought (Figure 3(b) - produced with an assumption indices, although much more complex and
that an average drought (mean precipitation minus inclusive of a large number of indicators (10 to 50),
mean precipitation deficit) occurs simultaneously are well known (e.g., O’Brien et al. 2004;
over the globe). The variations of the same index www.sopac.org/tiki-index.php?page=EVI). The index
relate to different aspects of availability of water used in this study only includes two proxy
(a)
(b)
FIGURE 3. Agricultural Water Crowding (population sharing one cubic kilometer of precipitation falling on croplands
within 0.50 grid cell) with respect to (a) mean annual precipitation, and (b) under drought conditions.
8
FIGURE 4. Agricultural Water Crowding with respect to mean annual river discharge (population sharing
one cubic kilometer of river water available for croplands within 0.50 grid cell).
indicators - the WB’s Rural Access Index (RA) and minutes) of an all-season road as a proportion of the
the percentage of population with access to an total rural population; and Idw = Percentage of
improved water source (IDW). The data for both people having access to (able to obtain at least 20
components are available for a large number of liters per person per day from a source within 1 km
countries (web.worldbank.org/). Both components of a dwelling) an improved water source (household
determine the adaptive capacity of agriculture and connection, public standpipe, protected well or
rural communities to current climate variability and spring, etc.). The index has a score of 0-100 with
associated droughts. The composite Infrastructure 100 implying maximum vulnerability (Figure 5).
Vulnerability Index (IFI) was constructed in a similar The Biophysical Vulnerability Index of O’ Brien
manner to UNDP’s Human Development Index et al. (2004) applied to India at the resolution of
(UNDP 2006), in which the values of each individual states, consists of three sub-indices:
component indicator were normalized to the range Depth of Soil Cover Index (DS), Soil Degradation
of values in the dataset: Severity Index (SD) and Groundwater Scarcity
Index (GWS). A similar Biophysical Vulnerability
= RA + IDW
IFI (3) Index (BVI) (Figure 6) was constructed in this
2 study by adding a fourth dimension: Surface Water
(Runoff) Scarcity Index (SWS).
= − Raactual − Ra min
RA 100 ⋅ 100 (4)
Ra max− Ra min
= DS + SD + GWS + SWS
BVI (6)
4
= − Idwactual − Idw min
IDW 100 ⋅ 100 (5)
Idw max− Idw min
= − Dsactual − Ds min
DS 100 ⋅ 100 (7)
Ds max− Ds min
where: Ra = World Bank’s Rural Access Index -
percentage of rural people who live within 2 = Sdactual − Sd min
SD ⋅ 100 (8)
kilometers (km) (typically equivalent to a walk of 20 Sd max− Sd min
9
FIGURE 5. Infrastructure Vulnerability Index based on the percentage of people having access to an improved
water source and general accessibility of rural areas through the road network.
FIGURE 6. Biophysical Vulnerability Index based on mean annual surface runoff, mean annual groundwater recharge,
soil depth and soil degradation severity within 0.50 grid cell.
Degradation Severity; Gr = Annual Groundwater
Gractual Gr min
GWS 100 100 (9) Recharge; and MAS = Mean Annual Surface
Gr max Gr min
Runoff. The final composite index was mapped at
0.5o
MAS  resolution. Data for proxy variables Ds, Sd
actual MAS min
SWS 100 100 (10) and MAS are available at that resolution (e.g., Ds
MAS max MAS min
– from FAO Digital Media Series (www.fao.org/
where: DS = Depth of Soil Cover Index; SD = Soil landandwater/lwdms.stm); Sd – from ISRIC
Degradation Severity Index; GWS = Groundwater (www.isric.org); and MAS from UNH
(wwdrii.sr.unh.edu/download.html)). Since high
Scarcity Index; SWS = Surface Runoff Scarcity resolution global groundwater data are not available
Index; Ds = Depth of Soil Cover; Sd = Soil in the public domain, country-scale groundwater
10
recharge data from WRI (earthtrends.wri.org/ Socioeconomic Drought Vulnerability Index. IDI and
index.php) were converted into 0.5o resolution grid EDI use World Bank Indicators (web.worldbank.
data - for mapping purposes. The index may be org/): percentage contribution from agriculture to
seen as a measure of sensitivity of agriculture to national GDP (Av), and percentage employed in
droughts. It has a score of 0-100 with 100 implying agriculture (% of total employment) (Ea),
maximum vulnerability. Areas with higher respectively, as proxy variables. The proxy variable
biophysical vulnerability are those which are most in CDI is the Crops Diversity Index (Ci) suggested
vulnerable to agricultural drought whenever by Jülich (2006). A weight of 0.4 is assigned to
meteorological drought occurs. each of IDI and EDI, while a weight of 0.2 is
The Socioeconomic Drought Vulnerability Index assigned to CDI in the composite index. The latter
(SDI) (Figure 7) measures the vulnerability of is done to emphasize that the importance of crops
individual countries to socioeconomic drought. It is diversity in a country depends on how large a
formulated on the consideration that higher GDP contribution is made by the agricultural sector to
contributions from non-agricultural sectors, lower the country’s economy. SDI has a score of 0-100
percentage employment in the agricultural sector with 100 implying maximum vulnerability.
and higher crops diversity will collectively lower a
country’s chances of developing socioeconomic
SDI = 0.4IDI + 0.4EDI + 0.2CDI (11)
drought when meteorological drought occurs. Three
sub-indices, namely, the Income Diversity Index
Avactual Avmin
(IDI), Employment Diversity Index (EDI) and the IDI 100 (12)
Crop Range Index (CDI), make up the composite Avmax Avmin
FIGURE 7. Socioeconomic Drought Vulnerability Index based on the crop diversity of individual countries and their
dependence on agriculture for income and employment generation.
11
Eaactual Eamin (13) - Bernardino and Corte Real 2004), but not
EDI 100
Eamax Eamin globally. Some authors attempted to combine
these measures to derive composite Drought Risk
Ciactual Ci min Indices (Zongxue et al. 1998; Loucks 1997;
CDI 100 (14)
Ci max Ci McMahon et al. 2006). This study attempted to
min
map several such indices. Two maps of drought
2 risks are presented here to avoid showing too
Ci P (15)
many maps (which are often similar). Drought Risk
Index (DRI) is calculated as:
where: IDI = Income Diversity Index;
EDI = Employment Diversity Index; CDI = Crop = 1 − REL + 2RV
DRI (16)
Range Index; Av =  percentage contribution from 3 3
agriculture to national GDP; Ea =  percentage
employed in agriculture (% of total employment); Ns V
where: REL = ; RV = ;
N MMP
Ci =  Crops diversity Index suggested by Jülich
(2006); and P = Fractional cropped area out of N
total cropped area for each type of crop. Cropped ∑ Im ax
area data for seven types of crops (cereals, V = i=1 ; V = Vulnerability; Im ax  =
vegetables, fruits and nuts, oil crops, roots and ND
tubers, pulses, and fibers) in FAO’s ProdSTAT Maximum drought intensity (maximum individual
database (faostat.fao.org/site/526/default.aspx) were deficit per time step) in each drought run; ND =
used in calculating Ci . Smaller Ci values indicate
Number of drought runs; RV =  Relative Vulnerability;
higher crops diversity.
A number of indices are proposed in literature, MMP = Mean Monthly Precipitation or Mean
which measure the performance of water resources Monthly River Discharge; Ns = number of intervals
systems in terms of reliability, resilience and (months) that the target demand (Mean Monthly
vulnerability of water resources (e.g., Hashimoto et Precipitation or Mean Monthly River Discharge) was
al. 1982). Reliability in essence is a probability that fully met; and N =  total number of intervals
monthly precipitation (or discharge) is larger than (months). The DRI was mapped at 0.5o resolution for
its long-term monthly mean value. Vulnerability, in both monthly precipitation (Figure 8) and monthly
this context, refers to the likely magnitude of a river discharge (Figure 9). Regardless of its seemingly
failure (maximum drought intensity) if one occurs. complex formulation, the DRI in essence is an
Relative vulnerability is the vulnerability divided by integrated index which shows the combined drought
the expected threshold value (Hashimoto et al. risk at any given location in terms of precipitation/river
1982; McMahon et al. 2006): in this study - long- discharge reliability and vulnerability. It ranges from 0-
term monthly mean precipitation or discharge. 1. Higher DRI values imply that the area has less
Resilience may be interpreted as a measure of reliable precipitation/discharge. The datasets used for
how quickly a system is likely to recover from calculating drought risk indices are monthly
failure once failure has occurred. Vulnerability and precipitation and monthly river discharge from the
Resilience are, hence, effectively complementary. University of East Anglia (www.cru.uea.ac.uk/cru/data/
Some of these measures were mapped before for hrg.htm) and UNH (www.grdc.sr.unh.edu/html/Data/
certain geographical regions (e.g., parts of Europe index.html), respectively.
12
FIGURE 8. Drought Risk Index with respect to Monthly Precipitation based on the frequency of meteorological (precipitation)
drought occurrence and drought intensity (deficit below long-term mean).
FIGURE 9. Drought Risk Index with respect to Monthly River Discharge based on the frequency of hydrological (river
discharge) drought occurrence and drought intensity (deficit below long-term mean).
Drought Duration is another important actual duration can. Figure 10 shows the
characteristic which varies globally very distribution of the mean drought run duration
significantly. It is possible to distinguish between based on monthly river discharge (sum of
the actual duration of a drought (which can last durations of all indentified drought runs divided by
more than a year – a drought ‘run’) and the number of runs). This map was produced using
duration of an annual drought (i.e., how long can 0.5° resolution monthly river discharge grids from
a drought last in a single year). The latter case UNH (www.grdc.sr.unh.edu/html/Data/index.html).
refers to a number of dry months within a year The distribution of annual drought duration is
and cannot be more than 12 months, while the broadly similar.
13
A few indices were mapped, which aim to
capture the adequacy of water storage capacity in SC
SCI = (17)
a country or other spatial unit to meet its annual ARW
water withdrawals in the event of a drought.
Storage Capacity (SC) as a proportion of Total where: SCI = Storage Capacity Index;
Annual Renewable Freshwater Resources (ARW) SC = Storage Capacity; and ARW =  Total Annual
within a country (Figure 11) is an indicator of the Renewable Freshwater Resources within a country.
extent of exploitation of national water resources in The Storage–Drought Duration (length) Index
a country. Total annual renewable freshwater (SLI) is the ratio between the duration (in months)
resources include both surface water as well as that the storage capacity in a country (SC) is able
groundwater. White (2005) has calculated this ratio to satisfy national water needs (based on monthly
for a few countries with reservoir storage in excess surface water withdrawals (SW)), and annual
of half the total annual freshwater resources. hydrological drought duration (DDM) (in months),
FIGURE 10. Mean Drought Run Duration based on monthly river discharge (sum of durations of all identified
drought runs divided by the number of runs).
FIGURE 11. Storage as a Proportion of a Country’s Total Annual Renewable Freshwater Resources.
14
calculated relative to an arbitrary drought threshold Monthly river discharge grids (0.5o resolution)
(long-term mean monthly river discharge). The from UNH (www.grdc.sr.unh.edu/html/Data/
Storage–Drought Deficit Index (SDI) is an indicator index.html) were used in calculating both
of how much of the annual (hydrological) drought indices. Only grid cells with MAR > 0.01 MCM
deficit (MAD) (relative to long-term mean) is were considered. The annual drought duration,
satisfied by the existing storage capacity (SC) in and the annual drought deficit were initially
calculated at a 0.5o
a county.  resolution and averaged
across each country, while storage capacity,
SC MAR and water withdrawal data were available on
SLI = SW (18) a country scale. Finally, SLI and SDI were
DDM mapped at a country scale (Figures 12 and 13).
For mapping the three storage related indices,
SC Storage Capacity data were obtained from the
SDI = (19)
MAD World Register of Dams, ICOLD (www.icold-
cigb.net/), AQUASTAT (www.fao.org/nr/water/
aquastat/data/query/index.html) and the dams,
where: SLI =  Storage–Drought Duration Index; lakes and reservoirs database of UNH
SDI =  Storage–Drought Deficit Index; SC = (wwdrii.sr.unh.edu/download.html). Total Annual
Storage Capacity;  SW = monthly surface water Renewable Freshwater Resources data were
withdrawals;  DDM = annual hydrological drought obtained mainly from AQUASTAT, Earthtrends
duration (months); and  MAD = annual Searchable Database of WRI (earthtrends.wri.org/
(hydrological) drought deficit relative to long-term index.php) and World’s Water database of the
mean. Pacific Institute (www.worldwater.org/).
FIGURE 12. Storage–Drought Duration (Length) Index - ratio between i) the duration (in months) that the storage
capacity in a country is able to satisfy national water needs, and ii) annual hydrological drought duration, calculated
relative to the long-term mean monthly river discharge.
15
FIGURE 13. Storage–Drought Deficit Index (how much of the long-term annual hydrological drought deficit is satisfied by the
existing storage capacity in a country).
Discussion
The maps presented in Figure 1 effectively describe areas which are naturally arid or semi-arid (e.g.,
the natural availability of water resources in any receiving less rainfall over the long term) also tend
specific region. This availability certainly determines to have higher CV of mean annual precipitation
whether droughts are seen as a severe problem or and, consequently, higher probability of drought
not. In arid areas, there may even be a lack of occurrence - at least in the case of an ‘annual’
distinction between drought and aridity (Smakhtin drought. This partially explains the occasional
and Schipper 2008). Aridity is a measure of how confusion between drought and aridity and also
dry/wet a region is on average over the long term; suggests, that management measures taken in arid
it is a permanent climatic characteristic of an areas to alleviate unreliable water supplies, whether
area. Drought is a deviation from this long-term in a drought or not, are similar.
mean (which is different in different physiographic More insights may be inferred if population is
areas). Thus, droughts come and go, but aridity in added to the picture. Per capita availability of mean
an area remains. In arid areas, however, the intra- annual river discharge (Figure 2) allows areas of
annual variability of precipitation is generally higher both ‘climate-driven’ and ‘population- driven’ water
than in humid areas. Figure 1 illustrates this point. scarcity to be identified (Falkenmark et al. 2007).
Figures 1(a) and 1(b) show the distribution of mean For example, Afghanistan, Iran and Pakistan,
annual precipitation on a global scale and the which together occupy a comparable land area
distribution of the coefficient of variation (CV) of (3,193,340 square kilometres (km2)) with India
mean annual precipitation, respectively. Figure 1(c) (3,287,260 km2), collectively generate only some
shows the probability that annual rainfall in an area 20% of India’s MAR of 1,858 cubic kilometres
will fall below the threshold of 75% of the long-term (km3). However, India on one hand and the other
mean annual precipitation. The latter threshold is three countries (on average) on another, have close
used here as an arbitrary limit, below which a year per capita MAR (1,613 cubic meters (m3) and 1,300
can be considered a ‘drought year’. It appears, that m3, respectively). Due to India’s higher population
16
density, this observation may be interpreted as km3 of water. Figure 4 presents a completely
‘population-driven’ water scarcity in India as different picture with more than half of the cropped
opposed to ‘climate-driven’ water scarcity in the areas of the world under the same condition (the
other three countries. Southeastern China, Thailand only exception being major river corridors). A closer
and East Africa are other areas more likely to be look at South and Southeast Asia, or the Murray–
experiencing population–driven water scarcity, Darling Basin in Australia, suggests that if mean
although they are some of the wettest parts of the precipitation is considered a measure to calculate
world (Figure 1). Australia, Southwest and Central water crowding, then most of the cropped areas
Asia, North Africa, northern China, Mongolia, are not under agricultural water stress (Agricultural
southern Africa, western United States, Latin Water Crowding is less than or equal to 600,000
America and southern American countries such as people per 1 km3 of water) (Figure 3(a)). Figure 4,
Argentina and Paraguay are, on the other hand, on the contrary, points to escalated water crowding
more likely to experience “climate–driven” water if river flow is used as a measure. In a drought year
scarcity being in arid or semi-arid environments. At (Figure 3(b)), agricultural water crowding increases,
the same time, almost all of them are also depending on the severity of a drought. In Figure
categorized as having or approaching “demand- 3(b), a drought year in a grid cell is defined as a
driven” scarcity (Comprehensive Assessment of year when precipitation is less than its long-term
Water Management in Agriculture 2007). In drought mean value by the long-term mean annual
years, per capita water availability drops. The precipitation deficit. However, agricultural water
overall distribution pattern remains the same, but crowding levels in Figure 3(b) are still much lower
regions with limited per capita flow availability than those in Figure 4, suggesting that even in a
increases. In the earlier example, if a global drought year precipitation water availability is
drought year is defined as a year when annual river higher than that of river discharge under long-term
discharge is 75% of long-term MAR, then India’s (normal) conditions, which is also true in areas
per capita river discharge drops by 402 m3 while which rely heavily on river water for agriculture.
that of the other three countries (Afghanistan, Iran Therefore, there may be a potential for rainwater
and Pakistan) drops by only 325 m3. In a ‘global’ use in agriculture that can be tapped by enhanced
drought year, the highest per capita water losses rainwater harvesting. This is yet another argument
occur not so much in the driest regions, but rather in support of frequent calls to view rainfall as the
in regions which are not normally water scarce due ultimate source of water (Comprehensive
to climate. Assessment of Water Management in Agriculture
The two maps of agricultural water crowding 2007; Falkenmark et al. 2007), instead of focusing
(Figures 3(a) and 4) illustrate much higher values of only on river flow/groundwater. According to the
crowding with respect to river discharge than with earlier definition of a drought year (which is equally
respect to precipitation. The obvious reason for this applicable to river discharge), the number of people
is that annual precipitation is higher than annual living under Falkenmark’s chronic (agricultural) river
runoff in any part of the world due to various losses water scarcity (1,000,000 people per km3) may
on the ground. According to the Comprehensive reach 3.3 billion, of which over 2 billion would be
Assessment of Water Management in Agriculture in areas with an extreme crowding of 2,000,000
(2007), globally, about 39% of rain contributes to people per km3. While a similar dry year could not
river discharge and groundwater collectively. Only in happen simultaneously over the entire planet, the
a few countries of the Middle East; South, East, estimates above point to the danger of droughts in
and Central Asia; and Northern and Western Africa various parts of the world.
in Figure 3(a), the cropped areas appear to be Infrastructure development of any country
under Chronic Agricultural Water Scarcity determines, amongst others, the level of its
(Falkenmark 1989; FAO 2000), i.e., where water preparedness to drought. The availability of
crowding is greater than 1,000,000 people per 1 improved drinking water and general accessibility of
17
rural areas (where most of the world’s poor reside) Figures 8 and 9 illustrate that the river
through the road network are two important factors discharge Drought Risk Index is higher than the
determining a country’s anti-drought coping precipitation Drought Risk Index throughout the
capacity. The countries most vulnerable to adverse world, except for a few pockets in South America,
societal impacts due to drought are those which Africa and Southeast Asia. This comparison
already have low MAP and high CV (thus having highlights the unreliable and vulnerable nature of
higher probability of occurrence of drought – see river discharge, and further confirms the widely
Figure 1). They often score similarly low in voiced dangers of relying on river water alone. In
infrastructure development terms and have lower general, the arid and semi-arid areas have a higher
institutional capacity to respond effectively or drought risk index than the rest of the world
mitigate the effects of drought. It is evident that the implying frequent drought occurrence and higher
African continent is lagging behind the rest of the drought intensity (deficit below long-term mean)
world (Figure 5) in this context. European countries when drought does occur. Europe is ‘better-off’ in
such as UK, Spain, France and the Netherlands terms of this index and Africa is the worst case.
score the lowest on the Infrastructure Vulnerability Figure 10 shows how average hydrological
Index (higher infrastructure development) while drought duration (run length) varies across the
Ethiopia, Somalia, Chad, Mali and Mozambique, as globe. A large part of Africa, South, Southwest
well as Afghanistan in Asia score the highest. and Central Asia and northern Australia (all arid
According to Figure 6, the arid and semi-arid and semi-arid regions) are more prone to multi-
areas of the world, especially the Sahel, Southern year hydrological droughts. An analysis of annual
Africa, Southwest Asia, parts of China and Latin drought durations (not mapped here), suggested
America show higher biophysical vulnerability. that these areas also experience longer annual
Comparison of Figures 1(a), 6 and 10 illustrates droughts (how long a drought can last in a single
that the above areas are also subject to prolonged year). Long-term droughts coupled with high
droughts and low MAP, which often results in low infrastructural (Figure 5) and socioeconomic
crop yields. (Figure 7) vulnerability contribute to poor soil
Socioeconomic Drought Vulnerability (Figure 7) quality, food shortage, malnutrition, disease,
is generally higher throughout Africa and Asia conflict and famine in Africa. However, large parts
since many African and Asian countries are largely of South and North America and most of Europe
agricultural economies. In contrast, North and appear to be less prone to multi-year hydrological
South America, Australia and Europe display much droughts, while they also have shorter annual
lower socioeconomic drought vulnerability. This is drought durations.
not surprising considering the fact that percentage The Storage–Drought Duration (Length) Index
employment in agricultural endeavors is as high as (Figure 12) indicates the fraction of the annual
93% in Bhutan and 92% in Burkina Faso while it drought duration in any country that its present
is as low as 1% in the United Kingdom and 2% in storage capacity is able to satisfy based on its
the United States. African Countries such as monthly surface water demand. An index value of 1
Guinea-Bissau, Ethiopia and Niger, and Asian implies that the country’s present storage capacity
countries Lao PDR, Afghanistan and Cambodia is satisfactory in comparison to its surface water
score the highest on this index (i.e., most demand and mean annual drought duration. Out of
vulnerable), while Hong Kong, Macau and all the areas having comparatively longer drought run
Singapore score the lowest. Agricultural economies durations (Figure 10), southern Africa, Australia,
are much more vulnerable to adverse societal most of South and Central America and the United
impacts due to meteorological drought. The more States seem to be able to satisfy most of their
complex economies of developed countries needs with the current storage facilities, unlike
insolate the population to fluctuations in agricultural some countries in Central and South Asia, where
productivity due to drought. this index is lower than 0.5. Overall, Africa appears
18
to be more ‘drought-ready’ than South Asia with value of Storage–Drought Deficit Index does not
respect to reservoir storage. The worst cases necessarily mean that a particular country is
include Saudi Arabia, Oman, Madagascar, Somalia, unable to meet its freshwater demands during
Kuwait, Syria, Slovakia, Hungary and Nepal. A look drought. Australia, for example, has enough
at Figure 11, which maps the present storage storage to last twice as long as the annual
capacity as a percentage of total available annual drought duration when compared with its monthly
freshwater resources, reveals that many of the water withdrawal or monthly demand (Figure 12).
countries which score low on this index (especially However, according to Figure 13, its storage
those in Asia) have no apparent hydrological barriers volume is 0.25-0.5 of the annual drought deficit
for increasing storage in the future except perhaps (with respect to long-term mean), which implies
Libya which is already storing 05-0.75% of its that its annual demand is much less than the
annual freshwater resources. annual deficit. Therefore, those countries which
Only a few countries score high on Storage– score high on the Storage–Drought Duration
Drought Deficit Index (Figure 13). They are Egypt, (length) Index can be reasonably assumed to
Morocco, Ghana, Cote-d’Ivoire, Burkina Faso, possess satisfactory storage to meet their
Zambia, Malawi, Zimbabwe, Burundi, South Africa, freshwater demands during drought. On the other
China, Uzbekistan, Kyrghystan, Tajikistan, Iraq, hand, those countries which score high on
Turkey, Azerbaijan, Romania and Spain. They are Storage-Drought Deficit Index are also often the
also the countries “performing satisfactorily” on ones which are more susceptible to river
both storage indices (Figures 12 and 13) while fragmentation and over–exploitation of freshwater
having the highest ratios of storage to total resources (e.g., China, Egypt, South Africa)
available freshwater resources (Figure 11). A low (Revenga et al. 2000).
Conclusions
This study reviewed all previous known attempts to been produced by integrating a number of publicly
approach the issue of drought analysis at the available global datasets.
global scale as well as attempts to map disaster This study should not be seen as exhaustive,
risks, water scarcity, climate change and related but rather as a starting point for global analysis of
subjects. The review showed that there has been drought patterns, impacts and preparedness. The
little, if any, attempt to date to comprehensively limited set of maps designed and analyzed in this
describe and map various aspects and impacts of study may, with subsequent contributions from
a drought as an individual natural disaster and as other research groups, develop with time into a
a global multi-faceted phenomenon. Hence, the comprehensive global drought indicators’ ‘atlas’.
study aimed to start filling this niche by producing There are many possibilities on this avenue. At the
a set of global maps of various drought-related same time, it is critically important to note that the
characteristics. These characteristics reflect various occurrence of a drought and a specific location’s
aspects of drought patterns and impacts ranging vulnerability to drought is the result of a
from global distribution of meteorological and combination of many local factors. This study gives
hydrological drought risks to social vulnerability and a rather general, ‘global’ illustration of various
indices related to water infrastructure. The maps - drought-related factors, and should not be used to
either at a country level or regular grid scale - have make sweeping generalizations at the local scale.
19
The present study used monthly rainfall and (low, moderate, high) can be quantified by
flow data as they are the only globally available combining GIS coverages of individual
hydrological data so far. Impacts and response meteorological and basin parameters (e.g., soil
options for short-term droughts (weeks to months) root zone available holding capacity, land-use type,
and long-term droughts (years to decades) may be etc). Such vulnerability coverage can provide
different. Future research should examine the information on which crops are better in which
differences between short-term (e.g., dry spells) parts of the state/country. Vulnerability indices
and long-term droughts more closely. However, for could be based on damage incurred, population
the former, daily precipitation and flow time series affected, number of droughts relative to land area,
are needed at the global scale – these are etc. It should be possible to map drought
currently not available or are not reliable. vulnerability at smaller administrative subdivisions
Quantifying and indexing vulnerability to within countries. But similarly important is to
droughts represents another challenge and evaluate it at the local level and at the level of
research niche. A number of attempts are made to households, where different indicators are needed.
quantify vulnerability to climate change and natural Drought indicator mapping eventually feeds
disasters (e.g., Downing et al. 2001; into development of a scientific knowledge base for
www.vulnerabilitynet.org; www.eci.ox.ac.uk; operational drought tools such as drought
unfccc.int/files/adaptation/methodologies_for/ monitoring, drought early warning systems, which,
vulnerability_and_adaptation/application/pdf/ in turn, should form part of national drought
vulnerability_indices.pdf; www.fao.org/sd/EIdirect/ preparedness plans. It is also necessary to note,
EIre0049.htm).  Vulnerability indices can help that since droughts are projected to become more
identify and target vulnerable regions or populations, severe, longer or frequent in many parts of the
raise awareness, and form part of a monitoring and world in the future (e.g., Bates et al. 2008),
adaptation strategy. However, vulnerability understanding and quantifying drought patterns and
definitions vary a lot between various sectors and anticipated impacts is becoming a matter of ever-
disciplines. Vulnerability to agricultural drought increasing importance.
References
Adejuwon, J. 2006. Food Security, Climate Variability and Climate Change in Sub Saharan West Africa. AIACC
Final Reports: Project No. AF 23. Washington, DC, USA: The International START Secretariat.
Bates, B.C.; Kundzewicz, Z. W.; Wu, S.; Palutikof, J. P. (eds.) 2008. Climate Change and Water: Technical Paper
of the Intergovernmental Panel on Climate Change. Geneva: IPCC Secretariat. 210 pp.
BCPR (UNDP Bureau for Crisis Prevention and Recovery). 2004. Reducing Disaster Risk: A Challenge for
Development.
Below, R.; Grover-Kopec, E.; Dilley, M. 2007. Documenting Drought-Related Disasters. The Journal of
Environment & Development 16(3): 328-344.
Bernardino, M. C.; Corte Real, J. 2004. A Drought Risk Assessment for Europe. Geophysical Research Abstracts
6: 03742.
Burke, E. J.; Brown, S. J.; Christidis, N. 2006. Modeling the Recent Evolution of Global Drought and Projections
for the Twenty-First Century with the Hadley Centre Climate Model. Journal of Hydrometeorology 7: 1113–
1125.
20
Cardona, O. D. 2007. A System of Indicators for Disaster Risk Management in the Americas. Proceedings of
Conference on Globalization, Diversity, and Inequality in Latin America: The Challenges, Opportunities, and
Dangers. Pittsburgh: University of Pittsburgh.
Comprehensive Assessment of Water Management in Agriculture. 2007. Water for food, water for life: A
comprehensive assessment of water management in agriculture. London: Earthscan, and Colombo:
International Water Management Institute.
Cramer, W.; Bondeau, A.; Woodward, F. I.; Prentice, I. C.; Betts, R. A.; Brovkin, V.; Cox, P. M.; Fisher, V.; Foley, J.
A.; Friend, A. D.; Kucharik, C.; Lomas, M. R.; Ramankutty, N.; Sitch, S.; Smith, B.; White, A.; Young–Molling,
C. 2001. Global response of terrestrial ecosystem structure and function to CO2 and climate change: results
from six dynamic global vegetation models. Global Change Biology 7: 357–373.
Dai, A.; Trenberth, K. E.; Qian, T. 2004. A Global Dataset of Palmer Drought Severity Index for 1870–2002:
Relationship with Soil Moisture and Effects of Surface Warming. Journal of Hydrometeorology 5: 1117-1130.
Dettinger, M. D.; Diaz, H. F. 2000. Global Characteristics of Stream Flow Seasonality and Variability. Journal of
Hydrometeorology 1(3): 289-310.
Dilley, M.; Chen, R. S.; Deichmann, U.; Lerner-Lam, A.; Arnold, M. 2005. Natural disaster hotspots: A global risk
analysis. Washington, DC: World Bank, Hazard Management Unit.
Dow, K.; Downing, T. E. 2006. The Atlas of Climate Change. Mapping the World’s Greatest Challenge. London:
Earthscan and Stockholm: Stockholm Environment Institute.
Downing, T. E.; Butterfield, R.; Cohen, S.; Huq, S.; Moss, R.; Rahman, A.; Sokona, Y.; Stephen, L. 2001.
Vulnerability Indices: Climate Change Impacts and Adaptation. Policy Series 3: United Nations Environment
Programme.
European Environment Agency. 2003. Mapping the impacts of recent natural disasters and technological
accidents in Europe. Copenhagen, Denmark: European Environment Agency.
European Environment Agency. 2004. Impacts of Europe’s Changing Climate: An indicator-based assessment.
Luxembourg: Office for Official Publications of the European Communities.
Falkenmark, M. 1989. The Massive Water Scarcity Now Threatening Africa – Why isn’t It Being Addressed?
Ambio 18: 112-118.
Falkenmark, M.; Berntell, A.; Jägerskog, A.; Lundqvist, J.; Matz, M.; Tropp, H. 2007. On the Verge of a New Water
Scarcity: A Call for Good Governance and Human Ingenuity. SIWI Policy Brief. Sweden: Stockhom
International Water Institute.
FAO (Food and Agriculture Organization of the United Nations). 2000. New Dimensions in Water Security; Water,
Society and Ecosystem Services in the 21 st Century. Rome: Food and Agriculture Organization of the United
Nations.
Fischer, G.; Shah, M.; van Velthuizen, H. 2002a. Climate Change and Agricultural Vulnerability. Special report
to the World Summit on Sustainable Development, Johannesburg 2002. Laxenburg, Austria: International
Institute for Applied Systems Analysis (IIASA).
Fischer, G., van Velthuizen, H.; Shah, M.; Nachtergaele, F. 2002b. Global Agro-ecological Assessment for
Agriculture in the 21st Century: Methodology and Results. RR-02-02. Laxenburg, Austria: International Institute
for Applied Systems Analysis (IIASA) and Rome, Italy: Food and Agriculture Organization of the United
Nations.
Fleig, A. K.; Tallaksen, L. M.; Hisdal, H.; Demuth, S. 2006. A global evaluation of streamflow drought
characteristics. Hydrology and Earth System Sciences 10: 535-552.
Hashimoto, T.; Stedinger, J. R.; Loucks, D. P. 1982. Reliability, Resiliency, and Vulnerability Criteria for Water
Resource System Performance Evaluation. Water Resources Research 18(1): 14-20.
21
ISRIC (International Soil Reference and Information Centre). 1990. Global Assessment of Human-induced Soil
Degradation (GLASOD) Data Set. Available at www.isric.org/UK/About+ISRIC/Projects/Track+Record/
GLASOD.htm (accessed in December 2008).
Jülich, S. 2006. Drought risk indicators for assessing rural households. Presentation made at Global Water
Hotspots: Water-related social vulnerabilities and resilience-building, Summer Academy for Social
Vulnerability. Hohenkammer, Germany: United Nations University and Munich Re Foundation. Available at
www.ehs.unu.edu/file.php?id=165 (accessed in December 2008).
Kilsby, C. G. (ed.). 2001. Final Report of the WRINCLE Project, Project Number ENV4-CT97-0452 of the
European Union Environment and Climate Research Programme, pp. 47–53. Available at www.ncl.ac.uk/
wrincle/ (accessed in December 2008).
Lehner, B.; Döll, P.; Alcamo, J; Henrichs, T.; Kaspar, F. 2006. Estimating the impact of global change on flood
and drought risks in Europe: A continental integrated analysis. Climatic Change 75: 273–299.
Lehner, B.; Henrichs, T.; Döll, P.; Alcamo, J. 2001. EuroWasser – Model-based assessment of European water
resources and hydrology in the face of global change. Kassel World Water Series 5. Chapters 5 and 7.
Kassel, Germany: Center for Environmental Systems Research, University of Kassel.
Lloyd-Hughes, B.; Saunders, M. A. 2002. A Drought Climatology for Europe. International Journal of Climatology
22: 1571–1592.
Loucks, D. P. 1997. Quantifying trends in system sustainability. Hydrological Sciences Journal 42(4): 513-530.
Malik, L. K.; Koronkevich, N. I.; Zaitseva, I. S.; Barabanova, E. A. 2000. Development of Dams in the Russian
Federation and NIS Countries, A WCD briefing paper prepared as an input to the World Commission on
Dams. Available at www.dams.org (accessed in September 2008).
McMahon, T. A.; Adeloye, A. J.; Zhou, S. L. 2006. Understanding performance measures of reservoirs. Journal
of Hydrology 324(1-4): 359-382.
Middleton, N.; Thomas, D. (eds.). 1997. World Atlas of Desertification, 2nd edition. 182 pp. London: Edward
Arnold.
Mitchell, T. D.; Carter, T. R.; Jones, P. D.; Hulme, M.; New, M. 2004. A comprehensive set of high resolution grids
of monthly climate for Europe and the globe: The observed record (1901-2000) and 16 scenarios (2001-
2100). Tyndall Centre Working Paper 55.
O’Brien, K.; Leichenko, R.; Kelkar, U.; Venema, H.; Aandahl, G.; Tompkins, H.; Javed, A.; Bhadwal, S. 2004.
Mapping Vulnerability to Multiple Stressors: Climate Change and Globalization in India. Global Environmental
Change 14: 303-313.
Peel, M. C.; McMahon, T. A.; Pegram, G. G. S. 2005. Global analysis of runs of annual precipitation and runoff
equal to or below the median: run magnitude and severity. International Journal of Climatology 25: 549-568.
Peel, M. C.; Pegram, G. G. S.; McMahon, T. A. 2004. Global Analysis of Runs of Annual Precipitation and Runoff
Equal to or below the Median: run length. International Journal of Climatology 24: 807-822.
Ramankutty, N.; Foley, J. A. 1998. Characterizing patterns of global land use: An analysis of global croplands
data. Global Biogeochemical Cycles 12(4): 667-685.
Revenga, C.; Brunner, J.; Henninger, N.; Kassem, K.; Payne, R. 2000. Pilot Analysis of Global Ecosystems:
Freshwater Systems. Washington, DC: World Resources Institute.
Schröter, D.; Acosta-Michlik, L.; Arnell, A. W.; Araújo, M. B.; Badeck, F.; Bakker, M.; Bondeau, A.; Bugmann, H.;
Carter, T.; de la Vega-Leinert, A. C.; Erhard, M.; Espiñeira, G. Z.; Ewert, F.; Fritsch, U.; Friedlingstein, P.;
Glendining, M.; Gracia, C. A.; Hickler, T.; House, J.; Hulme, M.; Kankaanpää, S.; Klein, R. J. T.; Krukenberg,
B.; Lavorel, S.; Leemans, R.; Lindner, M.; Liski, J.; Metzger, M. J.; Meyer, J.; Mitchell, T.; Mohren, F.; Morales,
P.; Moreno, J. M.; Reginster, I.; Reidsma, P.; Rounsevell, M.; Pla, E.; Pluimers, J.; Prentice, I. C.; Pussinen,
A.;  Sánchez, A.; Sabaté, S.; Sitch, S.; Smith, B.; Smith, J.; Smith, P.; Sykes, M. T.; Thonicke, K.;  Thuiller, W.;
22
Tuck, G.; van der Werf, G.; Vayreda, J.; Wattenbach, M.; Wilson, D. W.; Woodward, F. I.; Zaehle, S.; Zierl, B.;
Zudin, S.; Cramer, W. 2004. ATEAM Final Report. Potsdam, Germany: Potsdam Institute for Climate Impact
Research (PIK).
Sheffield, J. 2008. Global drought in the 20th and 21st centuries: Analysis of retrospective simulations and future
projections of soil moisture. PhD thesis. Princeton, USA: Princeton University.
Sheffield, J.; Wood, E. F. 2007a. Characteristics of global and regional drought, 1950–2000: Analysis of soil
moisture data from off-line simulation of the terrestrial hydrologic cycle. Journal of Geophysical Research
112: D17115. doi:10.1029/2006JD008288.
Sheffield, J.; Wood, E. F. 2007b. Projected changes in drought occurrence under future global warming from
multi-model, multi-scenario, IPCC AR4 simulations. Climate Dynamics 31(1): 79-105.
Smakhtin, V. U.; Schipper, E. L. 2008. Droughts: the impact of semantics and perceptions. Water Policy 10(2):
131-143.
Snidvongs, A. 2006. Vulnerability to Climate Change Related Water Resource Changes and Extreme
Hydrological Events in Southeast Asia. AIACC Final Reports, Project No. AS 07. Washington, DC, USA: The
International START Secretariat.
Sullivan, C.; Vörösmarty, C. J.; Craswell, E.; Bunn, S.; Cline, S.; Heidecke, C.; Storeygard, A.; Proussevitch, A.;
Douglas, E.; Bossio, D.; Günther, D.; Giacomello, A.; O’Regan, D.; Meigh, J. 2006. Mapping the links between
water, poverty and food security. Report on the Water Indicators workshop held at the Centre for Ecology and
Hydrology, Wallingford, UK, May 16-19, 2005. GWSP Issues in GWS Research No. 1. Bonn, Germany: Global
Water System Project, International Project Office. Available at www.gwsp.org (accessed in January 2008).
Thornton, P. K.; Kruska, R. L.; Henninger, N.; Kristjanson, P. M.; Reid, R. S.; Atieno, F.; Odero, A. N.; Ndegwa,
T. 2002. Mapping poverty and livestock in the developing world. Nairobi, Kenya: International Livestock
Research Institute (ILRI).
UNDP (United Nations Development Programme). 2006. Calculating the Human Development Indices,
Technical Note 1 in Human Development Report, 2006.
UNEP (United Nations Environment Programme). 1992. World Atlas of Desertification. London: Edward Arnold.
UNEP. 2008. Africa: Atlas of Our Changing Environment. Nairobi, Kenya: Division of Early Warning and
Assessment (DEWA), United Nations Environment Programme.
Vörösmarty, C. J.; Federer, C. A.; Schloss, A. L. 1998. Potential evapotranspiration functions compared on US
watersheds: Possible implications for global-scale water balance and terrestrial ecosystem modeling.
Journal of Hydrology 207: 147-169.
White, R. P.; Nackoney, J. 2003. Drylands, people, and ecosystem goods and services: A web-based geospatial
analysis. Washington, DC, USA: World Resources Institute (WRI). Available at www.wri.org/publication/drylands-
people-and-ecosystem-goods-and-services-web-based-geospatial-analysis (accessed in August 2008).
White, W. R. 2005. A Review of Current Knowledge: World Water Storage in Man–Made Reservoirs. FR/R0012.
Marlow, UK: Foundation for Water Research.
Wilhite, D. A. (ed.) 2005. Drought and Water Crises: Science, Technology and management Issues. CRC Press,
Taylor and Francis Group. 406 pp.
Wood, E. F.; Sheffield, J.; Goteti, G. 2003. The Occurrence of Global Drought under Future Climate Scenarios.
American Geophysical Union, Fall Meeting 2003. Abstract #H11H-04.
Zaidman, M. D.; Rees, H. G.; Gustard, A. 2000. An Electronic Atlas for Visualisation of Stream Flow Drought.
Assessment of the Regional Impact of Droughts in Europe (ARIDE) Technical Report 7. Wallingford, UK:
Centre for Ecology and Hydrology.
Zongxue, X.; Jinno, K.; Kawamura, A.; Takesaki, S.; Ito, K. 1998. Performance Risk Analysis for Fukuoka Water
Supply System. Water Resources Management 12: 13-30.
23

IWMI Research Reports
133. Mapping Drought Patterns and Impacts: A Global Perspective. Nishadi Eriyagama,
Vladimir Smakhtin and Nilantha Gamage. 2009.
132. Malaria Transmission in the Vicinity of Impounded Water: Evidence from the
Koka Reservoir, Ethiopia. Solomon Kibret, Matthew McCartney, Jonathan Lautze
and Gayathree Jayasinghe. 2009.
131. Implementing Integrated River Basin Management: Lessons from the Red River
Basin, Vietnam. François Molle and Chu Thai Hoanh. 2009.
130. Economic Gains of Improving Soil Fertility and Water Holding Capacity with Clay
Application: The Impact of Soil Remediation Research in Northeast Thailand.
Rathinasamy Maria Saleth, Arlene Inocencio, Andrew Noble and Sawaeng
Ruaysoongnern. 2009.
129. Adaptive Water Resource Management in the South Indian Lower Bhavani Project
Command Area. Mats Lannerstad and David Molden. 2009.
128. Importance of Irrigated Agriculture to the Ethiopian Economy: Capturing the Direct
Net Benefits of Irrigation. Fitsum Hagos, Godswill Makombe, Regassa E. Namara
and Seleshi Bekele Awulachew. 2009.
127. Drivers and Characteristics of Wastewater Agriculture in Developing Countries:
Results from a Global Assessment. Liqa Raschid-Sally and Priyantha Jayakody.
2008.
126. Climate Change Impacts on Hydrology and Water Resources of the Upper Blue
River Basin, Ethiopia. Ungtae Kim, Jagath J. Kaluarachchi and Vladimir U.
Smakhtin. 2008.
125. The Lower Krishna Basin Trajectory: Relationships between Basin Development
and Downstream Environmental Degradation. Jean-Philippe Venot, Bharat R.
Sharma and Kamineni V. G. K. Rao. 2008.
124. Institutions, Impact Synergies and Food Security: A Methodology with Results
from the Kala Oya Basin, Sri Lanka. Rathinasamy Maria Saleth, Ariel Dinar,
Susanne Neubert, Bandi Kamaiah, Seenithamby Manoharan, Sarath
Abayawardana, Ranjith Ariyaratne and Shyamalie de Silva. 2007.
123. India’s Water Future to 2025–2050: Business-as-Usual Scenario and Deviations.
Upali A. Amarasinghe, Tushaar Shah, Hugh Turral and B. K. Anand. 2007.
Electronic copies of IWMI's publications are available for free.
Visit
www.iwmi.org/publications/index.aspx
IWMI
RESEARCH
R E P O R T
Related Publications Mapping Drought Patterns 
Ahmad, S.; Hussain, Z.; Qureshi, A. S.; Majeed, R.; Saleem, M. 2004. Drought mitigation in Pakistan: 
current status and options for future strategies. Colombo, Sri Lanka: International Water 
Management Institute (IWMI). 54p. (IWMI Working Paper 85: Drought Series Paper 3). 133 and Impacts: 
www.iwmi.org/Publications/Working_Papers/working/WOR85.pdf
IWMI. 2004. Assessment and mitigation of droughts in South-West Asia: issues and prospects. A Global Perspective  
Background Document for the Regional Workshop on Drought Assessment and Mitigation. 
Colombo, Sri Lanka: International Water Management Institute (IWMI). 20p.
www.iwmi.org/droughtassessment/files/pdf/workshop%20docs/Background.pdf 
Samra, J. S. 2004. Review and analysis of drought monitoring, declaration and management in 
India. Colombo, Sri Lanka: International Water Management Institute. 38p. (IWMI Working Paper 
84: Drought Series Paper 2).
www.iwmi.org/Publications/Working_Papers/working/WOR84.pdf
Smakhtin, V. U.; Hughes, D. A. 2004. Review, automated estimation and analyses of drought Nishadi Eriyagama, Vladimir Smakhtin and Nilantha Gamage
indices in South Asia. Colombo, Sri Lanka: International Water Management Institute. 29p. (IWMI 
Working Paper 83: Drought Series Paper 1).
www.iwmi.org/Publications/Working_Papers/working/WOR83.pdf
Biophysical
Thenkabail, P. S.; Gamage, M. S. D. N.; Smakhtin, V. U. 2004. The use of remote sensing data for Vulnerability
Index
drought assessment and monitoring in southwest Asia. Colombo, Sri Lanka: International Water 0 - 20
20 - 30
Management Institute (IWMI). 30p. (IWMI Research Report 85). 30 - 40
40 - 50
www.iwmi.org/Publications/IWMI_Research_Reports/PDF/pub085/RR85.pdf 50 - 60
Annual River 60 - 70
Discharge 70 - 80
(1,000 m3 80 - 99
per person) water bodies
0 - 0.001 no data
0.001 - 1 no agriculture
1 - 10
Postal Address 10 - 100
P O Box 2075 100 - 1,000
1,000 - 10,000
Colombo 10,000-100,000
Sri Lanka >100,000
low population
Location no data
127, Sunil Mawatha
Pelawatta
Battaramulla
Sri Lanka
Telephone
+94-11-2880000
Fax
+94-11-2786854
E-mail
iwmi@cgiar.org
Website
www.iwmi.org
I n t e r n a t i o n a l
I n t e r n a t i o n a l
Water Management ISSN: 1026-0862 Water  Management
I n s t i t u t e ISBN: 978-92-9090-711-4 I n s t i t u t e