research paper about water shortage

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 09 December 2016

The world’s road to water scarcity: shortage and stress in the 20th century and pathways towards sustainability

• M. Kummu 1   na1 ,
• J. H. A. Guillaume 1 , 2   na1 ,
• H. de Moel 3   na1 ,
• S. Eisner 4   na1 ,
• M. Flörke 4   na1 ,
• M. Porkka 1   na1 ,
• S. Siebert 5   na1 ,
• T. I. E. Veldkamp 3   na1 &
• P. J. Ward 3   na1  

Scientific Reports volume  6 , Article number:  38495 ( 2016 ) Cite this article

83k Accesses

535 Citations

228 Altmetric

Metrics details

• Environmental impact

Water scarcity is a rapidly growing concern around the globe, but little is known about how it has developed over time. This study provides a first assessment of continuous sub-national trajectories of blue water consumption, renewable freshwater availability, and water scarcity for the entire 20 th century. Water scarcity is analysed using the fundamental concepts of shortage (impacts due to low availability per capita) and stress (impacts due to high consumption relative to availability) which indicate difficulties in satisfying the needs of a population and overuse of resources respectively. While water consumption increased fourfold within the study period, the population under water scarcity increased from 0.24 billion (14% of global population) in the 1900s to 3.8 billion (58%) in the 2000s. Nearly all sub-national trajectories show an increasing trend in water scarcity. The concept of scarcity trajectory archetypes and shapes is introduced to characterize the historical development of water scarcity and suggest measures for alleviating water scarcity and increasing sustainability. Linking the scarcity trajectories to other datasets may help further deepen understanding of how trajectories relate to historical and future drivers, and hence help tackle these evolving challenges.

Similar content being viewed by others

Reassessing the projections of the World Water Development Report

Future global urban water scarcity and potential solutions

Pollution exacerbates China’s water scarcity and its regional inequality

Introduction.

The overexploitation of freshwater resources threatens food security and the overall wellbeing of humankind in many parts of the world 1 . The maximum global potential for consumptive freshwater use (i.e. freshwater planetary boundary) 2 , 3 is approaching rapidly 4 , regardless of the estimate used. Due to increasing population pressure, changing water consumption behaviour, and climate change, the challenge of keeping water consumption at sustainable levels is projected to become even more difficult in the near future 5 , 6 .

Although many studies have increased the understanding of current blue water scarcity 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , and how this may increase in the future 5 , 6 , 15 , the historical development of water scarcity is less well understood 10 . Trajectories of these past changes at the global scale could be used to identify patterns of change, to provide a basis for addressing future challenges, and to highlight the similarities and differences in water scarcity problems that humanity shares around the world. This requires crossing scales, performing analyses globally, but at a sub-national resolution. Identifying recurring patterns of change can further provide evidence of key drivers of scarcity and thus help to recognise types of problems and solutions. Understanding what has occurred previously can thus help us to avoid repeating mistakes and to build on past successes.

Like other forms of scarcity, physical blue water scarcity can be fundamentally divided into two aspects: shortage and stress. Water shortage refers to the impact of low water availability per person. In “crowded” conditions, when a large population has to depend on limited resources, the capacity of the resource might become insufficient to satisfy otherwise small marginal demands, such as dilution of pollutants in a water body, and competition may result in disputes 16 . Given a resource and per capita requirements, water shortage can therefore be seen as population-driven scarcity. Water stress refers to the impact of high water use (either withdrawals or consumption) relative to water availability. Use of a large portion of a resource 1 , 13 might lead to difficulties in accessing the resource, including side effects 16 , e.g. social and environmental impacts. Stress can be seen as demand-driven scarcity, potentially occurring even if the population is not large enough to cause shortage.

These two aspects have commonly been assessed in isolation from each other 7 , 10 , despite being combined in the seminal work on water scarcity by Falkenmark 1 , 16 , 17 , as well as some later works 15 , 18 . Indeed, the indicators of water shortage and stress are fundamentally related through per capita water use, and therefore provide a more complete picture when used together:

There are, however, multiple ways each of the terms can be defined, yielding different families of indicators for shortage and stress. For example, use can refer to consumption or withdrawals. Availability might refer to water from different sources, of different quality, or at decadal, annual or seasonal time scales. The population in question might be that which is dependent on a resource, which is physically located within a region, or only that which has access to the resource.

Given the complexity of the impacts, these are clearly crude indicators of actual impacts involved in stress and shortage. There is substantial uncertainty in determining at what value of the stress and shortage indicators, stress and shortage impacts actually occur. Even when justified thresholds are selected, the value of the indicator is typically also reported, so that the reader can form their own opinion of whether stress and shortage have really occurred.

Despite their high level of abstraction, and the multiple ways in which they can be used, the concepts of shortage and stress and their defining indicators are central to understanding the development of scarcity over time. Therefore, they provide an obvious first step in analysing trajectories of past changes.

This paper first explores how water consumption has evolved globally over the entire 20 th century. The analysis uses recently released spatially explicit data for the entire past century on socio-economic development 19 and irrigation 20 , which allow us to assess past water consumption trends in greater detail, using the WaterGAP2 hydrological and water use models 19 , 21 (see Methods). This evolution is put into context by assessment of water scarcity based on the concepts of shortage, stress and per-capita consumption, structured graphically using a Falkenmark matrix 1 , 16 , 17 . Archetypes and shapes of the trajectories are introduced as new concepts to characterize the historical development of water scarcity in regions, and hence to assess the effectiveness of potential alleviation strategies and define pathways towards sustainability.

The version of the shortage and stress indicators we use consider decadal scale water availability and consumption at sub-national scales. They therefore capture the effect of long term sub-national water scarcity, but not the seasonal variation in demand and supply, inter-annual variability or sub-regional variation. We focus on physical blue water scarcity, meaning that issues of access are omitted, and emphasis is on water in lakes, rivers and renewable groundwater rather than “green” water, soil water from precipitation directly used by plants, or non-renewable fossil groundwater. Moderate (high) shortage is deemed to occur when total water availability drops below a requirement of 1700 m 3 cap −1 yr −1 , (1000 m 3 cap −1 yr −1 ) 1 , 7 . Moderate (high) stress is deemed to occur when more than 20% (40%) of available water is consumed 1 . The stress threshold was originally applied to water withdrawals but is used here for water consumption to account for substantial return flows that are still available for downstream users 22 , 23 . The focus on consumption also means that water degradation caused by return flows is not considered as part of stress, though it is still (indirectly) captured through population-driven pollutant load as part of shortage.

This study’s findings show a nearly 16-time increase in population under water scarcity since the 1900s although total population increased only 4-fold over the same time period. Per capita water consumption only shows a slight and irregular increase over the past century, while the expansion of water scarcity is predominantly explained by the effects of spatial distribution of population growth relative to water resources.

Water consumption

The global population has almost quadrupled over the past hundred years, and it reached 6.5 billion in the last time step of the study period, i.e. the 2000s (given decadal results are averages over specified decades, in this case 2001–2010) 24 . Over the same period, annual consumptive blue water use per capita (see Methods for details) increased only from 209 m 3 cap −1 yr −1 in the 1900s (i.e., 1901–1910) to 230 m 3 cap −1 yr −1 in the 2000s, with some variation between decades and a maximum of 256 m 3 cap −1 yr −1 in the 1960s ( Fig. 1B ). The increases in population and per capita water consumption resulted in a total water consumption increase from 358 km 3 yr −1 in the 1900s to 1500 km 3 yr −1 in the 2000s ( Fig. 1B ).

Regional ( A ) and global ( B ) consumptive water use trends over the 20 th century. The filled area represents per capita water consumption trends while the dashed line represents the total water consumption trends. The per capita consumption is divided into different water use sectors. The trend in per capita consumption at the FPU scale is shown as a background. [Adobe Illustrator CS5, ArcGIS 9.2 and Matlab 2015b softwares were used to create the figure; http://www.adobe.com/products/illustrator.html , http://www.esri.com , and http://www.mathworks.com ].

The trends of water consumption over the 20 th century were not, however, similar across the globe ( Fig. 1A ). The consumption per capita seems to have remained rather stable in many regions, such as Southern Africa and South America, but declined in the Middle East (since the 1950s), Northern Africa and South Asia. However, per capita consumption increased rapidly in Australia-Pacific, being over 6-fold greater in the 2000s compared to the 1900s. Increases were also found in Eastern Europe & Central Asia (until the 1990s) and Western Europe, although less rapid.

At the FPU (i.e., food production unit; see Methods) scale, this dataset shows that trends in per capita water consumption also varied significantly within the regions ( Fig. 1A ). A good example is North America, where the west coast experienced a decreasing trend while on the east coast, water consumption per capita increased. Of the world population, 46%, 25% and 29% live in FPUs where per capita consumption respectively increased, decreased, or showed no statistically significant trend over time (two-sided p -value > 0.05 with the Mann-Kendall test).

Although the trend in per capita water consumption varied between regions, total water consumption increased in all regions due to increased population except in Eastern Europe and Central Asia, where the total consumption decreased slightly (~7%) since the collapse of the Soviet Union in 1990 ( Fig. 1A ). Growth was greatest in Australia-Pacific (30-fold increase) followed by Central America, Southern Africa, and Southeast Asia (approximately eight-fold). In a number of regions, consumption increased 3–4 fold, with the lowest increase in Northern Africa with about a three-fold increase.

Globally, irrigation was by far the largest water consumer over the entire study period, with a share ranging over time between 90–94% of global water consumption ( Supplementary Fig. 1B ). It had the largest share in South Asia (96–98%) due to extensive rice cultivation, and in the Middle East (97–99%) due to arid conditions 20 . In Western Europe, the irrigation share of total water consumption was lowest (62–74%), as it includes areas where irrigation is not extensively practiced for food production. Moreover, the economy is more industrialised than, for example, in Asia. Globally, the second largest sector until the 1990s was domestic water consumption. However, this was surpassed by industrial water consumption in the final time step (2000s; domestic 3.7%, industrial 4.3%). A second notable global trend is the emergence of water consumption due to thermal electricity production (~1% share). Regionally, results show larger changes in the shares of different sectors, though the real-world significance of the changes is difficult to judge. In some areas (e.g. Western Europe, Australia/Pacific), the proportion of water consumption used for irrigation has increased and the proportion for domestic consumption has decreased. The opposite has occurred in other areas (e.g. North America, Supplementary Fig. 1A ).

Global and regional water scarcity

Despite only small variations in per capita water consumption over time ( Fig. 1A ), rapidly expanding local populations and increases in total water consumption resulted in a nearly 16-fold overall increase in the population under water scarcity within the 20 th century ( Figs 2 and 3 ). Whilst in the 1900s just over 200 million people (14% of global population) lived in areas under some degree of water scarcity, this number increased to over two billion by the 1980s (42%), and reached 3.8 billion people (58%) by the 2000s ( Table 1 ; Fig. 2B ).

Regional ( A ) and global ( B ) water scarcity trajectories. Filled graphs represent the absolute population under water scarcity (in billions) while dashed lines represent the population relative to total regional population. M WStr refers to moderate water stress, H WStr to high water stress, M WSh to moderate water shortage, and H WSh to high water shortage. See definitions of these different water scarcity dimensions, and their combinations, in Table 1 and Fig. 4A . [Adobe Illustrator CS5, ArcGIS 9.2 and Matlab 2015b softwares were used to create the figure; http://www.adobe.com/products/illustrator.html , http://www.esri.com , and http://www.mathworks.com ].

Mapped water scarcity categories for years 1905 ( A ), 1935 ( B ), 1965 ( C ), 1985 ( D ), and 2005 ( E ). The definition for each scarcity category is given in Table 1 and Fig. 4A . [Adobe Illustrator CS5 and ArcGIS 9.2 softwares were used to create the figure; http://www.adobe.com/products/illustrator.html , http://www.esri.com ].

In the 2000s, roughly half of the people under water scarcity suffered either moderate water shortage or moderate water stress ( Table 1 ), while the other half lived in areas facing both water stress and water shortage. Of these, 1.1 billion people (17% of global population) lived in areas facing both high water shortage and high water stress ( Table 1 ; Fig. 2B ). Most of these people lived in South and East Asia, North Africa and Middle East ( Fig. 2A ), with 61–89% of the population under water scarcity. The regions with the lowest proportion of population under water scarcity were Australia-Pacific, South America, North America, and Southeast Asia (7–29%, Fig. 2A ). Around a half of the population under water scarcity in the 2000s suffered water shortage alone, without water stress ( Table 1 ; Fig. 2B ). These areas are located in Sub-Saharan Africa, Central America, Europe, and South and East Asia ( Figs 2A and 3E ). A small part of the population (2%) suffered water stress alone ( Table 1 ), occurring mostly in North America, Middle East, and Australia ( Fig. 3E ).

A global water scarcity trend-plot ( Fig. 2B ) reveals that the population under water shortage, or a combination of high water stress and water shortage, has increased rapidly since the 1960s, while water stress alone has remained rather low over the entire study period. There are, however, differences in regional trajectories ( Fig. 2A ), indicating that, for example, in the Middle East, Northern Africa and North America, scarcity has developed gradually over the whole study period while in many other regions (e.g. Central America, Southern Africa, South Asia, Southeast Asia, and East Asia) there has been a steep increase in scarcity trend since the 1960s.

Different FPUs show distinct population dynamics, climate patterns, and developments of water consumption per capita. An FPU’s long-term water scarcity trajectory over time is visualised using the Falkenmark matrix 16 ( Fig. 4 ) that distinguishes between population-driven water shortage and demand-driven water stress, and highlights the relationship with per capita consumption using superimposed diagonal lines. Drivers and adaptation strategies are strongly dependent on the level and type of water scarcity an FPU is experiencing ( Fig. 4B ). As defined in Table 2 and discussed below, the notions of archetypes and shapes help to make sense of these trajectories. The archetype refers to the positioning within the Falkenmark matrix ( Fig. 5 ), whilst shape ( Fig. 6 ) refers to the direction of change over time.

FPU water scarcity trajectories by scarcity archetypes in a map ( A ) and within the Falkenmark matrix ( B–G ). Archetypes categorise FPUs according to their water scarcity status (corresponding to position on the plot) and where both shortage and stress occur, according to which occurs first (which is related to the level of per capita consumption). The trajectories are grouped based on irrigation zone 20 they are located in. See Table 2A for definitions and Supplementary Table 2 for percentage of population in each archetype – irrigation zone combination. Note: only FPUs with more than one million people are included. [Adobe Illustrator CS5 and R studio softwares were used to create the figure; http://www.adobe.com/products/illustrator.html , https://www.rstudio.com ].

FPU water scarcity trajectory shapes.

( A and B ) FPU shapes shown as map, separated according to whether scarcity has been experienced ( B ) or not ( A ). ( C ) Examples of shapes of FPU water scarcity trajectories. The diagonal lines refer to per capita consumption isolines and numbers to FPUs (location indicated in tile B ). See Table 2B for definition of each shape category and Supplementary Fig. 2 for each FPU trajectory categorised by their shape. [Adobe Illustrator CS5 and R studio softwares were used to create the figure; http://www.adobe.com/products/illustrator.html , https://www.rstudio.com ].

FPU water scarcity trajectories: archetypes

The concept of water scarcity trajectory archetypes captures issues related to water scarcity status and per capita consumption. Trajectory archetypes are thus also useful to identify possible adaptation measures in an FPU. Their definitions are summarised in Table 2A while Fig. 5 maps the regions belonging to each archetype, and displays their trajectories. Each archetype is discussed further below.

The archetypes stress alone or stress first (before shortage) are experienced if per capita consumption is high ( Fig. 5F,G ), such that scarcity is demand-driven. FPUs in this category would thus benefit most from demand-side oriented adaptation strategies. The archetypes shortage alone or shortage first ( Fig. 5C,D ) are experienced if per capita consumption is low, such that scarcity is population-driven. This calls for supply-side adaptation strategies in particular. This division of adaptation strategies also corresponds to a distinction between ‘soft’ behaviour-change and ‘hard’ infrastructure-based solutions, respectively 1 , 17 , 25 , 26 ( Supplementary Table 1 ).

Specifically, using a threshold of stress of 20% and a per capita water availability (shortage) threshold of 1700 m 3 cap −1 yr −1 , the switch-over point between stress first and shortage first occurs at a per capita consumption of 340 m 3 cap −1 yr −1 ( Fig. 5A–F ; Methods). The stress and shortage at same time archetype is a borderline case, in which per capita consumption varies near that switch-over point. For that archetype, both adaptation strategies may be relevant. When an FPU is of a no scarcity archetype, no direct adaptation measures are necessary. However, as population grows, the per capita consumption of an FPU sets it on a trajectory towards either stress first or shortage first, and so the above introduced guidelines may apply.

For stress first and stress alone archetypes, the need for demand management rather than supply side measures 1 is motivated by the common ideological point of view that high per capita water consumption should be reduced. In practice, however, there seems to be a tendency to meet demand first, for example in the case of trajectories with a constant per capita demand shape (see Fig. 6C ). This might be explained in terms of the “hydraulic mission” 27 , common around the world in the 20 th century, which aims to dominate nature in order to increase food production and provide water and food security. This has to some extent been curbed by increased emphasis on social and environmental impact assessment 27 , 28 . Ideally, adaptation strategies should focus first on increasing water productivity (domestic, agricultural, and industrial) or on shifting to lower water footprint goods and services. The latter might include reducing virtual water exports 29 and/or increasing virtual water imports 30 . Several of these actions would not be captured by the data and analysis applied, and may have already occurred, as suggested by recent studies 29 , 31 , 32 , 33 , 34 .

For cases where shortage occurs before stress, supply-side options are in principle preferred because lower per capita water consumption provides less potential for demand-side intervention than when stress occurs first. There are, however, two main ways to handle water shortage: (i) increasing available water, or (ii) limiting population. Available water can be increased by using desalination (in coastal areas) 35 , introducing physical water transfers 36 , 37 and/or reducing non-productive evaporation 38 . Increased storage capacity is likely to play a smaller role at decadal scale, but is a common strategy to increase seasonal or inter-annual water availability. Emigration and lowered birth rates may limit population, but are perhaps better treated as side-effects of other developments rather than explicit water scarcity strategies. Moreover, an area can adapt to water shortage by using the strategies to reduce per capita water consumption. Possibilities for reducing water requirements include more efficient irrigation 39 , reduction of food losses 40 , reduction of water-intensive goods 41 , 42 , and reduction of leakages in public supply systems 43 .

The potential for reducing blue water consumption notably depends on green water availability (soil water from precipitation), especially in the case of agriculture 13 , but also, for example, on urban parks and golf courses. Areas with reliable green water resources tend to have lower blue water consumption, and hence less stress. While this study does not quantify green water availability, it does show that different archetypes occur depending on the reason for irrigation consumption (which is the largest water-consumption sector in most areas). As discussed in Siebert et al . 20 , irrigation is notably driven by: (i) the desire to make agriculture possible in arid areas; (ii) the desire to increase productivity in semi-arid and temperate areas; or (iii) weed-suppression by controlling the water level when growing rice. The results by irrigation zones 20 (see Fig. 5 for trajectories by irrigation zones, and tabulated results for population in Supplementary Table 2 ) indicate, for example, that most of the high per capita consumption stress first (90% of FPUs within those archetypes) or stress alone (82% of FPUs) trajectories occur in arid regions, consistent with higher crop water requirements due to irrigation. Shortage alone in turn occurs commonly in wet areas (50% of FPUs), consistent with low water requirements and high population pressure.

In practice, it appears that shortage is not directly tackled until stress occurs. Moderate shortage is tolerated, perhaps buffered by low consumption and other water sources, such as virtual water imports, green water and fossil groundwater. This avoids tackling the underlying issue of population growth, and stress is reached some time later. For example, in North-eastern Mainland China, some FPUs have experienced shortage since before 1905, and others more recently since 1925 and 1975 ( Supplementary Fig. 3 ). Stress followed years or decades later, as population grew. Groundwater and a number of inter-basin transfers are already in use, and additional south-north transfers are in development 44 , 45 . These FPUs are good examples where per capita blue water consumption is low enough that shortage occurred first. There is, however, significant potential for further reductions due to large virtual water exports, which could avoid the need for inter-basin transfers 45 .

FPU water scarcity trajectories: shapes

When FPU trajectories are distinguished by their shape , it is possible to understand the dynamics of consumption over time, and how that has impacted on the scarcity type (shapes are summarised in Table 2B ; example trajectories for each shape are shown in Fig. 6C and all trajectories in Supplementary Fig. 2 ). Further, shapes can be used to assess what needs to be done for an FPU to be put on a sustainable pathway, avoiding both water stress and water shortage in the long term. The majority of FPUs show significant temporal variation in per capita water consumption, stress, and shortage, consistent with the expected tension between population growth, water supply and demand management. In general, achieving sustainable water consumption on a decadal scale requires a combination of stabilising population, enforcing limits of sustainable supply, mitigating impacts of water stress and/or reducing water requirements.

All these strategies are likely to be required to deal with FPUs in the shape categories increasing scarcity and other . The former face both incessant population growth and intensification of water consumption, which currently leads to strictly increasing stress and shortage (6.6% of global population in 2000s, Fig. 6 ), for example in parts of the Balkans (FPU 169, Fig. 6 ). The other shape category (32.2% of the population) shows complex trajectories for which specific recommendations cannot be made without other economic or demographic data.

In FPUs where the trajectory shape is determined by constant per capita demand (29% of population), changes in scarcity are predominantly determined by population growth. Constant per capita demand is visible as a (relatively) straight diagonal trajectory in the Falkenmark matrix ( Figs 4B and 6C ). As long as per capita consumption is kept in check, stabilising population is an effective strategy for FPUs with any trajectory shape as it avoids increases in shortage and total consumption, and hence stress.

In FPUs with strictly increasing stress but varying shortage (4.9% of population), consistent intensification of water consumption is the key concern, for example in northern France (FPU 121, Fig. 6 ). Recognising the socio-economic importance of exploitation of the local water resource and potential difficulty in curbing water consumption, achieving sustainability may involve mitigation measures to allow greater water consumption than would otherwise be possible. Examples include improving water allocation and other governance mechanisms, providing storage and channelling engineering works, optimising environmental flows, and benefit-sharing to compensate other impacted users. This corresponds to the idea of ‘decoupling’ growth from impacts 46 .

In FPUs with strictly increasing shortage but varying stress (15% of population), the key concern is strong population growth, as in northern India (FPU 494, Fig. 6 ). Recognising that addressing the drivers of population growth may take time, achieving sustainability may involve reducing local water requirements, so that consumption does not grow in parallel with population. This corresponds to decoupling growth from resource use and may be achieved by improved water productivity or decreasing water-dependent production 40 , 41 . Decoupling from resource use already appears to be occurring in many areas, as shown by decreasing trends for per capita consumption ( Fig. 1 ). In FPUs where irrigation is important, per capita consumption is particularly influenced by area equipped for irrigation and a combination of irrigation efficiency and climate effects. However, the most prominent examples of decoupling from local resource use are FPUs dominated by cities, taking as an example FPU 307 in western Africa (32 million people in 2000s), which includes the megacity of Lagos in Nigeria. While some food and other water-dependent products are produced in the hinterland, they can also be imported from elsewhere (along with virtual water) 47 . Such areas can therefore have relatively low local blue water requirements, mainly for domestic and industrial water supply (83% of total water consumption at FPU 307). The sustainability of such FPUs depends largely on their interactions with regional and global water resources.

In addition to cases where trends suggest that decoupling is occurring, the analysis identifies some cases with a stress decrease -shape (10% of population), or where stress stabilised ( stress ceiling -shape, 2% of population). In most cases, this occurs as a result of decreases in consumption, but appears to be driven often by socio-economic factors rather than limited water availability. Results show that FPUs that have reached a stress ceiling are mostly those with high per capita consumption that suffer water stress alone ( cf. Figs 3 and 6B ) in North America, Central Asia, or Africa. However, stress ceilings occur even with a stress level of 10% (e.g. in Northern Africa), and decreases in stress in FPUs that are not water scarce in large parts of the former Soviet Union ( Fig. 6A ), following the dissolution of the Soviet Union. This may thus be related to the region’s political and economic changes. Consistent with the idea of a “hydraulic mission” 27 , 28 , dams and canals increased supply to allow irrigation demand to expand. Reductions in consumption then occurred not just due to improvements in irrigation efficiency but also due to a shift from exported cotton (and virtual water 29 ) to food self-sufficiency in the newly independent nation states 48 , 49 . Water scarcity trajectories and their sustainability are closely tied with other socio-economic and political issues.

This study highlights key issues in understanding global historical water scarcity and pathways for future adaptation. Considering both forms of water scarcity, this analysis provides an improved understanding of blue water consumption and trajectories of past water scarcity development globally at sub-national level for the entire 20 th century. The results show that more people are under water scarcity than previously estimated ( Supplementary Table 4 ).

Only a few previous studies assessed historical water scarcity using multiple water use sectors 10 , 19 , 50 , and even then only for the past 50 years. This study’s results compare well with previous trends and estimates of water consumption since 1960, the starting period of existing assessments 10 , 50 ( Supplementary Table 3 ). The largest improvement in this study, in terms of water consumption trends, is the use of historical spatially explicit irrigation maps 20 rather than national values. This results in large differences in the location and extent of irrigation areas, particularly in large countries, such as the USA 20 .

Findings for population under stress and shortage separately also show good agreement with existing studies of historical water scarcity ( Supplementary Table 4 ). The existing studies focus on water stress alone 10 or water shortage alone 7 , or assess both forms of scarcity at only one or two time steps 16 or scenarios 29 , with the exception of one study 18 that assesses the interannual variability of blue water scarcity. Assessing both shortage and stress over several decades provides additional insights on the development of water scarcity. The FPU-level trajectories show signs not just of differences in resource endowments and local history, but also similarities due to shared problems and diffusion of solutions, suggestive of a global shared destiny for which collaboration is essential. Classifying sub-national water scarcity trajectories in terms of archetypes ( Fig. 5 ) helps to highlight possible adaptation actions to cope with shortage and/or stress, depending on the level of water consumption in per capita terms. Classifying trajectories in terms of their shape ( Fig. 6 ) helps to highlight different approaches to put FPUs on a sustainable pathway. Nearly all FPUs show an increase in scarcity over time as population increases ( Fig. 6 ; Supplementary Fig. 2 ), indicating that understanding of scarcity adaptation actions and pathways to sustainability will only become more important in the future. These historical trajectories provide a common foundation from which further work can dig deeper to identify mistakes to avoid repeating, and past successes worth replicating, in order to better tackle future challenges of water scarcity.

As noted in Introduction, results presented correspond to a well-defined scope focussed on scarcity associated with a long-term view of consumptive blue water use. The selected indicators are widely adopted and can be linked to previous studies 8 , 9 , 10 , 14 , 18 . Additional information sources that would allow more sophisticated water scarcity analysis are not available for the entire study period. These include water quality, technological and social access to water and trade of virtual water. Future studies could include these aspects.

Furthermore, the analysis is commensurate with the significant uncertainty involved in the datasets and models used to cover the globe for the past 110 years 51 , 52 . In this study, two important datasets are combined: water availability and water use, both provided by the WaterGAP2 model. In order to reduce uncertainty in water availability estimates, the model has been calibrated in a basin-specific manner against mean annual river discharge using 1319 gauging stations 53 . Previous studies have reported that the model performs well in relation to other global hydrological models when compared to observations 51 , giving confidence in our water availability estimates. Water use data, on the other hand, is viewed as particularly uncertain 54 . For example, in a multi-model comparison, Wada et al . 55 show that modelled irrigation demand compares reasonably well to country-scale reported values (deviations in the range of +/− 15% in most cases) and conclude that most models are capable of simulating regional variability in irrigation water demand across the globe. Since irrigation constitutes the largest share to global total water consumption and is the dominant water-consuming sector in many parts of the world, it is very likely to also dominate the uncertainty in estimated total water consumption.

We compared the water consumption data of this study to two previous studies assessing the past water consumption 10 , 50 ( Supplementary Table 3 ), and found that the consumption estimates vary on the order of 35%, this study being the most conservative one. When our water scarcity results were compared to existing studies 10 , 18 ( Supplementary Table S4 ), we found that estimates of global population under shortage, and population under stress vary on the order of 15% and 30% respectively.

Besides these two key input data products, various assumptions have been made in the analysis itself. A notable assumption relates to the thresholds used to differentiate different states of water stress and shortage. Whilst these assumed thresholds directly affect the amount of population living under water scarcity, they do not affect the trajectory lines in the Falkenmark matrix themselves. Correspondingly, the shapes of the trajectories are not affected by these thresholds. However, trajectory archetypes would somewhat be impacted, as changing these thresholds would mean a specific FPU reaches a certain level of scarcity a decade earlier or later.

As a result, our emphasis is on drawing coherent insights rather than providing precise estimates. In this context, specific numbers represent one possible realisation in the context of significant uncertainty. This is important when comparing our results for a specific year with other studies. The key conclusions of this study are, however, robust, namely the interpretation of sub-national shortage and stress trajectories and the importance of population growth and per capita water consumption in determining local development of scarcity. They are consistent with existing understanding, and strongly influenced by patterns in input data (e.g. population growth and expansion of irrigation area) that are independent of other assumptions made in the analyses.

The analytical approach used and the initial insights it provides could also be used as a foundation for further research. Additional information about uncertainty could be obtained by systematically repeating the analysis with other models and forcing datasets, as has been done in comparable contexts 5 . This would, however, require a carefully chosen, meaningful set of scenarios. A range of different assumptions can be used regarding scarcity thresholds and indicators, focussing on different issues delimiting different perspectives on safe and just operating spaces for socio-ecological systems 3 , 56 . Calculating indicators at seasonal 11 , 57 or annual scale 18 , 58 would allow investigation of how shortage and stress occur at shorter time scales, more closely related to every-day operations rather than long-term planning. Ideally, availability would be tied to access, which would help alleviate problems related to selection of spatial scale 59 . Focussing on water quality 60 , 61 , unsustainable water sources 62 , and on spatially explicit environmental flow requirements 4 , 63 (the thresholds used for water stress assume global environmental flow requirements of 30% 17 ) would explicitly identify the portion of available water that should not be used to avoid stress according to different criteria. Similarly, focussing on self-sufficiency of water and food 12 , 58 , 64 would identify specific water requirements for shortage, though it would also require greater consideration of both blue and green water 13 .

Whether self-sufficiency is required is particularly relevant in the context of trade 65 and virtual water transfers 31 , which are not captured in this study. From an economics perspective, scarcity is not intrinsically problematic, but rather raises questions of optimal allocation of the scarce resources, trade to make use of comparative advantages, and the inclusion of externalities. Prominent issues include the role of water quality and safety 66 , and accessibility and equity determined by social, economic and political circumstances 25 , 67 , 68 , 69 , 70 , 71 . Linking the trajectories to other datasets may help deepen understanding, expanding and better explaining the shapes introduced here ( Table 2B ), and how they relate to historical and future drivers as well as limits to adaptation.

Analysis unit: Food production units

This study used food production units (FPUs), a combination of river basin and administrative boundaries 7 , 72 , 73 , as an analysis unit. These are reported to be suitable for water scarcity studies 7 , 58 . For this project, a set of FPUs were developed that are consistent with the basin delineation of the WaterGAP2 hydrological and water use models, resulting in 548 FPUs. It is important to use the same delineation for FPUs as watersheds of the WaterGAP2 model, as the way water availability is dealt with (see Fig. 7 ) requires that FPUs do not cross the borders of large river basins. Results are also aggregated from the FPU scale to regional ( n  = 12) scale. The regions are based on UN macro regions aggregating the countries to larger units 74 with the difference that some of the largest regions were divided into smaller regions by Kummu et al . 7 to be more suitable for (historical) water analyses.

Water availability calculations in a large basin with several FPUs, i.e. each FPU is a sub-basin for the large basin. A: schematic illustration of a basin with four FPUs; B: Runoff of each grid cell in km 3 yr −1 ; and C: discharge of each grid cell in km 3 yr −1 . The share of available water resources is calculated as the sum of discharges of each grid cell within an SBA divided by the sum of discharges of all grid cells within a basin. The available water resources are then calculated by multiplying that share with the total available runoff of the whole basin. [Adobe Illustrator CS5 –software was used to create the figure; http://www.adobe.com/products/illustrator.html ]

Water availability

This analysis used the global hydrological model WaterGAP2 53 to derive gridded estimates for runoff and river discharge at 30 arc-min spatial resolution for the study period of 1901–2010. Based on daily meteorological forcing fields and spatially distributed physiographic information (e.g. soil, land cover), the model simulates the terrestrial water cycle by a sequence of storage equations for the storage compartments canopy, snowpack, soil, renewable groundwater, and surface water bodies. For this study, simulations were driven by WATCH Forcing Data (WFD) which is available for the period 1901–2001 75 . Since it is not recommended to combine WFD with other similar data-sets 53 , 76 in order to derive full coverage over the study period 1901–2010, simulations for the period beyond the year 2001 were based on 1990s climate forcing.

Since this analysis focuses on long-term trends in water scarcity, the 10-yr annual average over each decade was calculated for both discharge and runoff to compensate for inter-annual variability. These data were then used to assess the water availability in each FPU. The calculation of water availability can be divided into two cases:

In cases when an FPU consisted of one basin or several small basins, water availability was simply the sum of annual runoff generated within the area of a specific FPU.

In cases of large river basins that were divided into several FPUs, a simple ‘water sharing rule’ was used to assign the available freshwater resources within each FPU 5 , 12 . This was developed in a way that it would be usable for both water shortage and water stress calculations, i.e. the sum of water availability of the FPUs within the basin cannot exceed the annual runoff of the basin. The water sharing rule was based on a discharge proportion of FPUs within a basin multiplied with the annual runoff, as illustrated in Fig. 7 .

The water use model of WaterGAP2 simulates water withdrawals and consumption of the following sectors: i) irrigation, ii) livestock farming, iii) thermal electricity production, iv) manufacturing industries, and v) households and small businesses (domestic).

To indicate the area equipped for irrigation (AEI), the analysis used the HID product by Siebert et al . 20 , which gives spatially explicit AEI for the entire 20 th century. The proportion of irrigated harvested rice area was based on the MIRCA-2000 dataset 77 . The proportions were kept at year 2000 level throughout the study period due to lack of historical data. As in the case of the water availability simulations (see above), to simulate the irrigation water consumption beyond 2001, climate forcing data from the 1990s were used. The estimate of consumption for the 2000s should therefore not be included when assessing trend in per capita consumption. Irrigation water consumption is the amount of water that must be applied to the crops by irrigation in order to achieve optimal crop growth. Monthly consumptive irrigation requirements are therefore based on climate, the spatial extent of AEI and crop type (rice and non-rice). Return flows, i.e. water withdrawal minus water consumption, which account for water that infiltrates and returns to the water cycle, are not quantified in this study.

Livestock water consumption was calculated on the basis of gridded information on the number of livestock units and water consumption per head and year, taking into account 10 livestock types 21 . Due to limited data prior to the year 1960, livestock water consumption for the period of 1900–1960 was kept at the level of 1960. Overall, this may lead to an underestimation or overestimation in livestock water consumption depending on the FPU 78 , which is expected to be minor as the amount of livestock water consumption is small compared to the other sectors. Water consumption estimates for electricity, manufacturing, and domestic sectors were based on the methodologies described in Flörke et al . 19 . In brief, domestic water consumption is estimated from population and domestic water use intensity, taking into account structural and technological changes. Country-scale water consumption in the manufacturing sector is calculated from manufacturing structural water use intensity, gross value added, and consumption coefficients; again taking into account technological change. The amount of water withdrawn and consumed for cooling purposes in thermoelectric power production is determined from the annual thermal electricity production and the water use intensity of each power station, distinguishing three cooling system types (once-through, pond, and tower cooling systems) and several fuel types (fossil/biomass/waste-fuelled, nuclear, natural gas/oil combined, coal/petroleum residuum-fuelled). Based on this information, the model approach distinguishes 14 combinations of plant type (PT) and cooling system (CS). In 2010, about 2.8% of cooling water abstractions evaporated, i.e. most of the water withdrawn was discharged back into rivers (Flörke et al . 19 ).

To get the total water consumption, all the water use sectors are summed together. Trends in per capita consumption (see background in Fig. 1A ) were determined with the Mann-Kendall test, calculating the Kendall correlation of demand with time. A p -value of 0.05 was used as part of a two-sided test of whether the correlation was statistically significantly different from zero.

Water stress calculations

The indicator of blue water stress is the water use to availability ratio. We use consumption rather than withdrawals, such that water ‘use’ means that water is no longer available for other users. The indicator was calculated for each decade and for each FPU. The water stress thresholds used are, however, those for the withdrawal-based water stress index (WSI) developed by Falkenmark 16 , and used by a number of other studies 8 , 10 , 57 , 78 :

WSI <0.2: no water stress

WSI = 0.2–0.4: moderate water stress

WSI >0.4: high water stress

Using withdrawals risks over-estimating the actual stress as a substantial part of the withdrawals are available for downstream users as return flows 22 , 23 . On the other hand, using water consumption, as in this study, might underestimate the water stress. Recent work by Munia et al . 79 uses consumption and withdrawals as minimum and maximum levels of scarcity, respectively. They show that the difference between these two estimates results in an 18 percent point difference in the amount of population under water stress. Similar uncertainties in the absolute amount of people under water scarcity should be considered for the numbers quoted in this study. This may also be worthwhile approach for future work. Finally, it should be stressed that the thresholds used assume a global environmental flow requirements of 30% 17 .

Water shortage calculations

For water shortage calculations the analysis is based the water crowding index (WCI) developed by Falkenmark 17 , 80 . WCI is calculated by dividing the water availability by total population of an FPU. Here, historical, spatially explicit, population data is from HYDE 3.1 81 . The water shortage thresholds are as follows:

WCI >1700 m 3 cap −1 yr −1 : no water shortage

WCI = 1000–1700 m 3 cap −1 yr −1 : moderate water shortage

WCI <1000 m 3 cap −1 yr −1 : high water shortage

Water scarcity matrix and related calculations

To illustrate the combination of water stress and water shortage, the analysis used the Falkenmark water scarcity matrix ( Fig. 4 ). By plotting water stress against shortage over time, water scarcity trajectories were derived for each FPU. These trajectories in turn were categorised for archetypes and shapes ( Table 2 , and see below).

The formulas used for the indicators mean that for any combination of stress and shortage, per capita consumption can also be calculated (see diagonal lines in Fig. 4B ). For example, consider the point where an FPU is classified as under both water stress and water shortage:

The corresponding per capita consumption can be calculated for those values of stress and shortage (see also Fig. 4B ):

For a given per capita consumption, this formula can be rearranged to identify whether an FPU would already be stressed when the shortage threshold is reached (shortage = 1700 m 3 cap −1 yr −1 ).

Therefore, the following interpretation can be made when assuming shortage of 1700 m 3 cap −1 yr −1 :

If per capita consumption = 340 m 3 cap −1 yr −1 → stress = 0.2 (stress and shortage same time)

If per capita consumption >340 m 3 cap −1 yr −1 → stress >0.2 (stress occurs first)

If per capita consumption <340 m 3 cap −1 yr −1 → stress <0.2 (shortage occurred first)

Scarcity archetypes

The scarcity archetypes define the water scarcity status and level of per capita consumption (see Table 2A ). Scarcity categorisation for archetypes is based on the lowest stress (20%) and shortage thresholds (1700 m 3 cap −1 yr −1 ). ‘No scarcity yet’ are FPUs that have never reached the lowest threshold of water stress (20%) or shortage (1700 m 3 cap −1 yr −1 ). For ‘Shortage alone’, water availability has passed the threshold of 1700 m 3 cap −1 yr −1 , but stress has remained below the threshold of 20%. ‘Stress alone’ occurs where stress exceeds 20% but water availability (i.e. shortage) has never dropped below 1700 m 3 cap −1 yr −1 . ‘Stress first’, ‘Shortage first’ and ‘Stress and shortage at same time’ occur when the trajectory has exceeded both the stress and shortage thresholds, sub-categorised according to which type of strategy is reached first.

Scarcity shapes

The scarcity shapes, in turn, divide the trajectories into categories based on their shape when plotted in the Falkenmark matrix. Specific rules for each shape were developed as outlined in Table 2B .

Additional Information

How to cite this article : Kummu, M. et al . The world’s road to water scarcity: shortage and stress in the 20th century and pathways towards sustainability. Sci. Rep. 6 , 38495; doi: 10.1038/srep38495 (2016).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Falkenmark, M. Growing water scarcity in agriculture: future challenge to global water security. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 371 (2013).

Rockström, J. et al. A safe operating space for humanity. Nature 461, 472–475 (2009).

Article   ADS   CAS   PubMed   Google Scholar  

Dearing, J. A. et al. Safe and just operating spaces for regional social-ecological systems. Global Environmental Change 28, 227–238 (2014).

Article   Google Scholar  

Gerten, D., Hoff, H., Rockström, J., Jägermeyr, J., Kummu, M. & Pastor, A. V. Towards a revised planetary boundary for consumptive freshwater use: role of environmental flow requirements. Current Opinion in Environmental Sustainability 5, 551–558 (2013).

Schewe, J. et al. Multimodel assessment of water scarcity under climate change. Proceedings of the National Academy of Sciences 111, 3245–3250 (2014).

Article   ADS   CAS   Google Scholar  

Gosling, S. & Arnell, N. A global assessment of the impact of climate change on water scarcity. Climatic Change, 1–15 (2013).

Kummu, M., Ward, P. J., de Moel, H. & Varis, O. Is physical water scarcity a new phenomenon? Global assessment of water shortage over the last two millennia. Environmental Research Letters 5, 034006 (2010).

Article   ADS   Google Scholar  

Oki, T. & Kanae, S. Global Hydrological Cycles and World Water Resources. Science 313, 1068–1072 (2006).

Vörösmarty, C. J., Green, P., Salisbury, J. & Lammers, R. B. Global Water Resources: Vulnerability from Climate Change and Population Growth. Science 289, 284–288 (2000).

Article   ADS   PubMed   Google Scholar  

Wada, Y., van Beek, L. P. H. & Bierkens, M. F. P. Modelling global water stress of the recent past: on the relative importance of trends in water demand and climate variability. Hydrology and Earth System Sciences 15, 3785–3808 (2011).

Wada, Y., van Beek, L. P. H., Viviroli, D., Dürr, H. H., Weingartner, R. & Bierkens, M. F. P. Global monthly water stress: 2. Water demand and severity of water stress. Water Resoures Research 47, W07518 (2011).

Gerten, D., Heinke, J., Hoff, H., Biemans, H., Fader, M. & Waha, K. Global water availability and requirements for future food production. J Hydrometeorol 12, 885–899 (2011).

Rockström, J., Falkenmark, M., Karlberg, L., Hoff, H., Rost, S. & Gerten, D. Future water availability for global food production: The potential of green water for increasing resilience to global change. Water Resour Res 45, W00A12 (2009).

Alcamo, J. et al. Global estimates of water withdrawals and availability under current and future “business-as-usual” conditions. Hydrological Sciences Journal 48, 339–348 (2003).

Alcamo, J., Flörke, M. & Märker, M. Future long-term changes in global water resources driven by socio-economic and climatic changes. Hydrological Sciences Journal 52, 247–275 (2007).

Falkenmark, M. Meeting water requirements of an expanding world population. Philos Trans R Soc Lond B Biol Sci 352, 929–936 (1997).

Article   ADS   PubMed Central   Google Scholar  

Falkenmark, M., Berntell, A., Jägerskog, A., Lundqvist, J., Matz, M. & Tropp, H. On the Verge of a New Water Scarcity: A Call for Good Governance and Human Ingenuity. In: SIWI Policy Brief. Stockholm International Water Institute (SIWI) (2007).

Veldkamp, T. I. E. et al. Changing mechanism of global water scarcity events: impacts of socioeconomic changes and inter-annual hydro-climatic variability. Global Environmental Change 32, 18–29 (2015).

Flörke, M., Kynast, E., Bärlund, I., Eisner, S., Wimmer, F. & Alcamo, J. Domestic and industrial water uses of the past 60 years as a mirror of socio-economic development: A global simulation study. Global Environmental Change 23, 144–156 (2013).

Siebert, S., Kummu, M., Porkka, M., Döll, P., Ramankutty, N. & Scanlon, B. R. A global data set of the extent of irrigated land from 1900 to 2005. Hydrology and Earth System Sciences 19, 1521–1545 (2015).

Alcamo, J. et al. Development and testing of the WaterGAP 2 global model of water use and availability. Hydrological Sciences Journal 48, 317–338 (2003).

Hoekstra, A. Y. & Mekonnen, M. M. The water footprint of humanity. Proceedings of the National Academy of Sciences 109, 3232–3237 (2012).

Kendy, E., The False Promise of Sustainable Pumping Rates. Ground Water 41, 2–4 (2003).

Article   CAS   PubMed   Google Scholar  

United Nations. World Population Prospects: The 2012 Revision, Methodology of the United Nations Population Estimates and Projections. United Nations, Department of Economic and Social Affairs, Population Division. ESA/P/WP.235 (2014).

Gleick, P. H. Water in crisis: Paths to sustainable water use. Ecol Appl 8, 571–579 (1998).

Gleick, P. H. Global Freshwater Resources: Soft-Path Solutions for the 21st Century. Science 302, 1524–1528 (2003).

Molle, F., Mollinga, P. P. & Wester, P. Hydraulic bureaucracies and the hydraulic mission: flows of water, flows of power. Water Alternatives 2, 328–349 (2009).

Google Scholar  

Abdullaev, I. & Rakhmatullaev, S. Transformation of water management in Central Asia: from State-centric, hydraulic mission to socio-political control. Environ Earth Sci 73, 849–861 (2015).

Porkka, M., Kummu, M., Siebert, S. & Flörke, M. The Role of Virtual Water Flows in Physical Water Scarcity: The Case of Central Asia. International Journal of Water Resources Development 28, 453–474 (2012).

Allan, J. A. Virtual Water: A Strategic Resource Global Solutions to Regional Deficits. Ground Water 36, 545–546 (1998).

Article   CAS   Google Scholar  

Carr, J. A., D’Odorico, P., Laio, F. & Ridolfi, L. Recent History and Geography of Virtual Water Trade. PLoS ONE 8, e55825 (2013).

Article   ADS   PubMed   PubMed Central   CAS   Google Scholar  

Hoekstra, A. Y. & Hung, P. Q. Globalisation of water resources: international virtual water flows in relation to crop trade. Global Environmental Change 15, 45–56 (2005).

Chapagain, A. K. & Hoekstra, A. Y. The global component of freshwater demand and supply: an assessment of virtual water flows between nations as a result of trade in agricultural and industrial products. Water International 33, 19–32 (2008).

Kumar, M. D. & Singh, O. P. Virtual Water in Global Food and Water Policy Making: Is There a Need for Rethinking? Water Resources Management 19, 759–789 (2005).

Wada, Y., Gleeson, T. & Esnault, L. Wedge approach to water stress. Nature Geosci 7, 615–617 (2014).

Gupta, J. & van der Zaag, P. Interbasin water transfers and integrated water resources management: Where engineering, science and politics interlock. Physics and Chemistry of the Earth, Parts A/B/C 33, 28–40 (2008).

Gohari, A., Eslamian, S., Mirchi, A., Abedi-Koupaei, J., Massah Bavani, A. & Madani, K. Water transfer as a solution to water shortage: A fix that can Backfire. Journal of Hydrology 491, 23–39 (2013).

Rockström, J. Water for food and nature in drought–prone tropics: vapour shift in rain–fed agriculture. Philos Trans R Soc Lond B Biol Sci 358, 1997–2009 (2003).

Article   PubMed   PubMed Central   Google Scholar  

Jägermeyr, J., Gerten, D., Schaphoff, S., Heinke, J., Lucht, W. & Rockström, J. Integrated crop water management might sustainably halve the global food gap. Environmental Research Letters 11, 025002 (2016).

Kummu, M., de Moel, H., Porkka, M., Siebert, S., Varis, O. & Ward, P. J. Lost food, wasted resources: global food supply chain losses and their impacts on freshwater, cropland, and fertiliser use. Sci Total Environ 438, 477–489 (2012).

Jalava, M., Guillaume, J. H. A., Kummu, M., Porkka, M., Siebert, S. & Varis, O. Diet change and food loss reduction: What is their combined impact on global water use and scarcity? Earth’s Future 4, 62–78 (2016).

Jalava, M., Kummu, M., Porkka, M., Siebert, S. & Varis, O. Diet change – a solution to reduce water use? Environmental Research Letters 9, 074016 (2014).

European Environment Agency. Towards efficient use of water resources in Europe. (2012).

Shao, X., Wang, H. & Wang, Z. Interbasin transfer projects and their implications: A China case study. International Journal of River Basin Management 1, 5–14 (2003).

Ma, J., Hoekstra, A. Y., Wang, H., Chapagain, A. K. & Wang, D. Virtual versus real water transfers within China. Philosophical Transactions of the Royal Society of London B: Biological Sciences 361, 835–842 (2006).

Article   PubMed   Google Scholar  

UNEP. Decoupling natural resource use and environmental impacts from economic growth. A Report of the Working Group on Decoupling to the International Resource Panel. Fischer-Kowalski, M., Swilling, M., von Weizsäcker, E.U., Ren, Y., Moriguchi, Y., Crane, W., Krausmann, F., Eisenmenger, N., Giljum, S., Hennicke, P., Romero Lankao, P., Siriban Manalang, A., Sewerin, S. United Nations Environment Programme (2011).

Hoff, H., Döll, P., Fader, M., Gerten, D., Hauser, S. & Siebert, S. Water footprints of cities–indicators for sustainable consumption and production. Hydrol Earth Syst Sci 18, 213–226 (2014).

Guillaume, J., Kummu, M., Eisner, S. & Varis, O. Transferable Principles for Managing the Nexus: Lessons from Historical Global Water Modelling of Central Asia. Water 7, 4200 (2015).

Aus der Beek, T., Voß, F. & Flörke, M. Modelling the impact of Global Change on the hydrological system of the Aral Sea basin. Physics and Chemistry of the Earth, Parts A/B/C 36, 684–695 (2011).

Shiklomanov, I. A. Appraisal and assessment of world water resources. Water International 25, 11–32 (2000).

Haddeland, I. et al. Multimodel Estimate of the Global Terrestrial Water Balance: Setup and First Results. J Hydrometeorol 12, 869–884 (2011).

Wada, Y. et al. Modeling global water use for the 21st century: the Water Futures and Solutions (WFaS) initiative and its approaches. Geosci Model Dev 9, 175–222 (2016).

Müller Schmied, H. et al. Sensitivity of simulated global-scale freshwater fluxes and storages to input data, hydrological model structure, human water use and calibration. Hydrol Earth Syst Sci 18, 3511–3538 (2014).

Döll, P., Douville, H., Güntner, A., Müller Schmied, H. & Wada, Y. Modelling Freshwater Resources at the Global Scale: Challenges and Prospects. Surveys in Geophysics 37, 195–221 (2016).

Wada, Y. et al. Multimodel projections and uncertainties of irrigation water demand under climate change. Geophysical Research Letters 40, 4626–4632 (2013).

Raworth, K. A Safe and Just Space for Humanity: Can we live within the doughnut. Oxfam Policy and Practice: Climate Change and Resilience 8, 1–26 (2012).

Hoekstra, A. Y., Mekonnen, M. M., Chapagain, A. K., Mathews, R. E. & Richter, B. D. Global Monthly Water Scarcity: Blue Water Footprints versus Blue Water Availability. PLoS ONE 7, e32688 (2012).

Kummu, M., Gerten, D., Heinke, J., Konzmann, M. & Varis, O. Climate-driven interannual variability of water scarcity in food production potential: a global analysis. Hydrol Earth Syst Sci 18, 447–461 (2014).

Salmivaara, A., Porkka, M., Kummu, M., Keskinen, M., Guillaume, J. H. A. & Varis, O. Exploring the Modifiable Areal Unit Problem in Spatial Water Assessments: A Case of Water Shortage in Monsoon Asia. Water 7, 898–917 (2015).

Jiang, J., Sharma, A., Sivakumar, B. & Wang, P. A global assessment of climate–water quality relationships in large rivers: An elasticity perspective. Sci Total Environ 468–469, 877–891 (2014).

Vörösmarty, C. J. et al. Global threats to human water security and river biodiversity. Nature 467, 555–561 (2010).

Wada, Y. & Bierkens, M. F. P. Sustainability of global water use: past reconstruction and future projections. Environmental Research Letters 9, 104003 (2014).

Pastor, A. V., Ludwig, F., Biemans, H., Hoff, H. & Kabat, P. Accounting for environmental flow requirements in global water assessments. Hydrol Earth Syst Sci 18, 5041–5059 (2014).

Porkka, M., Gerten, D., Schaphoff, S., Siebert, S. & Kummu, M. Causes and trends of water scarcity in food production. Environmental Research Letters 11, 015001 (2016).

Porkka, M., Kummu, M., Siebert, S. & Varis, O. From food insufficiency towards trade-dependency: a historical analysis of global food availability. PLoS ONE 8, e82714 (2013).

Yang, H., Wright, J. A. & Gundry, S. W. Water accessibility: Boost water safety in rural China. Nature 484, 318–318 (2012).

Kundzewicz, Z. W. et al. Freshwater resources and their management. In: Climate Change 2007: Impacts, Adaptation and Vulnerability . Contribution of the Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (2007).

Arnell, N. & Delaney, E. K. Adapting to climate change: Public water supply in England and Wales. Climatic Change 78, 227–255 (2006).

Wutich, A., White, A. C., White, D. D., Larson, K. L., Brewis, A. & Roberts, C. Hard paths, soft paths or no paths? Cross-cultural perceptions of water solutions. Hydrol Earth Syst Sci 18, 109–120 (2014).

Ohlsson, L. Water conflicts and social resource scarcity. Phys Chem Earth Pt B 25, 213–220 (2000).

Turton, A. R. Water scarcity and social adaptive capacity: towards an understanding of the social dynamics of water demand management in developing countries. MEWREW Occasional Paper No. 9, Water Issues Study Group, School of Oriental and African Studies (SOAS) (1999).

Cai, X. & Rosegrant, M. W. Global Water Demand and Supply Projections. Water International 27, 159–169 (2002).

De Fraiture, C. Integrated water and food analysis at the global and basin level. An application of WATERSIM. Water Resources Management 21, 185–198 (2007).

United Nations. United Nations World Macro Regions and Components. In: UN Map Library . United Nation (UN): http://www.un.org/depts/dhl/maplib/worldregions.htm , accessed December 2009 (2000).

Weedon, G. P. et al. Creation of the WATCH Forcing Data and Its Use to Assess Global and Regional Reference Crop Evaporation over Land during the Twentieth Century. J Hydrometeorol 12, 823–848 (2011).

Weedon, G. P., Balsamo, G., Bellouin, N., Gomes, S., Best, M. J. & Viterbo, P. The WFDEI meteorological forcing data set: WATCH Forcing Data methodology applied to ERA-Interim reanalysis data. Water Resour Res 50, 7505–7514 (2014).

Portmann, F. T., Siebert, S. & Döll, P. MIRCA2000-Global monthly irrigated and rainfed crop areas around the year 2000: A new high-resolution data set for agricultural and hydrological modeling. Global Biogeochemical Cycles 24, GB1011 (2010).

Palgrave Macmillan Ltd ed. International Historical Statistics. Palgrave Macmillan (2013).

Munia, H., Guillaume, J. H. A., Mirumachi, N., Porkka, M., Wada, Y. & Kummu, M. Water stress in global transboundary river basins: significance of upstream water use on downstream stress. Environmental Research Letters 11, 014002 (2016).

Falkenmark, M., Lundqvist, J. & Widstrand, C. Macro-scale water scarcity requires micro-scale approaches. Natural Resources Forum 13, 258–267 (1989).

Klein Goldewijk, K., Beusen, A. & Janssen, P. Long-term dynamic modeling of global population and built-up area in a spatially explicit way: HYDE 3.1. The Holocene 20, 565–573 (2010).

Download references

Acknowledgements

Study was funded by Academy of Finland project SCART (grant no. 267463), Emil Aaltonen foundation (‘eat-less-water’ project), Academy of Finland funded SRC project ‘Winland’, and Maa- ja vesitekniikan tuki ry . Additionally, P.J. Ward received funding from the Netherlands Organisation for Scientific Research (NWO) in the form of a VENI grant (grant no. 863-11-011) and T.I.E. Veldkamp from EU 7th Framework Programme through the projects ENHANCE (grant agreement no. 308438) and EartH2Observe (grant agreement no. 603608). Authors are grateful to Suvi Sojamo and Olli Varis for their comments and support.

Author information

Kummu M. and Guillaume J. H. A. contributed equally to this work.

Authors and Affiliations

Water & Development Research Group (WDRG), Aalto University, Espoo, Finland

M. Kummu, J. H. A. Guillaume & M. Porkka

National Centre for Groundwater Research and Training & Integrated Catchment Assessment and Management Centre, The Fenner School of Environment and Society, The Australian National University, Australia

J. H. A. Guillaume

Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, Amsterdam, The Netherlands

H. de Moel, T. I. E. Veldkamp & P. J. Ward

Center for Environmental Systems Research (CESR), University of Kassel, Germany

S. Eisner & M. Flörke

Institute of Crop Science and Resource Conservation (INRES), University of Bonn, Germany

You can also search for this author in PubMed   Google Scholar

Contributions

M.K., J.H.A.G., H.d.M., S.E., S.S. and P.J.W. designed this study in consultation with M.F. and T.I.E.V. The modelling was conducted by S.E. and M.F. supported by M.K., J.H.A.G., H.d.M. and M.P. Analyses were conducted by H.d.M., J.H.A.G. and M.K. in consultation with S.E., S.S. and P.J.W. Statistical analyses for trajectory classification were conducted by J.H.A.G. M.K., J.H.A.G., S.E. and T.I.E.V. wrote the article, with contributions from all co-authors.

Ethics declarations

Competing interests.

The authors declare no competing financial interests.

Electronic supplementary material

Supplementary information, supplementary dataset 1, supplementary dataset 2, rights and permissions.

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Reprints and permissions

About this article

Cite this article.

Kummu, M., Guillaume, J., de Moel, H. et al. The world’s road to water scarcity: shortage and stress in the 20th century and pathways towards sustainability. Sci Rep 6 , 38495 (2016). https://doi.org/10.1038/srep38495

Download citation

Received : 20 May 2016

Accepted : 11 November 2016

Published : 09 December 2016

DOI : https://doi.org/10.1038/srep38495

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Half of twenty-first century global irrigation expansion has been in water-stressed regions.

• Piyush Mehta
• Stefan Siebert
• Kyle Frankel Davis

Nature Water (2024)

Examination of regional water governance and water insecurity issues in Central Asia

• Young-Jin Ahn
• Zuhriddin Juraev

Sustainable Water Resources Management (2024)

Multi-objective optimal allocation of water resources based on improved marine predator algorithm and entropy weighting method

• Zhaocai Wang
• Haifeng Zhao

Earth Science Informatics (2024)

Habitability of low-lying socio-ecological systems under a changing climate

• Tom Spencer
• Alexandre K. Magnan
• Colette C. C. Wabnitz

Climatic Change (2024)

Geospatial analysis of soil resistivity and hydro-parameters for groundwater assessment

• Pradeep Kumar
• Karamat Ali

Discover Geoscience (2024)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.

Accessibility Links

• Skip to content
• Skip to search IOPscience
• Skip to Journals list
• Accessibility help
• Accessibility Help

Click here to close this panel.

Purpose-led Publishing is a coalition of three not-for-profit publishers in the field of physical sciences: AIP Publishing, the American Physical Society and IOP Publishing.

Together, as publishers that will always put purpose above profit, we have defined a set of industry standards that underpin high-quality, ethical scholarly communications.

We are proudly declaring that science is our only shareholder.

Global water scarcity including surface water quality and expansions of clean water technologies

Michelle T H van Vliet 1,2 , Edward R Jones 1 , Martina Flörke 3 , Wietse H P Franssen 2 , Naota Hanasaki 4 , Yoshihide Wada 5,1 and John R Yearsley 6

Published 26 January 2021 • © 2021 The Author(s). Published by IOP Publishing Ltd Environmental Research Letters , Volume 16 , Number 2 Citation Michelle T H van Vliet et al 2021 Environ. Res. Lett. 16 024020 DOI 10.1088/1748-9326/abbfc3

You need an eReader or compatible software to experience the benefits of the ePub3 file format .

Article metrics

35652 Total downloads

Share this article

Author e-mails.

[email protected]

Author affiliations

1 Department of Physical Geography, Utrecht University, P.O. Box 80.115, 3508 TC, Utrecht, The Netherlands

2 Water Systems and Global Change Group, Wageningen University, P.O. Box 47, 6700 AA, Wageningen, The Netherlands

3 Ruhr-Universität Bochum, Universitätsstr. 150, DE-44801, Bochum, Germany

4 National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Japan

5 International Institute for Applied Systems Analysis (IIASA), Schlossplatz 1, A-2361, Laxenburg, Austria

6 Department of Civil and Environmental Engineering, University of Washington, Seattle, WA, 98195, United States of America

Michelle T H van Vliet https://orcid.org/0000-0002-2597-8422

Naota Hanasaki https://orcid.org/0000-0002-5092-7563

Yoshihide Wada https://orcid.org/0000-0003-4770-2539

John R Yearsley https://orcid.org/0000-0002-2630-9589

• Received 5 May 2020
• Accepted 9 October 2020
• Published 26 January 2021

Peer review information

Method : Single-anonymous Revisions: 2 Screened for originality? Yes

Buy this article in print

Water scarcity threatens people in various regions, and has predominantly been studied from a water quantity perspective only. Here we show that global water scarcity is driven by both water quantity and water quality issues, and quantify expansions in clean water technologies (i.e. desalination and treated wastewater reuse) to 'reduce the number of people suffering from water scarcity' as urgently required by UN's Sustainable Development Goal 6. Including water quality (i.e. water temperature, salinity, organic pollution and nutrients) contributes to an increase in percentage of world's population currently suffering from severe water scarcity from an annual average of 30% (22%–35% monthly range; water quantity only) to 40% (31%–46%; both water quantity and quality). Water quality impacts are in particular high in severe water scarcity regions, such as in eastern China and India. In these regions, excessive sectoral water withdrawals do not only contribute to water scarcity from a water quantity perspective, but polluted return flows degrade water quality, exacerbating water scarcity. We show that expanding desalination (from 2.9 to 13.6 billion m 3 month −1 ) and treated wastewater uses (from 1.6 to 4.0 billion m 3 month −1 ) can strongly reduce water scarcity levels and the number of people affected, especially in Asia, although the side effects (e.g. brine, energy demand, economic costs) must be considered. The presented results have potential for follow-up integrated analyses accounting for technical and economic constraints of expanding desalination and treated wastewater reuse across the world.

Export citation and abstract BibTeX RIS

Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 license . Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

1. Introduction

A growing worldwide population strongly increases the demands for clean water for different sectoral water uses (e.g. irrigation, domestic, energy, manufacturing uses) (Biswas and Tortajada 2018 ). Climate change induced increases in the frequency and intensity of hydro-climatic extremes (e.g. droughts, floods) (Dankers et al 2014 , Prudhomme et al 2014 , Trenberth et al 2014 ), combined with increasing intensification of agriculture, industrialisation, urbanisation, and water extractions and uses, aggravate water quality deterioration, particularly in developing countries (Macdonald et al 2016 , UNEP 2016 , Sinha et al 2017 ). These changes will challenge sustainable management of 'clean accessible water for all', one of the UN Sustainable Development Goals (SDGs) for 2030 (UN 2015 ).

So far, water scarcity assessments have focussed mainly on water quantity (Schewe et al 2014 , Liu et al 2017 , Cui et al 2018 ). A widely used index of water scarcity or water stress considers the proportion of the freshwater use (withdrawal) relative to the available freshwater resources (Liu et al 2017 ). This indicator has been used for several scientific studies (Kummu et al 2016 , Liu et al 2017 , Vanham et al 2018 ) and is also presented as SDG-indicator 6.4.2 for estimating levels of water stress (UN 2015 ). Previous work by Vanham et al ( 2018 ) evaluated the shortcomings of this water scarcity indicator, including the absence of water quality, the lack of consideration of unconventional water resources, and the weak temporal (annual) and spatial (country) resolutions used in most water scarcity assessments (Vanham et al 2018 ).

It is imperative that we understand regional hotspots of water scarcity in terms of both water quantity and quality, as the usability of water for human purposes and ecosystem health depends on both sufficient water quantity and suitable water quality (van Vliet et al 2017 ). Earlier studies used the water poverty index combining water quantity and quality data (Sullivan et al 2003 ) or included water pollution drivers in calculations of threat indices to human water security and biodiversity (Vörösmarty et al 2010 ). For Chinese cities and river basins (Zhao et al 2016 , Liu et al 2017 ), the ratio of total water demands to freshwater availability has been combined with the grey water footprint (Hoekstra and Mekonnen 2012 ) (i.e. the amount of water required to dilute pollutants in wastewater to sufficiently meet environmental water quality standards). Following on this concept, an indicator of water scarcity has been developed including water quality requirements for different sectoral water uses (van Vliet et al 2017 ). However, an assessment of water scarcity accounting for water quality and clean water technologies is still lacking, in particular at the large scale.

Typically, water scarcity solutions focus on decreasing sectoral water uses (e.g. improved water use efficiencies) or by increasing water availability (e.g. increasing reservoir storage capacity). These solution options have been included in earlier water scarcity assessments (e.g. Ward et al 2010 , Wada et al 2014a , Jägermeyr et al 2015 ). Clean water technologies suiting both the water quantity and water quality demands, such as desalinated water use and treated wastewater reuse, are fast-growing (Elimelech and Phillip 2011 , Gude 2017 , Jones et al 2019 ). Both options are considered as a key component to reduce water pollution and freshwater scarcity globally (SDG targets 6.3 and 6.4) (UN 2015 ). While some first steps have been made to implement seawater desalination in water scarcity quantifications (Oki and Kanae 2006 , Hanasaki et al 2016 ), these assessments have ignored desalination of inland (brackish, river) waters and other sources (e.g. wastewater, brine), which together contribute to almost 40% of the desalination water use worldwide (Jones et al 2019 ). Furthermore, the desalination data used in previous studies did not consider sector-specific uses of desalinated water. Another previous study of Parkinson et al ( 2019 ) focused on improved water access, treatment and efficiencies towards the SDG6 targets using an integrated assessment modelling approach, but water quality conditions were disregarded.

Here we fill in the knowledge gap by presenting global hotspots of water scarcity driven by both surface water quantity and water quality issues, and quantify expansions in desalination and treated wastewater reuse to reduce the number of people suffering from water scarcity as required by UN's SDG6.

2.1. Water scarcity indicators and framework

We developed new indicators and a globally applicable model framework of water scarcity including a water demand versus supply dimension from both a surface water quantity and water quality perspective (figure 1 ). Our framework includes global gridded simulations at 0.5° × 0.5° spatial resolution of surface water availability and sectoral water use (section 2.2 ), surface water quality and sector water quality requirements (section 2.3 ) and spatially-explicit data of desalination and treated wastewater reuse capacities (section 2.4 ). We focus on these two water technologies, because they are fast-growing technologies suiting both water quantity and quality demands (Elimelech and Phillip 2011 , Gude 2017 ) and because they are considered as a key component to reduce freshwater scarcity globally (UN 2015 ). Next to this, wastewater treatment impacts are also included in the modelling of pollutant loadings and hence in surface water quality concentrations. The focus of this water scarcity assessment is on surface water resources, and with particular focus on impacts of surface water quality, desalination and treated wastewater reuse on water scarcity levels. We do not include groundwater resources in our study, given the current lack of a globally-applicable groundwater model accounting for both water quality and quantity.

Figure 1.  Water scarcity framework and three water scarcity indicators: 1. water scarcity based on only quantity (WS), 2. water scarcity including both water quantity and water quality (WSq); and 3. water scarcity based on both water quantity and quality, and including desalination and treated wastewater reuse (WSq _ desal + wwr ).

Download figure:

Water scarcity was quantified using three different water scarcity indicators: 1) water scarcity based on only surface water quantity (WS); 2) water scarcity including both surface water quantity and water quality for different sectoral uses (WSq); and 3) water scarcity based on both surface water quantity and quality, and including desalination of various sources (sea water, inland resources) and treated wastewater reuses per sector (WSq _ desal + wwr ). These three indicators are briefly discussed below and in more detail (including their equations) in the supplementary section 1 (available online at stacks.iop.org/ERL/16/024020/mmedia ).

Water scarcity based only on quantity (WS) is calculated as the ratio of water withdrawals for all sectors to surface water availability ('criticality ratio'), accounting also for environmental flow requirements (EFRs). The 'criticality ratio' is a widely use approach (Alcamo et al 2003b , Raskin et al 1997 , Liu et al 2017 , Vanham et al 2018 ), which has also been adopted by the UN SDG-indicator 6.4.2. (for details see supplementary section 1.1 and supplementary equation 1).

Water scarcity including both water quantity and water quality (WSq) is estimated by using the ratio of the sectoral water withdrawals of acceptable quality to the water availability. Under conditions that a water quality constituent i does not meet the sector water quality requirements for sector j ( C i > C max i,j ) we quantify the additional water needed to obtain acceptable water quality by dilution (dq i,j ). For this, we estimate for each water use sector the most critical water quality constituent resulting in the highest dilution water demands (dq max j ). Here we follow on from the concepts of grey water footprint and previous work (Hoekstra and Mekonnen 2012 , van Vliet et al 2017 , Ma et al 2020 ), but focussing on dilution of water withdrawn for specific-sector uses rather than total water availability. For thermoelectric water withdrawals, which largely depend on water temperature as the critical water quality constituent, we used spatially-explicit values of maximum permitted water temperature for cooling water use (see supplementary section 3.5, supplementary table 1). Where the water temperature exceeds permissible levels, we calculate the extra surface water withdrawal needed to dilute thermal effluents from power plants (i.e. dissipate the same waste heat). This is in line with the increase in water demands for power plant cooling under higher water temperatures as quantified in earlier work (Koch and Vögele 2009 , van Vliet et al 2012 ). Additional dilution water demands can also be estimated to obtain acceptable salinity, for instance, for irrigation uses. This dilution water suiting water quality for sectoral use can in principle originate from various sources (e.g. treated (waste) water or groundwater). In our global assessment we do however not specify the origin of these alternative water resource, but we quantify the potential dilution water requirement needed to obtain acceptable quality. This results in a calculated additional 'pressure' on the water system (higher water scarcity levels) in case water quality does not meet certain sectoral water quality requirements. Our water scarcity approach explicitly accounts for different quality requirements by different intended uses (for details see supplementary section 1.2 and supplementary equation 2).

We further developed this water scarcity indicator to account for spatially-explicit desalination uses and treated wastewater reuses. We distinguish between desalinated water from 'new' sources (i.e. beyond what is available from inland waters, e.g. seawater) and 'existing' (inland) sources (brackish water, river waters) per water use sector (see equation 1; for details see supplementary section 1.3). Both desalination and treated wastewater are subtracted from the sectoral water demand and hence also cause a reduction in the volume of water required for dilution of water to obtain an acceptable quality (dq).

Where WSq desal + wwr = water scarcity including water quality, desalination and treated wastewater reuse [−]; D = water withdrawal for sector j [m 3 s −1 ]; Q = water availability [m 3 s −1 ]; EFR = environmental flow requirements [m 3 s −1 ]; dq i,j = extra water withdrawals for dilution to obtain acceptable quality for sector j and water quality constituent i [m 3 s −1 ]; dq max j = maximum required water withdrawals for dilution to obtain acceptable quality for sector j based on the most critical water quality constituent [m 3 s −1 ]; N j = desalinated water of 'new' sources (e.g. seawater, brine) for sector j [m 3 s −1 ]; E j = desalinated water of 'existing' sources (inland brackish, river water resources) used for sector j [m 3 s −1 ]; W j = treated wastewater reuse for sector j [m 3 s −1 ]; C i = actual water quality level of water quality constituent i [unit depends on water quality constituent considered, e.g. mg l −1 for concentrations, °C for water temperature]; C max i,j = maximum water quality threshold for water quality constituent i for water use sector j [e.g. mg l −1 ,°C].

All water scarcity calculations are at 0.5° × 0.5° spatial resolution globally and with a monthly timestep, focussing on the period of 2000–2010. We identified water scarcity levels higher than 0.4 as 'severe water scarcity' in line with previous work (Liu et al 2017 , Hanasaki et al 2018 ), facilitating comparisons with previous studies. The average population under 'severe' water scarcity was quantified by combing our monthly water scarcity calculations with gridded (0.5°) population data (Goldewijk et al 2005 , 2010 ).

2.2. Global water resources and sectoral water use modelling

For the water quantity component of our water scarcity framework we used global gridded simulations of surface water availability (i.e. discharge), sectoral water use (i.e. withdrawal and consumption) at 0.5° × 0.5° spatial resolution and on a monthly time step for 1979–2010 from four global hydrological models: PCR-GLOBWB (van Beek et al 2011 , Wada et al 2011 , 2014b , Sutanudjaja et al 2018 ), H08 (Hanasaki et al 2008 ), WaterGAP2 (Döll et al 2003 , Alcamo et al 2003a , Flörke et al 2013 , Müller Schmied et al 2016 ) and VIC (Liang et al 1994 , Lohmann et al 1998 , Hamman et al 2018 , Droppers et al 2019 ). We used simulated actual water withdrawal and consumption for the main water use sectors irrigation, domestic, manufacturing and thermoelectric water uses. These global hydrological models were selected because of their ability to simulate both water availability and sectoral water use on a global scale. Multi-model mean results were calculated to account for uncertainties in water availability and sectoral water use (withdrawal and consumption) simulations. EFRs were calculated using the monthly variable flow method (Pastor et al 2014 ), on the multi-model average discharge. For further details and results on the global water resource and water use modelling we refer to supplementary section 2.

2.3. Global water quality modelling and sector water quality requirements

The water quality component of our water scarcity framework accounts for surface water temperature, salinity (total dissolved solids; TDS), organic pollution (biochemical oxygen demand; BOD) concentrations), total nitrogen (TN) and total phosphorous (TP) concentrations. These water quality constituents are selected because they are key in constraining different sector water uses and ecosystem health (Scheffer et al 2001 , von der Ohe and Liess 2004 , Dumont et al 2012 , Herbert et al 2015 ). In addition, most of these water quality constituents are also part of SDG indicator 6.3.2 ('Proportion of bodies of water with good ambient water quality').

We developed a process-based global gridded surface water quality model to simulate surface water temperature, salinity (TDS concentrations), and organic pollution (BOD concentrations) using the approaches described in supplementary section 3. This water quality model was applied on 0.5° × 0.5° spatial resolution globally and monthly timestep for the period 1979–2010. Simulated return flows from the global hydrological models per water use sector (supplementary section 2, supplementary figure 3) were calculated and used as input to estimate pollutant loadings for the surface water quality modelling. Thermoelectric return flows were used to simulate impacts of heat effluents from power plants on surface water temperature. For calculating TDS loadings, irrigation and manufacturing return flows, together with population numbers and TDS excretion rates were used. For organic pollution (BOD) loadings, we used manufacturing return flows, population and livestock numbers, and excretion rates per capita and livestock type (cattle, chickens, ducks, goats, pigs and sheep). Next to this, pollutant loadings were calculated including the country-based fractions of wastewater treatment types (primary, secondary, tertiary and advanced treatment) and removal efficiencies per pollutant and treatment level (for details see supplementary section 3). In addition, global grid-based (0.5° × 0.5°) simulations of in-stream concentrations of TN and TP were produced with the IMAGE-GNM model (Beusen et al 2015 , 2016 ) (supplementary section 3).

Model validation against observed surface water quality records show that the observed water quality conditions are represented realistically by the global surface water quality model (supplementary figures 5–10 and supplementary section 3 for more details). However, the station density and number of water quality measurements for model validation is low particularly in the relatively dry regions of the world (e.g. parts of Africa, Australia, Asia and Middle East). This limits analyses of the water quality model performances in those regions. Next to this, the uncertainties in the simulated water availability are also highest in particular in these dry regions (supplementary figure 4), which likely also results in higher uncertainties in simulated in-stream concentrations and water scarcity levels. The results of simulated water availability, water quality and water scarcity are therefore masked in the global maps for the very dry regions of the world (with surface water availability less than 1 m 3 s −1 ). These very dry regions contribute to less than 1.6% of the global population.

Water quality requirements for irrigation, domestic, manufacturing and thermoelectric water uses and for ecosystem health used in our water scarcity framework are derived per sector from international standards (supplementary section 3.5). The exception is water temperature for which standards were considered only for energy (thermoelectric) uses and ecosystem health. An overview of the selected water quality thresholds for all water use sectors and selected water quality constituents, as well as the corresponding sources is presented in supplementary table 1.

2.4. Desalination and treated wastewater reuse globally

Desalination capacity was derived from a global spatially-explicit desalination plant database (GWI 2019a ) including results of 15 906 operational desalination plants (supplementary section 4.1). This database accounts for desalinated water use per main water use sector (domestic, manufacturing, energy, irrigation) individually (supplementary figure 12). Our approach thus accounts for sector-specific desalination uses and considering different sources: seawater, inland (brackish, river) sources, brine and wastewater.

A global spatially-explicit dataset of wastewater reuse was developed on 0.5° × 0.5° spatial resolution by downscaling of country data of existing sources (AQUASTAT 2019 , GWI 2019b ) with total gridded population numbers (Klein Goldewijk 2005 , Klein Goldewijk et al 2010 ) as described in more detail in supplementary section 4.2 (supplementary figure 13).

2.5. Expansion in desalination and treated wastewater reuse towards water scarcity mitigation

We quantify how much expansion in desalination and treated wastewater reuse would potentially be required compared to current capacities to 'substantially reduce the number of people suffering from water scarcity' in line with SDG target 6.4. For this final part of the analyses, we focus on water scarcity levels below 0.2 as target towards water scarcity mitigation, because a water scarcity threshold of 0.2 has typically been set as a limit towards 'moderate water scarcity' in contrast to 0.4 representing 'severe water scarcity' (Liu et al 2017 , Hanasaki et al 2018 ). For the analyses of potential expansion in desalination and treated wastewater reuse towards water scarcity mitigation (in line SDG target 6.4) we therefore consider technological expansions aiming at water scarcity levels below 0.2 as an appropriate target. We calculated the required expansion in desalination capacity of both seawater and inland water resources and treated wastewater reuse volume needed under the present (2000–2010) levels to obtain water scarcity levels below 0.2 (WSq _ desal + wwr ≤ 0.2). We consider an increase in desalination capacity (from both sea water and inland surface water resources) required to fulfil sector demands for the domestic, manufacturing and energy sectors, which are the dominant users of desalinated water, accounting for 97% of the world's desalination capacity (Jones et al 2019 ). Expansion in sea water for desalination are constrained to locations proximate to the coastline (<100 km) where increases in seawater desalination are economically feasible and technically viable (Zhou 2005 ). In locations without ready access to seawater, expansions in desalination are assumed to be covered by existing inland water resources (e.g. river water, brackish water), and are constrained by available water resources in contrast to desalination expansion of seawater, which was considered as an 'unlimited source'. Expansion in treated wastewater reuse towards water scarcity reduction is used for the irrigation sector only, which is the dominant user in terms of treated wastewater (Qadir et al 2007 , WWAP 2017 , Zhang and Shen 2017 ). Expansion in treated wastewater reuse capacity were constrained by the available total wastewater produced per gridcell (for details see supplementary section 4.3, supplementary table 3).

3.1. Water scarcity hotspots driven by water quantity and water quality

Our results show that including water quality contributes to an increase in the percentage of the global population currently suffering from severe water scarcity from an annual average of 30% (22%–35% monthly range; only quantity) to 40% (31%–46%; including water quality) for 2000–2010. We focus here on water scarcity levels equal or higher than 0.4, which has typically been set as a limit towards 'severe water scarcity' (Liu et al 2017 , Hanasaki et al 2018 ). Water scarcity levels and hence the number of people affected differ per month, with the largest inter-annual variability in Australia and lowest in North America (figure 2 (a)). Water scarcity intensification by accounting for water quality occurs in South America and Africa, but also in particular in the severe water scarcity regions, such as in South and East Asia (India and China), Middle East, Southern Europe and Mexico (figures 2 (a) and (b)). In most of these water scarcity hotspots we find that water scarcity is driven by a combination of water quantity and water quality issues (figure 2 (c)). Here, excessive sectoral water withdrawals result in high water quantity-driven water scarcity, but polluted return flows degrade water quality, depending also on wastewater treatment efficiencies and capacities.

Figure 2.  Water scarcity driven by water quantity and water quality issues for 2000–2010. Impacts of water quality on global water scarcity levels including both water quantity and various water quality constituents (water temperature, salinity (TDS), organic pollution (BOD), total nitrogen and total phosphorous concentrations) and bar plots with percentage of population affected by severe water scarcity (a), percentage increase in water scarcity levels by including water quality compared to the approach focussing solely on water quantity (b) and identification of main drivers of water scarcity (water quantity, water quality or combined) (c). Regions with water availability less than 1 m 3 s −1 are masked (white).

While different water quality constituents (pollutants) may have different impacts in terms of constraints for sectoral uses, overall consistent water quality hotspots are identified for organic pollution, salinity and nutrients (figure 3 ). This is due to common pollution sources and contributing sectors, and is in line with previous large-scale water quality assessments covering multiple water quality constituents (Kroeze et al 2016 , UNEP 2016 , Strokal et al 2019 , van Vliet et al 2019 ). In particular in north-eastern China, but also in other parts of central Asia, and parts of the Mediterranean, western US and Mexico, are identified as water quality hotspots regions in terms of high salinity (TDS), organic pollution (BOD) and nutrients (TN, TP) concentrations.

Figure 3.  Global surface water quality hotspots. Average simulated in-stream concentrations presented for simulated organic pollution as indicated by biochemical oxygen demand (BOD) (a), salinity as indicated by total dissolved solids (TDS) (b), total nitrogen (TN) (c) and total phosphorous (TP) (d) concentrations. Regions with water availability less than 1 m 3 s −1 are masked (white). For details of water quality modelling and validation results see supplementary section 3, and supplementary figures 5–10.

Degraded water quality may result in unsuitability for sectoral water uses, exacerbating water scarcity levels. For the most severe water scarcity hotspots in the world, including eastern China and parts of central Asia, water scarcity is strongly water quality-driven (figure 2 (c)). Here, we see that water scarcity levels increase by an order of magnitude compared to water scarcity based on only water quantity. Exacerbation of water scarcity due to water pollution in China is in line with previous water scarcity studies for this region (Zhao et al 2016 , Ma et al 2020 ). In contrast, in regions with low sectoral water uses and pollution levels and with high water availability (e.g. tropical regions and high latitudes) we find that water scarcity levels, while being low, are mainly water-quantity driven (figure 2 (c)).

We identify for each water use sector and for freshwater ecosystems the main critical water quality constituent that has the strongest contribution in water scarcity intensification. This shows that increases in water scarcity for the energy (thermoelectric) sector are in particular driven by high water temperatures (figure 4 (a)), with higher water scarcity particularly in eastern China, India, US, Europe, and parts of Australia. High salinity (TDS) levels mainly constrain irrigation and manufacturing uses in most regions (figures 4 (b) and (c); for 72% and 77% of area for irrigation and manufacturing, respectively). Domestic water scarcity levels increase due to various critical water quality constituents; high organic pollution (BOD concentrations) in particular in eastern Asia, high TN mainly in Europe, and high salinity (TDS) levels mainly in arid regions (e.g. western US) (figure 4 (d)). Surface water quality thresholds for ecosystem health are exceeded in particular for BOD (45%) and phosphorous (42%) (figure 4 (e)).

Figure 4.  Main critical water quality constituents contributing to water scarcity for human uses and freshwater ecosystems in water scarcity regions. Maps presenting the water quality constituents that contributed strongest to increases in water scarcity for energy (a), irrigation (b), manufacturing (c), domestic uses (d), and freshwater ecosystems (e).

3.2. Reducing (clean) water scarcity

In a next step, the newly developed water scarcity indicator (WSq _ desal + wwr ) was used to calculate the potential expansion in desalination and treated wastewater reuse towards water scarcity alleviation. From a water resource perspective, the population under severe water scarcity (water scarcity levels equal or higher than 0.4) could potentially be reduced from 40% (31%–46%) to 14% (7.0%–16%) under maximum expansions (figure 5 (a)). This would require a worldwide increase in desalination capacity from 2.9 to 13.6 billion m 3 month −1 and an increase in treated wastewater reuse from 1.6 to 4.0 billion m 3 month −1 . In terms of treated wastewater reuse, the strongest increases are calculated for China and India, which together contribute to 60% of the required expansion in treated wastewater reuse worldwide (figures 5 (b) and 6 (a), (c)). In these densely populated regions, large amounts of wastewater is produced that could potentially be treated and reused to fulfil the high irrigation water demands in these regions. The highest potential expansion in desalination capacity towards water scarcity mitigation is quantified for the USA, China and India and several European countries contributing to water scarcity reduction for domestic and industrial uses (figures 5 (c) and 6 (b), (d)). In some regions, such as India, eastern Asia (China), but also parts of the USA, Europe and other regions, these calculated potential expansions in desalination and treated wastewater reuse are still insufficient to meet the sectoral water demands. This is due to limited available wastewater resources that can be treated and reused, a lack of close access to seawater or limited availability of inland surface resources for desalination. While saline or brackish groundwater desalination could potentially be used in some of these regions, it should be noted that this has been disregarded in our analyses.

Figure 5.  Impacts of expanding desalination and treated wastewater reuse and required capacity increase towards water scarcity mitigation. Reduction in population under severe water scarcity (a) and required mean expansion in treated wastewater reuse capacity (b) and desalination (c) towards water scarcity mitigation. Circular barplots (b), (c) show results for a selection of 30 countries with highest increase in required total expansion of desalination and treated wastewater capacity towards water scarcity mitigation (in million m 3 month −1 ) and situated in different world regions (NA = North America, SA = South America, EU = Europe, AF = Africa, AS = Asia and Russia, AU = Australia and Oceania).

Figure 6.  Required potential expansion (absolute and relative increase) in treated wastewater reuse (a), (c) and desalination (b), (d) towards water scarcity mitigation.

4. Discussion and conclusions

We developed a new global water scarcity framework including a water quality dimension and the use of fast-growing clean water technologies suiting both water quantity and quality aspects (i.e. desalination and treated wastewater re-use). Our global study shows that water scarcity levels and percentage of people affected by severe water scarcity are substantially higher when we account also for water quality (on average 40%) rather than solely water quantity (30%). Moreover, we show that water scarcity in most hotspots regions (India, China, Middle East, Mediterranean and Mexico) is driven by a combination of water quantity and water quality issues (figure 2 ). We show that desalinated water use and treated wastewater reuse can potentially strongly reduce the number of people affected by water scarcity (SDG target 6.4), especially in Asia (figures 5 and 6 ). These results are relevant in terms of defining water investment strategies and water resources exploitation potentials on a longer term (Cobbing and Hiller 2019 , Damania et al 2019 ). For instance, in regions where water scarcity is also strongly driven by water quality, investments in clean water technologies or pollution prevention measures (Damania et al 2019 ) would be recommended in addition to traditional measures focusing on water supply management (e.g. reservoir construction).

While our global assessment has been limited to a selection of water quality constituents (i.e. water temperature, salinity, organic pollution, total nitrogen and total phosphorous) relevant for various sectoral uses and ecosystem health, our water scarcity framework could potentially be used to add also other water quality constituents (e.g. pathogens, heavy metals, pesticides, pharmaceuticals and other emerging pollutants). This would require to expand the modelling of pollutant loadings and in-stream concentrations as well as the determination of suitable sectoral water quality standards for additional water quality constituents. The presented estimates of water quality impacts on water scarcity levels and number of people under severe water scarcity could then potentially increase, depending on whether sectoral water quality standards for those water quality constituents are exceeded.

It is important to note that our current water scarcity assessment is limited by the lack of global data for groundwater resources availability and water quality. Previous studies have highlighted a major role of groundwater resources availability and changes on water scarcity levels (Döll 2009 , Foster and Macdonald 2014 , Richey et al 2015 , Damkjaer and Taylor 2017 ). While the focus of our water scarcity assessment is on surface waters and the impacts of surface water quality and water technologies, the water scarcity concept developed in our study has the potential to include groundwater resources. This will provide a more comprehensive understanding of water scarcity, accounting for both surface and groundwater resources from both a water quantity and water quality perspective across different scales. Inclusion of groundwater resources would in particular be important in regions with a relative high contribution of water withdrawals from groundwater resources, such as India (Döll et al 2012 , Wada et al 2014b ), regions where the quality of groundwater resources is deteriorating (Macdonald et al 2016 , Burri et al 2019 , Gleeson et al 2020 ) or where there is a potential for increased sustainable groundwater exploitation, such as in Sub-Saharan Africa (Cobbing and Hiller 2019 ).

Our estimates of expansion in desalination potential towards water scarcity alleviation should be considered as lower bound estimates, as these do not consider potential increases in desalination from brackish groundwater resources. Our study thus identifies the physical boundaries towards water scarcity mitigation from a surface water resources perspective. Next to this, the technical, socio-economic and environmental constraints (Kümmerer et al 2018 ) and side-effects of these technologies must also be considered. Desalination and wastewater treatment are both energy intensive technologies, aggravating greenhouse gas emissions if provisioned from fossil fuels (Martin-Gorriz et al 2014 ), and associated with high economic costs (Parkinson et al 2019 ). The production of by-products also poses problems. For example, the 15 906 desalination plants considered in our study produce at present 4.3 billion m 3 month −1 of brine (i.e. hypersaline concentrate), in addition to the 2.9 billion m 3 month −1 of freshwater for water scarcity alleviation (Jones et al 2019 ). Few economically feasible and environmentally sound management strategies exist for the safe disposal of brine. When disposed back to the source (e.g. seawater), increased salinity and toxicity levels in the receiving body can pose major risks to aquatic ecosystems (Gacia et al 2007 , Palomar and Losada 2011 ). Disposal of brine to inland water resources can also paradoxically increase local water scarcity driven by salinity issues, constraining other sectoral water uses and aquatic ecosystems (Meneses et al 2010 ). Brine production is mainly driven by desalination technology and salinity of feedwater type used, and volumetrically is typically smaller for inland water resources than for desalination of seawater (Jones et al 2019 ). However, suitable and economically viable brine disposal management options are highly important for expanding desalination of inland water resources (Morillo et al 2014 ). For wastewater re-use, health concerns and public perceptions are also potential constraints, particularly for the domestic and irrigation sectors (WWAP 2017 ). Improvements in treatment technologies, coupling with renewable energy sources and resource recovery of 'waste' products provide opportunities for reducing the costs and environmental concerns associated with these technologies.

Achieving 'clean water for all' and 'reducing the number of people suffering from water scarcity', as advocated by SDG6, requires that we expand our focus from solely water quantity solutions (e.g. increasing water use efficiencies and reservoir storage), to measures that contribute to both water quantity and water quality improvements. Moreover, water quality improvements and water scarcity reduction should be sustainable without compromising environmental objectives. In addition to the 'hard infrastructure' clean water technologies, a strong focus on reducing the pollutant emissions (Kümmerer et al 2018 ) is also paramount in meeting the sustainable management of clean and sufficient water for all.

Acknowledgments

Dr Michelle van Vliet was financially supported by a VENI-grant (project no. 863.14.008) of NWO. The Global Environment Monitoring System is kindly acknowledged for supplying observed water quality data worldwide for global water quality model validation purpose. Dr Arthur Beusen is kindly acknowledged for sharing source code for the IMAGE-GNM global nutrient model.

Data availability statement

The data that support the findings of this study are available upon reasonable request from the authors.

Author contributions

MTHvV developed the study, performed the analyses and drafted the manuscript. EJ contributed to the implementation of desalination and wastewater reuse in the water scarcity assessment. WF, MF, NH, YW produced the global hydrological model results. WF and JRY contributed to the global water quality model development. All authors contributed to the manuscript.

Supplementary data

• < Previous

Home > Schools & Departments > HE > School of Social Sciences > Economics Dept. > Faculty Publications > 174

Economics Department Faculty Publications

Global water shortages: a philippines case study.

Rosalina Palanca-Tan , Ateneo de Manila University Follow

Document Type

Publication date.

Water constraint is a global problem that afflicts both developing and developed countries. More than a fourth, 2.1 out of 7.5 billion people worldwide still lack safely managed drinking water. This water shortage is the result of rapid population growth and poor governance that led to failure to put up the necessary water supply infrastructure and systems to meet the growing demand for water of a rapidly increasing population. In the Philippines, 9 million out of 101 million Filipinos still suffer from unimproved, unsafe and unsustainable water sources. Inadequate and intermittent water supply afflicts not only rural areas and low-income urban cities but also even the foremost urban center, Metro Manila. Governance issues and inefficiencies of government-managed water utilities result in poor access to adequate water both in terms of quantity and quality. Relying heavily on primary data collection methods — focus group discussion, key informant interview and a comprehensive household survey, this paper looks into the water provision aspect of the Philippine government’s housing program. Despite a stated imperative for adequate water provision in government housing program, failure of the designated water utility to put up the necessary water supply infrastructure that could meet the water requirements of the resident households results in several unmonitored small-scale water suppliers of highly-priced but poor quality water. The survey of household water purchases reveal that households buying water from the alternative small-scale water suppliers (neighbors with jetmatic pump wells and water tankers) pay more than 5-times but consume just about half of the consumption of those already served by the water utility, reflecting an overly constrained consumption that has wide-ranging health and well-being implications.

Recommended Citation

Palanca-Tan, R. (2020). Global water shortages: A Philippines case study. The Journal of Social, Political, and Economic Studies, 45(1/2), 46–62. https://www.proquest.com/docview/2395271609?pq-origsite=gscholar&fromopenview=true

Since March 01, 2022

Advanced Search

• Notify me via email or RSS
• Collections
• Libraries & Archives
• Ateneo Journals
• Disciplines

Author Corner

• Why contribute?
• Getting started
• Working with publishers and Open Access
• Copyright and intellectual property
• Contibutor FAQ

About Archium

• License agreement
• University website
• University libraries

Home About Help My Account Accessibility Statement

Privacy & Data Protection Copyright

Suggestions or feedback?

MIT News | Massachusetts Institute of Technology

• Machine learning
• Social justice
• Black holes
• Classes and programs

Departments

• Aeronautics and Astronautics
• Brain and Cognitive Sciences
• Architecture
• Political Science
• Mechanical Engineering

Centers, Labs, & Programs

• Abdul Latif Jameel Poverty Action Lab (J-PAL)
• Picower Institute for Learning and Memory
• Lincoln Laboratory
• School of Architecture + Planning
• School of Engineering
• School of Humanities, Arts, and Social Sciences
• Sloan School of Management
• School of Science
• MIT Schwarzman College of Computing

Study: Heavy snowfall and rain may contribute to some earthquakes

Press contact :.

Previous image Next image

When scientists look for an earthquake’s cause, their search often starts underground. As centuries of seismic studies have made clear, it’s the collision of tectonic plates and the movement of subsurface faults and fissures that primarily trigger a temblor.

But MIT scientists have now found that certain weather events may also play a role in setting off some quakes.

In a study appearing today in Science Advances , the researchers report that episodes of heavy snowfall and rain likely contributed to a swarm of earthquakes over the past several years in northern Japan. The study is the first to show that climate conditions could initiate some quakes.

“We see that snowfall and other environmental loading at the surface impacts the stress state underground, and the timing of intense precipitation events is well-correlated with the start of this earthquake swarm,” says study author William Frank, an assistant professor in MIT’s Department of Earth, Atmospheric and Planetary Sciences (EAPS). “So, climate obviously has an impact on the response of the solid earth, and part of that response is earthquakes.”

The new study focuses on a series of ongoing earthquakes in Japan’s Noto Peninsula. The team discovered that seismic activity in the region is surprisingly synchronized with certain changes in underground pressure, and that those changes are influenced by seasonal patterns of snowfall and precipitation. The scientists suspect that this new connection between quakes and climate may not be unique to Japan and could play a role in shaking up other parts of the world.

Looking to the future, they predict that the climate’s influence on earthquakes could be more pronounced with global warming.

“If we’re going into a climate that’s changing, with more extreme precipitation events, and we expect a redistribution of water in the atmosphere, oceans, and continents, that will change how the Earth’s crust is loaded,” Frank adds. “That will have an impact for sure, and it’s a link we could further explore.”

The study’s lead author is former MIT research associate Qing-Yu Wang (now at Grenoble Alpes University), and also includes EAPS postdoc Xin Cui, Yang Lu of the University of Vienna, Takashi Hirose of Tohoku University, and Kazushige Obara of the University of Tokyo.

Seismic speed

Since late 2020, hundreds of small earthquakes have shaken up Japan’s Noto Peninsula — a finger of land that curves north from the country’s main island into the Sea of Japan. Unlike a typical earthquake sequence, which begins as a main shock that gives way to a series of aftershocks before dying out, Noto’s seismic activity is an “earthquake swarm” — a pattern of multiple, ongoing quakes with no obvious main shock, or seismic trigger.

The MIT team, along with their colleagues in Japan, aimed to spot any patterns in the swarm that would explain the persistent quakes. They started by looking through the Japanese Meteorological Agency’s catalog of earthquakes that provides data on seismic activity throughout the country over time. They focused on quakes in the Noto Peninsula over the last 11 years, during which the region has experienced episodic earthquake activity, including the most recent swarm.

With seismic data from the catalog, the team counted the number of seismic events that occurred in the region over time, and found that the timing of quakes prior to 2020 appeared sporadic and unrelated, compared to late 2020, when earthquakes grew more intense and clustered in time, signaling the start of the swarm, with quakes that are correlated in some way.

The scientists then looked to a second dataset of seismic measurements taken by monitoring stations over the same 11-year period. Each station continuously records any displacement, or local shaking that occurs. The shaking from one station to another can give scientists an idea of how fast a seismic wave travels between stations. This “seismic velocity” is related to the structure of the Earth through which the seismic wave is traveling. Wang used the station measurements to calculate the seismic velocity between every station in and around Noto over the last 11 years.

The researchers generated an evolving picture of seismic velocity beneath the Noto Peninsula and observed a surprising pattern: In 2020, around when the earthquake swarm is thought to have begun, changes in seismic velocity appeared to be synchronized with the seasons.

“We then had to explain why we were observing this seasonal variation,” Frank says.

Snow pressure

The team wondered whether environmental changes from season to season could influence the underlying structure of the Earth in a way that would set off an earthquake swarm. Specifically, they looked at how seasonal precipitation would affect the underground “pore fluid pressure” — the amount of pressure that fluids in the Earth’s cracks and fissures exert within the bedrock.

“When it rains or snows, that adds weight, which increases pore pressure, which allows seismic waves to travel through slower,” Frank explains. “When all that weight is removed, through evaporation or runoff, all of a sudden, that pore pressure decreases and seismic waves are faster.”

Wang and Cui developed a hydromechanical model of the Noto Peninsula to simulate the underlying pore pressure over the last 11 years in response to seasonal changes in precipitation. They fed into the model meteorological data from this same period, including measurements of daily snow, rainfall, and sea-level changes. From their model, they were able to track changes in excess pore pressure beneath the Noto Peninsula, before and during the earthquake swarm. They then compared this timeline of evolving pore pressure with their evolving picture of seismic velocity.

“We had seismic velocity observations, and we had the model of excess pore pressure, and when we overlapped them, we saw they just fit extremely well,” Frank says.

In particular, they found that when they included snowfall data, and especially, extreme snowfall events, the fit between the model and observations was stronger than if they only considered rainfall and other events. In other words, the ongoing earthquake swarm that Noto residents have been experiencing can be explained in part by seasonal precipitation, and particularly, heavy snowfall events.

“We can see that the timing of these earthquakes lines up extremely well with multiple times where we see intense snowfall,” Frank says. “It’s well-correlated with earthquake activity. And we think there’s a physical link between the two.”

The researchers suspect that heavy snowfall and similar extreme precipitation could play a role in earthquakes elsewhere, though they emphasize that the primary trigger will always originate underground.

“When we first want to understand how earthquakes work, we look to plate tectonics, because that is and will always be the number one reason why an earthquake happens,” Frank says. “But, what are the other things that could affect when and how an earthquake happens? That’s when you start to go to second-order controlling factors, and the climate is obviously one of those.”

This research was supported, in part, by the National Science Foundation.

Share this news article on:

Related links.

• William Frank
• Department of Earth, Atmospheric and Planetary Sciences

Related Topics

• Earthquakes
• Earth and atmospheric sciences
• Environment
• Fluid dynamics
• Climate change
• Natural disasters
• Planetary science

Related Articles

Q&A: What sets the recent Japan earthquake apart from others?

Boom, crackle, pop: Sounds of Earth’s crust

Shaking up earthquake research at MIT

3 Questions: Understanding the Haiti earthquakes

Previous item Next item

More MIT News

Professor Emeritus David Lanning, nuclear engineer and key contributor to the MIT Reactor, dies at 96

Read full story →

Discovering community and cultural connections

MIT Supply Chain Management Program earns top honors in three 2024 rankings

New treatment could reverse hair loss caused by an autoimmune skin disease

How AI might shape LGBTQIA+ advocacy

Two MIT PhD students awarded J-WAFS fellowships for their research on water

• More news on MIT News homepage →

Massachusetts Institute of Technology 77 Massachusetts Avenue, Cambridge, MA, USA

• Map (opens in new window)
• Events (opens in new window)
• People (opens in new window)
• Careers (opens in new window)
• Accessibility
• Social Media Hub
• MIT on Facebook
• MIT on YouTube
• MIT on Instagram

Alleviating Worker Shortages Through Targeted Subsidies: Evidence from Incentive Payments in Healthcare

Worker shortages are common in many industries. This paper examines the effect of government subsidies to address these shortages in the context of a reform that tied Medicaid payments to nursing home staffing levels. We find that the reform substantially increased staffing, especially for facilities serving many Medicaid patients. Facilities responded primarily by hiring workers in lower-wage roles rather than increasing hours of incumbent or high-wage staff. This contrasts with null effects we estimate for a non-incentivized rate increase, suggesting that the incentive structure of government payments—rather than just the level—is key to boosting employment in sectors facing worker shortages.

We are very grateful to Andy Allison, David Grabowski, Brian McGarry, Melanie Wasserman, Maria Zhu, and participants at the Whistler Junior(ish) Health Economics Summit for helpful comments. Kevin Wang provided excellent research assistance. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.

I have received external research funding exceeding $10,000 during the last three years from Arnold Ventures, Washington Center for Equitable Growth, and PALTC Foundation. I received payments exceeding $10,000 from the NIA IMPACT Collaboration through Brown University for research related to nursing homes. I also received research funding exceeding $10,000 within the last three years from the National Institute on Aging. I also have received internal funding in excess of $10,000 from internal UCLA grants. To my knowledge, these organizations do not have any interests in the findings of this research. Moreover, they do not have any right to review the manuscript. In 2023, the Illinois Department of Healthcare and Family Services paid me $3,500 to participate as one of multiple academics advising on a proposed grant application for healthcare workforce collaboratives. This grant program was unrelated the paper.

MARC RIS BibTeΧ

Download Citation Data

• data appendix

More from NBER

In addition to working papers , the NBER disseminates affiliates’ latest findings through a range of free periodicals — the NBER Reporter , the NBER Digest , the Bulletin on Retirement and Disability , the Bulletin on Health , and the Bulletin on Entrepreneurship  — as well as online conference reports , video lectures , and interviews .

SA’s Water shortage to get worse in 2025

A S if load shedding was not enough to bring businesses to its knees, South Africans are now confronted with water shortages, with the situation expected to get worse in 2025.

A December 2023 report by the Department of Water and Sanitation (DWS) found that several water supply systems were operating close to or beyond their design capacity, and monitoring and compliance were severely deficient.

The report also stated that this made fixing problems impossible as the scale of the issues at stake was not being identified.

But this was nothing new, as ESI Africa reported in 2020 that: “South Africa is approaching physical water scarcity in 2025 where the country is expected to experience a water deficit of 17% by 2030, and climate change will worsen the situation.”

The effects of the water deficit are now being felt by Johannesburg residents who sometimes go a few days without water.

According to the DWS: “South Africa’s water security is threatened by a decrease in water supply due to negative impact on yields arising from climate change, degradation of wetlands and water resources, siltation of dams, whilst water losses and demand are escalating due to population and economic growth, urbanisation, inefficient use, and changing lifestyles.”

However, the Development Bank of Southern Africa (DBSA) stated in their recent report that one of South Africa’s most prominent water issues was that most people don’t have enough knowledge on how to preserve it.

Research carried out by the Institute for Security Studies found that: “South Africans use more water than the global average. South Africans currently use 234 litres of water per person daily, and the country’s per capita water consumption is higher than the global average of 173 litres. South Africans need to learn how to conserve water if they wish to avoid water scarcity.

“This can be done through tiered pricing, where users are charged when they consume a higher rate than what is considered necessary for daily activities. Other ways include having incentives for consumers to consider purchasing water-efficient appliances and go above and beyond to find ways to use less water,” read the DBSA report.

The DWS 2023 audit report found that the quality of the country’s drinkable water was getting worse. Nearly half (46%) of all water supply systems pose acute human health risks because of bacteria or other pathogens in the drinking water supply.

The report also found that more than two thirds (67.6%) of all wastewater treatment works are close to failure. On top of this it showed that over 47% of all clean and treated water was lost through leaks, or could not be accounted for.

“Water supply systems are in “poor and critical condition”. Almost half of all water supply systems (46%) do not comply with microbiological standards. In these water supply systems, drinking water is contaminated by sewage and bacteria. Viruses and parasites such as Legionella and Cyanobacteria may have grown in the piped water systems and or water sources,” read the report.

Making it worse, the report highlighted that more than half of the country’s municipalities (57%) do not notify water users when they discover that the water has been contaminated.

This placed citizens at risk of contracting water-borne illnesses and is an unacceptable practice due to the possible serious health repercussions of drinking contaminated water.

Associate Professor and Water Management Expert Anja du Plesis from Unisa stated on The Conversation that “The poor drinking water quality, lack of monitoring and unaccountability needed to receive immediate attention because of the human health risks involved.

“We cannot afford another tragic case such as Hammanskraal in South Africa’s Gauteng province, where 31 people died of cholera in May 2023 after drinking contaminated municipal water,” wrote Du Plessis.

Approached for comment, DWS spokesperson Kamogelo Magotsi said the department would respond to questions from the publication, but had not done so by the time of publication.

Meanwhile, the Ministry of Water and Sanitation said on Thursday that it expected to meet with Gauteng municipalities to further engage on plans to address water pollution on Monday next week in Pretoria.

“Minister of Water and Sanitation, Senzo Mchunu, together with Deputy Ministers David Mahlobo and Judith Tshabalala will meet with the Cities of Ekurhuleni, Johannesburg and Tshwane, as well as Mogale City Local Municipality to engage on the plans to address pollution affecting the Upper Crocodile and Upper Vaal Rivers,” read the statement.

The ministry stated that the discussions would address the performance and capacity of the municipalities’ wastewater treatment works, general catchment management that includes waste and stormwater management that have an impact on water quality.

[email protected]

Journal of Materials Chemistry A

Constructing z-scheme wo3/c3n4 heterojunctions with enlarged internal electric field and accelerated water oxidation kinetics for robust co2 photoreduction.

The solar-driven conversion of CO2 into value-added fuels is regarded as one of the most promising strategies to address the increasing greenhouse effect and energy crisis. However, severe charge recombination and sluggish H2O oxidation kinetics lead to its low efficiency. Herein, a Z-scheme WO3/C3N4 heterojunction was constructed to overcome these issues, with C3N4 nanosheets (NS) and WO3 NS acting as CO2 reduction center and H2O oxidation booster, respectively. Driven by the enlarged internal electric field, charge separation and migration in WO3/C3N4 heterojunction are promoted. Meanwhile, the WO3 NS provides highly oxidative photoinduced holes to accelerate H2O oxidation kinetics. Due to these advantages, the CO and CH4 yields of the optimal sample reached 9.4 μmol g-1 h-1 and 2.6 μmol g-1 h-1, respectively, representing approximately 4.5 times and 7.2 times that of the pristine C3N4 NS. Overall, this work contributes to an in-depth understanding of Z-scheme mechanism and offers a feasible approach to simultaneously modulate the charge separation and surface reaction for enhancing the efficiency of photocatalytic CO2 reduction.

• This article is part of the themed collection: Journal of Materials Chemistry A HOT Papers

Supplementary files

• Supplementary information PDF (2414K)

Article information

Download citation, permissions.

Z. Song, Q. Chen, Z. Sun, K. Chang, Z. Xie and Q. Kuang, J. Mater. Chem. A , 2024, Accepted Manuscript , DOI: 10.1039/D4TA01795A

To request permission to reproduce material from this article, please go to the Copyright Clearance Center request page .

If you are an author contributing to an RSC publication, you do not need to request permission provided correct acknowledgement is given.

If you are the author of this article, you do not need to request permission to reproduce figures and diagrams provided correct acknowledgement is given. If you want to reproduce the whole article in a third-party publication (excluding your thesis/dissertation for which permission is not required) please go to the Copyright Clearance Center request page .

Read more about how to correctly acknowledge RSC content .

Social activity

Search articles by author.

This article has not yet been cited.

Advertisements

Thursday, May 9, 2024 81° Today's Paper

Red Hill fuel spread through water quickly, expert testifies

By Kevin Knodell

May 1, 2024

Editors' Picks Red Hill Water Crisis

• Share on Facebook
• Share by email

Select an option below to continue reading this premium story.

Already a Honolulu Star-Advertiser subscriber? Log in now to continue reading.

Get unlimited access

From as low as $12.95 /mo.

Water system experts testified Tuesday on the second day of a class-action lawsuit against the U.S. government that tests may have missed contamination in the critical early days of the Red Hill fuel crisis.

The trial in federal court in Honolulu involves the first 17 plaintiffs claiming medical, emotional and financial injuries from the contamination. Another roughly 7,500 plaintiffs have joined other lawsuits also seeking compensation. The crisis began in November 2021 when jet fuel from the Navy’s underground Red Hill fuel storage facility tainted the Navy’s Oahu water system, which serves 93,000 people including service members, military families and civilians living in former military housing.

Government lawyers have argued that the amount of jet fuel that entered the water system was relatively low — too low to make most people sick — and that many people who reported symptoms weren’t exposed to fuel at all. They argue the ailments reported by thousands of people were stress-induced from hearing about contaminated water rather than from the contaminated water itself.

But Joseph Hughes, an engineering professor at Drexel University in Philadelphia, testified Tuesday as a witness for the plaintiffs and argued that contamination levels could be higher than the federal government acknowledges and that it would have spread quickly through the system.

On May 6, 2021, at least 20,000 gallons of JP-5 jet fuel spilled from one of Red Hill’s underground storage tanks, entering the facility’s fire suppression system. On Nov. 21, 2021, a worker driving a cart accidentally ruptured a pipe in the system, spilling the fuel in a tunnel near the Red Hill water well, which served the Navy waterline.

Navy officials responded and believed that they had contained the spill, and decided not to tell their superiors or state regulators. In the coming days, people on the waterline began reporting smelling fuel in the water and experiencing wide-ranging symptoms such as coughs, rashes and vomiting. The Navy shut off the well from its system on Nov. 28, 2021, and during the first months of 2022, the entire system was flushed.

Hughes testified that there were two types of testing carried out by the Navy after the November spill — “low flow” sampling and “bailer” sampling. Low-flow sampling is designed to take samples out of the water with minimal disruption to the water itself in an effort to minimize outside influences. Bailer sampling involves putting a sampling tube in the water and trapping a sample.

Hughes said that no bailer samples were taken from the Navy’s water system before the Navy shut down the Red Hill water well Nov. 29. He said that means potential contamination may have been missed because jet fuel is a light nonaqueous phase liquid — or LNAP — meaning it is less dense than water and tends to rise to the surface.

Hughes argued that low-flow sampling could have missed fuel closer to the water’s surface, which would have been captured by bailer sampling. He also said the higher concentrations likely would have been found in samples that did find contamination if those tests had been conducted earlier.

Under cross-examination from the government’s attorneys, Hughes acknowledged that not all the fuel that spilled into the Red Hill water well made it into the Navy’s drinking water system. The lawyers cited the accounts of Navy divers who were sent into the well as part of the response who reported that fuel was still seeping from the lava rock into the water well after the Navy shut it off from its water distribution system.

But Hughes said that based on his modeling of government data, he believes over 2,000 gallons of fuel entered the system from the contaminated Red Hill well before it was shut off. He told the court that “a little bit of volume is a lot of mass” when it comes to contamination of this sort.

Government lawyers questioned his credentials, noting he is neither a medical doctor nor epidemiologist, and that he was only analyzing data and wasn’t involved in any of the testing itself.

During the first and second day of the trial, Patrick Feidnt, a plaintiff in the trial, and his wife, Army Maj. Mandy Feidnt, testified about the symptoms they experienced. The government’s lawyers asserted that their home on Ford Island was never exposed to fuel and presented test results from their home. When presenting it to Mandy Feidnt, she testified it was the first time she had ever seen the document despite asking for it for years.

The government’s lawyers maintain that their symptoms during the crisis — including severe rashes, vomiting, diarrhea and other ailments — were all the result of preexisting conditions or stress-induced and cannot be tied to JP-5 fuel.

The Feidnts described the symptoms their young children suffered, including their daughter who still suffers neurological and psychological problems. They acknowledged that she had developmental problems from before the spill, but said the exposure to the fuel and the trauma it caused has exacerbated behavioral problems. They testified that their daughter now fixates on water and is scared of it, associating nearly all illnesses with drinking water.

The trial will continue through next week. Outside the courthouse Tuesday, the state Department of Health announced that it has rejected the Navy’s Red Hill Consolidated Groundwater Sampling Program as the military works to close the facility. The Navy has been monitoring groundwater for fuel while dumping treated water from the Red Hill well into Halawa Stream.

In a media release the DOH said it ordered the Navy “to implement a modified version of the program, which includes additional conditions relating to groundwater monitoring to support the Navy’s request to reduce the amount of water being discharged into Halawa Stream from 4.5 million gallons a day to 1.8 million gallons a day.”

“We have worked to hold the Navy accountable to create its own plan to reduce the amount of water discharged into Halawa Stream, while still maintaining a robust sampling program to track and stay ahead of any contaminant migration,” said Deputy Director for Environmental Health Kathleen Ho in a statement. “It’s unacceptable that the Navy would fail to produce such a plan following multiple rounds of regulatory review. That’s why we are taking the unusual step to reject the Navy’s plan and impose conditions necessary to protect public health and our environment.”

Understanding Australia’s teacher shortage: the importance of psychosocial working conditions to turnover intentions

• Open access
• Published: 08 May 2024

Cite this article

You have full access to this open access article

• Mark Rahimi   ORCID: orcid.org/0000-0002-9278-3537 1 &
• Ben Arnold   ORCID: orcid.org/0000-0003-2482-7611 1  

15 Accesses

Explore all metrics

Australian policymakers are currently attempting to address an ‘unprecedented teacher shortage’. Through a survey of 905 teachers in Australian government schools, this paper examines some of the key factors influencing the career intentions of teachers in Australian government schools. Drawing upon the concept of the psychosocial work environment from the field of occupational health, this analysis examines the relationship between key workplace demands, workplace resources, teachers’ experiences and attitudes towards work, teachers’ mental health outcomes, and their intentions to either remain in or leave their roles. The results reveal significant relationships between teachers’ intentions to leave their roles and challenging working conditions, adverse work experiences, as well as heightened levels of stress, burnout, and depressive symptoms. In contrast, remaining in their role was associated with factors indicative of a supportive psychosocial work environment, such as job recognition, trust in management, organisational justice, positive work experiences, and lower levels of mental health difficulties. The study underscores the urgent need to understand and assess the psychosocial work environments of teachers, and for multi-level strategies that address both protective and risk factors.

Avoid common mistakes on your manuscript.

Introduction

In the past 18 months of post-Covid, concerns about a critical teacher shortage associated with challenging working conditions and the status of the profession have emerged in Australia (Australian Government, 2023 ). Recent evidence indicates that most Australian teachers are either unsure about whether they will remain in the profession for their entire career or are intending to leave the profession prior to retirement (AITSL, 2022 ; Heffernan et al., 2022 ). For example, the proportion of Australian classroom teachers reporting that they plan to leave the profession before retirement has increased dramatically over recent years, from less than a quarter (22.25%) in 2020 to more than a third (34.21%) in 2022 (AITSL, 2022 ).

Over the past decade, the Australian government has implemented various policies to address teacher retention in the school teaching workforce. These policies have focussed on initiatives such as increased professional development opportunities, competitive remuneration packages, and enhanced support for early-career teachers through mentorship programs (Australian Government, 2023 ). More recently, strategies to recognise and reward exceptional teaching performance, efforts to reduce workload, along with measures to enhance the status and the perception of the teaching profession, have been important to the government’s approach (Australian Government, 2023 ).

In the context of growing concern about the sustainability of the Australian government school teaching workforce, understanding why teachers leave and factors related to turnover is crucial. Teacher turnover negatively impacts students, schools, and the education system. When teachers leave, it impacts student achievement, the quality of instruction, and the continuity of the learning experience (Ingersoll & Strong, 2011 ). Schools suffer due to the loss of institutional knowledge (Ingersoll & Strong, 2011 ) and increased organisational instability (Borman & Dowling, 2008 ), while school systems are affected by the substantial costs associated with recruiting and training new teachers (Ingersoll & Strong, 2011 ).

There are many different accounts of why teachers leave the profession, ranging from issues associated with education policies and the status of the teaching profession, such as standardisation, performance cultures, and systems of accountability (Nguyen et al., 2019 ; Perryman & Calvert, 2020 ), to school policies, conditions, and characteristics, including heavy workloads and the emotionally demanding nature of the work (Heffernan et al., 2022 ). Student, school, and teacher characteristics, teacher satisfaction, morale, and motivation are also factors (Madigan & Kim, 2021 ), as well as teacher mental health and wellbeing (Madigan & Kim, 2021 ). In this paper, focussing on teachers’ intentions to leave the profession, we present the findings from a national survey on Australian government school teachers’ work, health and wellbeing conducted in 2023. The survey focussed on teachers’ working conditions, experiences of work, teacher safety, teacher health, and future career intentions. We investigated this issue using the concept of the psychosocial work environment.

This study specifically focusses on teachers in Australia who work in government schools, a sector that caters to the majority of the student population (64.5%). While this study presents findings at a national level, it is important to note that education in Australia is primarily the responsibility of state governments, and variations in teachers’ working conditions and experiences of work are likely to vary within and between states.

The psychosocial work environment

The concept of the psychosocial work environment is important for understanding how the status of the teaching profession and their working conditions, influence their work experiences and their overall health—whether positively or negatively (Dicke et al., 2018 ; Riley et al., 2020 ). The psychosocial work environment is a concept from the fields of occupational health and psychology. It encompasses the social, psychological, and organisational factors that influence a worker’s experience of work and its impact on their mental health (Aust, et al., 2023 ). This concept includes organisational conditions, the culture of the organisation, interactions among individuals in a workplace, and the impact of an individual’s work on the employee health.

Rugulies ( 2019 ) describes the psychosocial work environment as an intermediate step in a causal pathway linking economic, social, and political structures with health and illness through psychological and psychophysiological processes. Adapting Rugulies’ framework for the teaching profession, we developed a conceptual framework for research into teachers’ psychosocial work environments and health that depicts the relation of the psychosocial work environment to phenomena at both the social and individual level (see Fig.  1 ).

Teachers psychosocial working conditions (adapted from Rugulies, 2019 )

Research on the psychosocial work environment goes beyond individual experiences of work, exploring macro- and meso-level structures that define and shape it. Understanding the impact of societal factors and work conditions on teacher health and career outcomes is crucial, highlighting the key role of the psychosocial work environment. It encompasses macro- and meso-level structures that define and shape the psychosocial work environment. Consequently, the psychosocial work environment is key for understanding how societal factors and work conditions affect teacher health and career outcomes. This research sheds light on how work impacts teachers’ wellbeing and job decisions.

The pathway begins with (i) broader societal factors, encompassing economic, social, and political structures like modes of production, division of teachers’ labour, school funding, school systems, and legal frameworks. These societal factors interact with (ii) workplace structures at an intermediate level, including the status and conditions of Australian teaching profession, such as job security, contracts, workforce supply, the authority of the profession, respect for the profession, and so on. These workplace structures, in turn, (iii) influence the psychosocial working conditions at the meso-level, such as job demands, work organisation, task content, and workplace relationships. These psychosocial working conditions are then encountered and processed by (iv) individuals or groups, triggering cognitive and emotional responses. As a result of these cognitive and emotional processes, the psychosocial working conditions subsequently give rise to (v) psychophysiological changes and (vi) health-related behaviours. Collectively, these factors (vii) impact the risk of developing somatic diseases and mental disorders.

Considerable evidence suggests that the psychosocial work environment significantly affects the overall health of workers (Fernandes & Pereira, 2016 ; Stansfield & Candy, 2006 ). Factors in work design or management that increase the risk of work-related stress are termed psychosocial risks or hazards. The World Health Organisation lists key aspects of the psychosocial work environment, such as job content, workload, work pace, scheduling, control, equipment, organisational culture, interpersonal relationships, role clarity, career development, and the balance between work and home life (Leka & Jain, 2010 ).

The heightened stress that workers experience due to psychosocial risks triggers a range of responses. Physiologically, it leads to neuroendocrine and immune reactions. Emotionally, it manifests as feelings of anxiety, depression, or depressive symptoms, as well as alienation and apathy, among other emotional reactions. Cognitively, it results in restricted perception, altered concentration capacity, reduced creativity, and difficulties in decision-making (Rugulies, 2019 ). Behaviourally, it can lead to actions such as substance abuse (e.g. alcohol, tobacco, drugs) and even acts of violence (Fernandes & Pereira, 2016 ). Conversely, positive psychosocial factors or resources are the are aspects of the work environment that reduce the risk of work-related stress, avoid damaging responses, and contribute positively to employees’ mental health.

A key strength of framing teachers’ work in terms of the psychosocial work environment is that it provides a clear framework for identifying and assessing the challenges faced by teaching professionals. Moreover, this framing encompasses established frameworks and strategies for responding to challenges, which can be employed to promote healthier and more sustainable work conditions for teachers. For instance, LaMontagne et al. ( 2014 ) propose an integrated approach to promoting mental health and wellbeing at work, highlighting the advantages of combining strategies at the individual, organisational, and policy levels. Given the widespread interest in the challenges facing the teaching profession and the health and wellbeing of teachers, lessons from the field of occupational health related to teachers’ work could prove to be very useful.

Psychosocial working conditions in schools and teachers’ intentions to leave

Psychosocial working conditions play a significant role in shaping teachers’ career intentions and decisions to remain or leave the profession. These conditions encompass various aspects within the school environment that influence teachers’ health and wellbeing, job satisfaction, and overall experience in the profession. Evidence shows that positive psychosocial work conditions lead to better career outlooks and retention among teachers (Arnold et al., 2023 ; Casely-Hayford et al., 2022 ; Rahimi et al., 2023 ). This section examines how key psychosocial factors relate to teachers’ plans to exit the profession.

Job demands

Job demands in the workplace are often associated with psychosocial risks because they require sustained physical or psychological effort from an employee. These demands include workload, work intensity, and the emotionally demanding nature of the work. Many studies have demonstrated the negative impact of heavy workloads on teachers’ career intentions (Li & Yao, 2022 ). In Australia, Heffernan et al. ( 2022 ) found that the primary reason that teachers planned to leave the profession was workload that was ‘excessive, unrealistic and unsustainable’. They reported that both the volume and diversity of tasks had grown, leading to unsustainable workloads. Alongside this, the emotionally demanding nature of teaching work, compounded by limited rest opportunities, was linked to higher turnover intentions. Similarly, Rajendran et al. ( 2020 ) determined that workload, work–family imbalance, and student misbehaviour were associated with emotional exhaustion that led to turnover intention. Understanding the extent of teachers’ work, including the number of hours, administrative tasks, and time spent on non-teaching activities, is crucial (Stacey et al., 2023 ). Additionally, the emotional dimensions of the workload have been shown to be vital to workforce sustainability (Tuxford & Bradley, 2015 ; Yin et al., 2019 ). Teachers facing psychological and emotional pressures, such as student care and behavioural challenges, experience more work–life conflict. This suggests a strong link between job demands, personal life, and career intentions in the teaching profession.

Job resources

Job resources refer to the various conditions within the workplace that help employees to cope with job demands and enhance their ability to achieve their professional goals. These resources can have a positive impact on employees’ health, satisfaction and motivation to remain in their roles. There are many different job resources that impact teachers’ work. However, evidence suggests that organisational environments or cultures that promote collaboration among employees can foster healthy and sustainable work roles. High social capital in organisations fosters collaborative climates based on trust and fairness. In a meta-analysis by Li and Yao ( 2022 ), trust was the top psychosocial predictor of turnover intention, followed by school climate, justice, and colleague support. There is evidence that when teachers have high levels of trust in school leadership (Karakuş et al., 2014 ), a positive school climate (Awang et al., 2015 ), and perceptions that their workplace is fair and just (Esop & Timmes, 2019 ) they are less likely to leave their roles.

Teachers’ experiences of work

In addition to psychosocial working conditions, teachers’ experiences of work and their relationships with their jobs have been shown to have important consequences for turnover intention. When teachers have positive experiences at work and positive evaluations of their jobs, they are more likely to express intentions to remain in their roles. For example, work engagement (Ferrer & Morris, 2013 ), job satisfaction (Skaalvik & Skaalvik, 2017 ; 2011 ), and workplace commitment (Esop & Timms, 2019 ; Lawrence et al., 2014 ) are negatively correlated with teachers’ intentions to leave. There is evidence that teachers who are highly engaged in their work tend to exhibit higher levels of commitment and a reduced intention to leave their positions (Ferrer & Morris, 2013 ). Conversely, dissatisfaction with their jobs, including dissatisfaction with working conditions (Ingersoll, 2001 ), salary or support from leadership (Guarino et al., 2006 ), increases the likelihood of teachers reporting the intention to leave their positions.

Teachers’ mental health

These challenges of teaching, alongside concerns about work-related safety, have led to concerns that for many teachers, teaching is a stressful profession in which mental health difficulties are common. In their daily work, teachers must contend with a multitude of stressors stemming from interactions with students, parents, and the demands of the school environment. Empirical studies have consistently demonstrated the adverse impact of stress on teacher attrition (Bukhari & Kamal, 2017 ; Klassen & Chiu, 2011 ; Kyriacou, 2001 ). Furthermore, research indicates that acute work-related stressors can lead to more serious mental health problems including burnout and depression (Stansfeld & Candy, 2006 ). There is evidence that teacher burnout predicts lower engagement at work, lower levels of satisfaction, increased absenteeism, increased motivation to leave the teaching profession, and increased teacher attrition (Madigan & Kim, 2021 ). In Australia, Heffernan et al. ( 2022 ) found that concern for their mental health and wellbeing, including experiences of exhaustion, stress, and burnout, led many teachers to consider leaving their job, while other studies have demonstrated that burnout is associated with turnover intention for Australian teachers (Goddard & Goddard, 2006 ; Rajendran et al., 2020 ).

Aim of the study

The principal aim of this paper is to investigate critical elements of the psychosocial work environment, teachers’ professional experiences, and their mental health as these factors relate to the intention to leave or remain in their roles. A deeper understanding of these issues will help to identify the key issues facing the Australian teaching workforce in government schools and offer opportunities for interventions at the policy, school, and individual teacher level.

Methodology

Participants.

Participants were recruited through email invitations sent to around 10,000 Australian teachers registered with Qualtrics panels. They were invited to access the survey through a secure online link. Out of those invited, 3328 opened the link. On the survey page, they were provided with information about the study in plain language, along with a consent form to be agreed upon before starting the survey. A dual-stage screening was applied at the start of the survey to exclude participants not meeting the study’s criteria, such as inactive teachers, those not teaching in government sectors, or lacking teaching qualifications. An additional screening ensured representativeness across key demographics like state/territory and school level. The final dataset contained 1005 records, with 905 active teachers in Australian government schools deemed viable for the study following the exclusion of entries with missing data.

In the final study sample, a significant majority were female, comprising around three-quarters of the respondents, with males making up the remainder. Most teachers worked in metropolitan locations, while smaller proportions were based in inner-regional, outer-regional, remote, and very remote schools. Regarding teaching levels, just over half taught at primary schools, a sizeable portion in secondary, and a small minority in combined levels. The majority of respondents were employed full time, with a smaller percentage working part-time and a minor fraction engaged as casual teachers. In terms of professional experience, roughly one-third had less than 5 years in the teaching role, while the majority had more extensive experience, with more than 5 years in the field (see Table  1 ).

As reported by the Australian Bureau of Statistics in 2022, the distribution of teaching staff was split between primary schools, which employed 156,019 (50.8%) teachers, and secondary schools, with 151,021 (49.2%) teachers. Notably, secondary schools boasted a higher proportion of male teachers at 38.6%, in contrast to primary schools, which had 18.0% (Australian Bureau of Statistics, 2022 ).

This paper is drawn on the data collected through the Copenhagen Psychosocial Questionnaire III (COPSOQ-III)—a validated tool for assessing psychosocial work environments, which is widely used across various occupations, including education. The Turnover Intention Scale (TIS-6) is similarly recognised for reliably measuring turnover intentions. These tools were selected for their effectiveness in capturing the nuances of the teaching profession’s psychosocial work environment and career intentions. We distributed the survey to primary and secondary school teachers in Australian government schools via a secure online platform. The majority of the survey consisted of the Copenhagen Psychosocial Questionnaire III (COPSOQ-III), an instrument known for its validity and reliability in assessing psychosocial work environments across various working populations, including Australian school leaders (Burr et al., 2019 ). The survey examined key aspects of teachers’ psychosocial work environments, including key job demands and job resources for teachers. Job demand factors, such as emotional and quantitative demands, and work–family conflict are seen as stressors that could potentially lead to strain and health risks. Job resources factors, including trust, justice, and recognition are aspects of work that can enable teachers to do their work. The survey also assessed teachers’ experiences of and attitudes to their work through three variables, job satisfaction, commitment to the workplace, and the quality of work. Finally, COPSOQ-III was also used to assess three key mental health-related outcomes: burnout, depressive symptoms, and stress.

To assess teachers’ career intentions, we used the short version of the Turnover Intention Scale (TIS-6) (Bothma & Roodt, 2013 ). TIS-6 has been shown to be a reliable measure of turnover intention that can distinguish between actual leavers and stayers (Bothma & Roodt, 2013 ). Study participants were prompted to think back on their experiences over the prior nine months and answer questions like: ‘how often have you thought about quitting your job?’, ‘how frequently have you felt frustrated by unachieved work goals?’, ‘how often have you fantasised about a different job?’, ‘how well does your job meet your personal needs?’, ‘how likely would you be to take another job with the same pay?’, and ‘how often do you look forward to another workday?’. These responses were totalled and normalised on a 0 to 100 scale, yielding a uniform metric for each teacher’s likelihood to leave their current role. A score above the 50 midpoint indicates a stronger inclination to leave, whereas a score below 50 suggests a weaker inclination.

Both instruments primarily utilised Likert scale responses. For the COPSOQ-III, participants were typically asked to respond on a 5-point scale, ranging from ‘always’ to ‘never’ or ‘to a very large extent’ to ‘to a very small extent’, depending on the question context. Similarly, the TIS-6 responses were structured on a Likert scale, where participants indicated their level of agreement or frequency of experiencing certain feelings related to turnover intentions. This methodology afforded a thorough, context-relevant exploration of factors affecting teachers’ workplace engagement and intentions. Open-ended survey questions also collected qualitative data, which were analysed and are presented in alignment with pertinent themes.

Data preparation and visualisations

The collected data underwent preparation, cleaning, and processing using a range of analytical programming methods in R, Python, and Microsoft Excel. In this study, we use scatter diagrams to visualise the relationships among groups of four variables. The utilisation of scatter graphs in this study enables a visual investigation into the relationships among four distinct variables related to teacher intention to leave. By plotting individual data points for different combinations of variables, such as work experience, job satisfaction, emotional demands, and intention to leave, scatter plots offer a way to discern underlying patterns and associations. This method aids in identifying clusters, trends, or outliers that might not be apparent through numerical data alone. By offering a clear visual representation, scatter plots enhance our understanding of the complex interplay between these factors within the teaching profession in Australian government schools.

COPSOQ measures of work and wellbeing

In evaluating the teachers’ work environment and mental wellbeing, the means of normalised scores derived from the Copenhagen Psychosocial Questionnaire III (COPSOQ) (Burr et al., 2019 ) serve as a vital measure. By calculating the average scores across various measures of each dimension for those that expressed an intention to stay and those that intend to leave we can gauge the prevailing conditions and attitudes within the two groups. This methodical approach, through the aggregation of individual responses, provides a comprehensive understanding of the broader trends in teacher morale, working conditions, their wellbeing, and job satisfaction. These mean scores allow for a comparative analysis between ‘leavers’ and ‘stayers’ and enable us to pinpoint areas that may require targeted intervention or support. The mean scores were used to provide a comparative analysis of the measures for teachers and the Australian workforce ( n  = 2442). The provisional unweighted measures for the Australian workforce were calculated based on the data that was separately collected from a random sample of working population of Australia. This approach thus provides insights specific to the teaching profession and situates these findings within the broader landscape of employment conditions across different sectors in Australia.

Correlations

In this study, correlation analysis plays a pivotal role in understanding the relationships between different factors influencing teacher retention within Australian schools. Specifically, Pearson’s correlation coefficient ( r ) was employed to measure the linear association between variables such as workload, support, job satisfaction, and intention to leave. The coefficient values range from + 1 to − 1, where + 1 denotes a perfect positive linear relationship, − 1 indicates a perfect negative linear relationship, and 0 signifies no linear correlation. The significance of each correlation was further tested to ascertain whether the observed relationships occurred by chance. The application of Pearson’s correlation coefficient enabled a rigorous and objective examination of how variables interrelate, shedding light on the underlying patterns that may guide teacher retention strategies. This statistical approach contributes valuable insights to the complex dynamics within the teaching profession, with implications for both policy and practice (Cohen et al., 2013 ).

• Intention to leave

In a comparison to the general Australian workforce, a greater proportion of teachers in Australian government schools report planning to leave their roles (See Fig.  2 ). The proportion of teachers planning to leave their roles was 4% higher among teachers (43.2%), compared to the general working population (39.2%). This difference underscores the unique challenges faced by the teaching community, drawing attention to a higher rate of turnover intention among educators.

Proportion (%) of teachers in Australian government schools that intend to stay or leave compared to the Australian general population

Table 2 shows that most Australian government school teachers, regardless of location or years of experience, intend to stay in their roles. However, a significant portion—between 35.7 and 47% across various categories—express a desire to leave. Primary teachers and those with fewer years in the field are more inclined to stay. The intention to leave or stay appears consistent across both full-time and part-time/casual employment status.

Psychosocial work environments and teachers’ career intentions

The next section of our analysis explores teachers’ intentions to leave in relation to several key groups of factors related to the concept of the psychosocial work environment: Job demands, Job resources, Experiences of work, and Mental health-related factors. Through a detailed analysis of these categories, we aim to explore the key factors that are associated with teacher career intentions and workforce sustainability.

Job demands: strains and challenges in the teaching environment

In school contexts, aspects of teachers’ work that demand significant and ongoing effort can potentially have a negative impact on teachers’ work–life balance and their career intentions. In response to our request to describe significant actions that would help retain teachers in the profession, several teachers identified issues related to job demands and the challenges of balancing work with their personal lives:

The workload is insane. The student misbehaviour and the amount of time spent on planning, assessment and correction is overwhelming. General conditions are alright. But workload is absolutely ridiculous. There are constantly new initiatives to ‘raise’ student achievements, so teachers are just finding their feet, then something new comes along and it’s new learning all over again. There are increasing numbers of challenging children due to trauma, behaviour and learning needs. And it’s ever harder to get support for them. More flexibility with hours and family/personal commitments. Despite its reputation otherwise, teaching is not a family-friendly occupation.

These teacher comments paint a picture of a demanding and complex work environment. Overwhelming workloads , frequently changing educational initiatives , and increasing behavioural and learning challenges in students create a challenging work environment. Teachers emphasise the need for more support and fewer new initiatives aimed at enhancing student learning , highlighting the emotional strain from managing student behaviour and continually changing their practice to accommodate new strategies and programs. Furthermore, they express a desire for more flexible work arrangements to better balance professional and personal commitments .

These comments highlight the importance of emotional demands, sheer quantity of work (quantitative demands), and work–family conflict. Teachers expressed a desire for more manageable workloads, including a smaller quantity or amount of work tasks (quantitative demands), more support to cope with the emotionally demanding aspects of their work and a better work–life balance.

The survey results supported these findings by demonstrating notable differences in perceptions of workload, emotional demands, and work–life imbalance between teachers planning to leave and stay (Fig.  3 ). Teachers intending to leave faced heavier workloads, greater emotional stress, and more work–life conflicts than those intending to stay and the average Australian worker. It is concerning that all of these measures for the government school teachers who intend to stay are even higher than the average for the Australian working population.

A comparison of the mean scores (out of 100) on the key job demand variables between teachers who intend to stay, teachers who intend to leave, and the average for the Australian working population

All three of the variables related to job demands show moderate correlations with intention to leave. Given the positive correlations of quantitative demands and intention to leave ( r  = 0.48, p  < 0.001), as well as between emotional demands and intention to leave ( r  = 0.44, p  < 0.001), it can be inferred that there is an association between higher quantitative and emotional demands at work and teachers’ intentions to leave their jobs. The positive correlation between work–family conflict and intention to leave ( r  = 0.50, p  < 0.001) suggests a positive relationship between work–family conflict and intention to leave the job.

Considering these associations, the scatterplot gives a visual depiction of these relationships. The size of each bubble in the plot corresponds to the level of work–family conflict experienced by the teachers. The colour of the bubbles represents the teachers’ intention to leave their current job, with green indicating a lower intention to leave (or higher intention to stay) and red indicating a higher intention to leave (see Fig.  4 ).

Association of Intention to leave with Emotional demands, Quantitative demands & Work–family conflict (bubble size)

The positioning of each bubble along the axes reveals how the emotional demands and quantitative demands interact with both work–family conflict and the intention to leave. Teachers experiencing high emotional demands and high quantitative demands (top-right quadrant, labelled ‘A: Higher int. to leave’) generally have larger bubbles (indicating higher work–family conflict) and a red colour (indicating a higher intention to leave), which aligns with the correlations observed in the matrix shown in Table  3 .

Job resources: supportive dynamics of the psychosocial work environment

In school contexts, job resources, which encompass aspects such as job recognition, trust in management, and organisational justice, can have a significant influence on a teacher’s decision to persist in or leave their profession. In response to our request to describe any significant actions that would help to retain teachers in the profession, several teachers identified issues related to trust, justice, and recognition within their schools and the profession more broadly:

Just trust our professionalism and let us teach. Stop the micromanagement. Less red tape, trusting and allowing teachers to use their knowledge and expertise like the professionals we are. Respect from Government and Community. Less nepotism in promotional positions. Career progression from a young age, not just waiting for someone to be ‘an age’ before they have access to that job.

These comments highlight the importance of recognition for work done and also emphasise the need for fairness and trust at work. The emphasis on trusting teachers’ professionalism and reducing micromanagement aligns with the significance of job recognition and trust in management, as noted by Li and Yao ( 2022 ) and Karakuş et al. ( 2014 ). These studies highlight trust as a crucial psychosocial predictor of turnover intention and the positive impact of trust in school leadership on retention. The call for less red tape and respect from government and community echoes the need for organisational justice and societal recognition, as discussed by Esop and Timmes ( 2019 ), who found that perceptions of a fair and just workplace significantly influence teachers’ decisions to stay.

These views were supported by survey results, showing significant disparities in perceptions of organisational justice, leadership trust, and job recognition between teachers planning to leave and stay (Fig.  5 ). Teachers intending to leave felt they worked in less fair settings, had less trust in leaders, and received less job recognition compared to those planning to stay and the average Australian worker.

A comparison of the mean scores (out of 100) on the key job resources variables between teachers who intend to stay, teachers who intend to leave, and the average for the Australian working population

All three variables of job recognition, trust in leadership, and organisational justice have moderate negative correlations with intention to leave. Given the negative correlations between job recognition and intention to leave ( r  = − 0.55, p  < 0.001), trust in leadership and intention to leave ( r  = − 0.55, p  < 0.001), and organisational justice and intention to leave ( r  = − 0.53, p  < 0.001), it can be inferred that higher levels of job recognition, trust in management, and organisational justice are associated with teachers’ intention to remain in their roles (see Table  4 ).

Figure  6 provides a visual representation of these relationships. The colour of the bubbles represents the teachers’ intention to leave their current job, with green indicating a lower intention to leave (or higher intention to stay) and red indicating a higher intention to leave. The size of each bubble in the plot (corresponding to job recognition) indicates the level of recognition perceived by the teachers.

Association of Intention to leave with Trust in management, Organisational justice & Job recognition (bubble size)

The positioning of each bubble along the axes then reveals how perceptions of organisational justice and trust in management interact with both job recognition and the intention to leave. Teachers with high trust in management and high organisational justice (top-right quadrant, labelled ‘A: Higher int. to stay’) generally have larger bubbles (indicating higher job recognition) and green colour (lower intention to leave), which aligns with the correlations observed in the matrix. This aligns with the correlations in Table  4 .

Teachers’ experiences of work: valuing engagement in teaching

In response to our request to describe significant actions that would help retain teachers in the profession, several teachers identified issues related to the quality of their work and the opportunities for engaging in meaningful work:

Less meaningless tasks more quality teaching time. More time to teach, less time on administrative tasks.

These comments emphasise the importance of teaching staff engaging in quality work by dedicating their time to what means the most to them: teaching students, rather than administrative tasks. These align with the findings of Ferrer and Morris ( 2013 ), who noted that work engagement is a critical factor in reducing teachers’ intention to leave. This sentiment is echoed in the call for ‘more time to teach, less time on administrative tasks’, underscoring the importance of job content and the balance of teaching versus administrative duties in influencing teachers’ job satisfaction, as highlighted by Skaalvik and Skaalvik ( 2017 ).

Survey results show significant differences in personal experiences between teachers planning to leave and those planning to stay (Fig.  7 ). Teachers wanting to leave had lower work quality, less role commitment, and lower job satisfaction compared to their staying counterparts and the average Australian worker.

A comparison of the mean scores (out of 100) on the key ‘experiences of work’ variables between teachers who intend to stay, teachers who intend to leave, and the average for the Australian working population

All three variables related to teachers’ experiences of their work show moderate negative correlations with intention to leave. Given the negative correlations between job satisfaction and intention to leave ( r  = − 0.58, p  < 0.001), commitment to the workplace and intention to leave ( r  = − 0.65, p  < 0.001), as well as quality of work and intention to leave ( r  = − 0.47, p  < 0.001), it can be inferred that higher job satisfaction, a greater commitment to the workplace, and a higher quality of work are associated with teachers’ intention to remain in their jobs (see Table  5 ).

Correlations between these variables and the intention to leave are displayed in Table  5 , with Fig.  8 visually depicting these relationships. Figure  8 positions each data point based on the teachers’ commitment to the workplace ( x -axis) and quality of leadership ( y -axis). The size of each point corresponds to the level of satisfaction on quality of work experienced by the teachers. The colour of the bubbles represents the teachers’ intention to leave their current job, with green indicating a lower intention to leave (or higher intention to stay) and red indicating a higher intention to leave.

Association of Intention to leave with Job satisfaction, Commitment to the workplace, and Quality of work (bubble size)

Teachers experiencing high commitment to the workplace and high quality of work (top-right quadrant, labelled ‘A: Higher int. to stay’) generally have larger bubbles (indicating higher satisfaction of work quality) and a green colour (indicating a lower intention to leave). These observations align with the correlations observed in the matrix (see Table  5 ).

Mental health-related factors: the importance of supporting teacher wellbeing

In response to our request to describe significant actions that would help retain teachers in the profession, several teachers identified issues related to the mental health and wellbeing:

Prioritising staff mental health and out of work hours. Better conditions to reduce end of year burnout. More focus on teacher wellbeing (burnout, mental health etc). Arrange psychological counselling, so that the teacher’s pressure can be relieved and relieved.

Teachers reported that prioritising mental health and reducing burnout through improved conditions, increased assistance, and a greater focus were essential. Results from the survey determined significant differences in teachers’ mental health outcomes between those who plan to leave and those who plan to stay (see Fig.  9 ). Teachers who intended to leave their roles reported higher levels of stress, more depressive symptoms, and increased burnout compared to both teachers who intend to stay and the average Australian worker.

A comparison of the mean scores (out of 100) on the key ‘mental health’ variables between teachers who intend to stay, teachers who intend to leave, and the average for the Australian working population

The positive correlations between job satisfaction and intention to leave ( r  = − 0.58, p  < 0.001), commitment to the workplace and intention to leave ( r  = − 0.65, p  < 0.001), as well as the quality of work and intention to leave ( r  = − 0.47, p  < 0.001), demonstrate that increased burnout, more depressive symptoms, and higher stress levels are associated with teachers’ intentions to leave their job (see Table  6 ).

The positive correlations with intention to leave for Burnout, Stress, and Depressive symptoms suggest that there is a moderate-to-strong relationship between mental health and intention to leave. Examining these connections further, Fig.  10 offers a graphical interpretation of these relationships. In this plot, the colour of the bubbles represents the teachers’ intention to leave their current job, with green indicating a lower intention to leave (or a higher intention to stay) and red indicating a higher intention to leave. The size of each bubble corresponds to the level of stress experienced by the teachers.

Association of Intention to leave with Burnout, Depressive symptoms & Stress (bubble size)

The positioning of bubbles along the axes discloses how burnout and depressive symptoms interact with both stress (bubble size) and the intention to leave. Teachers experiencing high levels of burnout and depressive symptoms (zone A at the top-right quadrant, labelled ‘A: Higher int. to leave’) generally have larger bubbles (indicating higher levels of stress) and a red colour (indicating a higher intention to leave), which aligns with the correlations observed in the matrix shown in Table  6 .

This paper also provides an exploration of the relationship between several key aspects of Australian government school teachers’ work and health and their intention to remain in or leave their roles. Framed by the concept of the psychosocial work environment, our analysis demonstrates the importance of the relationship between both the quantity of work that teachers face and the emotionally demanding nature of the role, and teachers’ intentions to leave their roles. In addition, there was a significant relationship between the experience of work–life conflict and teachers’ intentions to leave. Taken together, this aspect of the analysis demonstrates that perceived demands at work and the encroachment of work on teachers’ personal lives are important issues related to teachers’ career intentions.

Workload among the teaching profession is recognised as a major issue affecting the profession (Heffernan, 2021 ; Heffernan et al., 2022 ). We have demonstrated that teachers who express a desire to leave report heavier workloads than those staying. Developing a fine-grained understanding of the sheer volume of work that teachers must contend with, such as the number of hours worked, the number of administrative tasks that must be completed (Stacey et al., 2023 ), and the time spent on non-teaching tasks, is critical to the future of the profession.

In addition, our analysis aligns with other research demonstrating that the emotional dimensions of teacher workload are also critical to the sustainability of the workforce (Tuxford & Bradley, 2015 ; Yin et al., 2019 ). Teachers that intended to leave their roles were significantly more likely to report experiencing psychological and emotional pressures related to their work. Responding to the emotional needs of students, caring for students and their future, addressing behavioural challenges and being under pressure to perform can all contribute to the emotionally demanding nature of teaching work (Tuxford & Bradley, 2015 ; Yin et al., 2019 ). Teachers intending to leave experienced more work–life conflict than those planning to stay. This implies a significant link between job demands affecting personal life and career intentions.

In terms of the positive aspects of the psychosocial work environment, our analysis determined that teachers who intended to leave their roles had significantly lower levels of job recognition, trust in leadership, and organisational justice compared to their colleagues who intended to remain. Our analysis demonstrates the importance of the relationship between these three variables and teachers’ career intentions. In terms of job recognition, acknowledgment, appreciation, and validation of teachers’ contributions and efforts are associated with teachers’ career intentions (Arnup & Bowles, 2016 ). Perceptions of fairness and equity in the treatment of teachers and other staff, decisions affecting teachers and their colleagues, as well as work processes, and the belief that school and system leaders are honest, competent, transparent, and have the best interests of teachers, students, and the school in mind, are also important factors influencing teachers’ career intentions.

Psychosocial work environments significantly affect teachers’ satisfaction, commitment, and perceptions of their job (Rugulies, 2019 ). Those intending to leave were notably more dissatisfied and less committed, and they perceived lower job quality compared to those planning to stay. Our findings highlight the positive association between higher levels of commitment, job satisfaction, and work quality and teachers’ intentions to remain in their roles (Madigan & Kim, 2021 ). These factors, which align with Organisational Commitment Theory, correlate with intentions to stay and build on prior research (Watt & Richardson, 2008 ), suggesting that fostering a work environment that promotes professional commitment and satisfaction is key to enhancing teacher retention.

Ultimately, teachers’ psychosocial working conditions, as shaped by their experiences and evaluations of their work, contribute to specific mental health outcomes (Fernandes & Pereira, 2016 ; Rugulies, 2019 ). Our analysis demonstrates a positive association between mental health problems and teachers’ intentions to leave. Teacher stress, burnout, and depressive symptoms were all positively associated with intentions to leave. Teachers who intended to leave their roles experienced significantly more stress, higher levels of burnout, and more symptoms of depression than colleagues who intended to stay.

This aligns with evidence about the importance of mental health for employees’ intentions to remain in their roles (Madigan & Kim, 2021 ). Depressive symptoms, burnout, and stress can all influence an individual’s career intentions and increase their desire to seek alternative opportunities (Bakker & Demerouti, 2007 ; Demerouti et al., 2001 ; Ybema et al., 2014 ). As discussed through a broad spectrum of literature, our findings provide insights into global educational settings, highlighting universally relevant psychosocial factors in the teaching profession and offering a framework for international policy and practice in diverse educational contexts.

Implications

These findings emphasise the importance of the psychosocial work environment to create supportive work environments that not only mitigate the risks of stress and burnout but also contribute to the retention and overall health and wellbeing of teachers in the profession. The strategies and frameworks discussed here, though based on Australian data, have broader implications, suggesting ways to improve teacher retention and wellbeing worldwide.

The current analysis offers a framework for understanding the complexities facing Australia’s teachers. It advocates for a unified approach to manage these challenges, promote positive work aspects, and support mental health. Current approaches are fragmented, employing various measures across jurisdictions. While multiple research perspectives are crucial for understanding teachers’ work and health, it is key for policymakers to adopt consistent methods to identify and assess psychosocial risks. These data should then inform comprehensive policies to tackle challenges teachers face, along with providing organisational support.

LaMontagne et al.’s ( 2014 ) integrated approach to workplace health could be useful in framing a coherent response to retaining Australian government schools teachers. The integrated approach aims to achieve the following objectives:

Prevent Harm: Enhance mental health by mitigating workplace-related risk factors.

Promote the Positive: Foster mental wellbeing by nurturing positive elements of work and harnessing employees’ strengths and positive capabilities.

Manage Illness: Tackle mental health challenges faced by working individuals, regardless of the underlying causes.

The first objective focusses on protecting teachers’ mental wellbeing, emphasising the employer’s role in ensuring a healthy work environment. This means Departments of Education, in collaboration with teachers and school leaders and federal support, should address workforce challenges to foster safe conditions. Evidence suggests that preventive, organisational-level interventions can improve the psychosocial environment and employee health (Aust et al., 2023 ). Organisational-level interventions target changes in school policies, leadership styles, or working conditions.

While school-level intervention is important, teachers’ work in schools is heavily shaped by systems and policies beyond the school. The scope for change in government schools is severely limited by mandated policies, processes, and procedures at both state and federal levels. For interventions to improve the work environment in schools, there needs to be change in school systems. This is particularly important given that policymakers and systems administrators have dominated decisions about teachers’ work over recent decades (Stacey et al., 2023 ). Potential interventions that promote change at the school and policy level include the Total Worker Health approach (Sanetti et al., 2022 ). This involves policymakers, systems administrators, and teachers working together to understand and promote healthier working environments.

The second objective, promote the positive, focusses on harnessing the existing strengths, talents, and positive attributes within schools to enhance wellbeing. This approach focusses on promoting the positive aspects of work and teacher capabilities such as resilience and adaptability. Unlike strategies solely aimed at managing mental illness, it aims to empower teachers to better handle challenges. The third objective, manage illness, addresses mental health issues through qualified support like EAPs and psychoeducation programs (LaMontagne et al., 2014 ). Teachers with mental health problems receive confidential counselling, and literacy programs encourage early intervention and help-seeking.

According to LaMontagne et al. ( 2014 ) combining all three approaches could substantially improve the mental health outcomes over and above what might be achieved by each thread on its own. However, LaMontagne et al. ( 2014 ) also emphasise that there are considerable risks associated with this approach due to the tendency for employers to avoid the difficult and time-consuming work of reducing stress and preventing harm in favour of individual-directed interventions. This is evidenced by the prevalence of workplace mental health literacy and resilience-oriented positive psychology programs (LaMontagne et al., 2014 ). For the integrated approach to be successful, it is critical that employers avoid taking the easy option and focus on the more challenging and time-consuming work of changing the psychosocial work environment through the development of healthier school systems. This means that employers must listen to teachers and school leaders about how they can design systems and policies that support the development of healthy, sustainable workplaces for teaching professionals.

Limitations

The correlational approach shows the relationship between teachers’ intention to leave and various work-related factors, but cannot be used to confirm causality. Longitudinal analysis is needed for deeper insight. Additionally, the study is cross-sectional, limiting its scope in capturing the fluid and complex nature of teachers’ work environments and their impact on mental health and career intentions. Furthermore, this study concentrated solely on government school teachers due to the larger population size and accessibility of this sector. Including other school sectors would have required a more extensive and resource-intensive data collection effort, given their varied administrative structures.

This analysis demonstrated that the concept of the psychosocial work environment can be useful in distilling some of the key features of teachers’ work environments, their experiences of work and their mental health, and analysing their relationship with teachers’ intention to leave. Our analysis demonstrates strong positive relationships between higher workplace demands, lower workplace resources, adverse experiences of work, and poor mental health and intentions to leave. These findings provide insight into some of the key aspects of work and health that policymakers and researchers should focus on to promote the health and sustainability of the Australian government school workforce. The onus lies with policymakers and systems administrators to work with teachers and their representatives to develop a coherent, robust framework to understand the challenges that teachers face and take a proactive role in reimagining and restructuring the conditions in which our teachers work. This study, rooted in the Australian context, contributes to the global discourse on teacher retention, underscoring the universal relevance of psychosocial factors in the teaching profession. The methodical and conceptual approaches of this study, together, establish a foundation for the statistical modelling of the critical elements that affect teachers’ intentions to remain in or leave their profession, providing insights for crafting targeted retention strategies.

Availability of data and materials

Due to ethical constraints, the data supporting this study's findings cannot be made publicly available. Access is contingent upon further approvals from the ethics committee and consent from funding institution.

Arnold, B., Rahimi, M., & Riley, P. (2023). The mental health of leaders in Australian government schools: Stress and burnout over a decade (2011–2020). Educational Management Administration & Leadership . https://doi.org/10.1177/17411432231209713

Article   Google Scholar  

Arnup, J., & Bowles, T. (2016). Should I stay or should I go? Resilience as a protective factor for teachers’ intention to leave the teaching profession. Australian Journal of Education, 60 (3), 229–244.

Aust, B., Moller, J. L., Nordentoft, M., Frydendall, K. B., Bengtsen, E., Jensen, A. B., & Jaspers, S. O. (2023). How effective are organizational-level interventions in improving the psychosocial work environment, health, and retention of workers? A systematic overview of systematic reviews. Scandinavian Journal of Work Environment Health, 49 (5), 315–329.

Australian Institute for Teaching and School Leadership. (2021). Initial teacher education: data report 2021 . AITSL.

Australian Bureau of Statistics. (2022). Schools, Australia, 2022 . Australian Bureau of Statistics.

Australian Government. (2023). National teacher workforce action plan . Department of Education. Retrieved 15th February 2024 from https://www.education.gov.au/national-teacher-workforce-action-plan

Australian Institute for Teaching and School Leadership. (2022). Australian teacher workforce data: Teachers’ intentions to leave . Melbourne: AITSL.

Google Scholar  

Awang-Hashim, R., Kaur, A., & Noman, M. (2015). The interplay of socio-psychological factors on school engagement among early adolescents. Journal of Adolescence, 45 , 214–224.

Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the art. Journal of Managerial Psychology, 22 (3), 309–328.

Borman, G. D., & Dowling, N. M. (2008). Teacher attrition and retention: A meta-analytic and narrative review of the research. Review of Educational Research, 78 (3), 367–409.

Bothma, C. F., & Roodt, G. (2013). The validation of the turnover intention scale. SA Journal of Human Resource Management, 11 (1), 1–12.

Bukhari, I., & Kamal, A. (2017). Perceived organizational support, its behavioral and attitudinal work outcomes: Moderating role of perceived organizational politics. Pakistan Journal of Psychological Research, 32 (2), 271–288.

Burr, H., Berthelsen, H., Moncada, S., Nübling, M., Dupret, E., Demiral, Y., & Pohrt, A. (2019). The third version of the Copenhagen psychosocial questionnaire. Safety and Health at Work, 10 (4), 482–503.

Casely-Hayford, J., et al. (2022). What makes teachers stay? A cross-sectional exploration of the individual and contextual factors associated with teacher retention in Sweden. Teaching and Teacher Education, 113 , 103664.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2013). Applied multiple regression/correlation analysis for the behavioral sciences . Routledge.

Book   Google Scholar  

Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86 (3), 499–512.

Dicke, T., Stebner, F., Linninger, C., Kunter, M., & Leutner, D. (2018). A longitudinal study of teachers’ occupational well-being: Applying the job demands-resources model. Journal of Occupational Health Psychology, 23 (2), 262.

Esop, M., & Timms, C. (2019). Relevance of organisational support on academics’ affective commitment and turnover intentions. Journal of Applied Research in Higher Education, 11 (1), 118–128.

Fernandes, C., & Pereira, A. (2016). Exposure to psychosocial risk factors in the context of work: A systematic review. Revista De Saúde Publica . https://doi.org/10.1590/S1518-8787.2016050006129

Ferrer, J. L., & Morris, L. (2013). Engaging elitism: The mediating effect of work engagement on affective commitment and quit intentions in two Australian university groups. Higher Education Quarterly, 67 (4), 340–357.

Goddard, R., & Goddard, M. (2006). Beginning teacher burnout in Queensland schools: Associations with serious intentions to leave. The Australian Educational Researcher, 33 (2), 61–75.

Guarino, C. M., Santibanez, L., & Daley, G. A. (2006). Teacher recruitment and retention: A review of the recent empirical literature. Review of Educational Research, 76 (2), 173–208.

Heffernan, A. (2021). Retaining Australia’s school leaders in ‘challenging’ contexts: The importance of personal relationships in principal turnover decisions. International Journal of Educational Research, 105 , 101716.

Heffernan, A., Bright, D., Kim, M., Longmuir, F., & Magyar, B. (2022). ‘I cannot sustain the workload and the emotional toll’: Reasons behind Australian teachers’ intentions to leave the profession. Australian Journal of Education, 66 (2), 196–209. https://doi.org/10.1177/00049441221086654

Ingersoll, R. M. (2001). Teacher turnover and teacher shortages: An organizational analysis. American Educational Research Journal, 38 (3), 499–534.

Ingersoll, R. M., & Strong, M. (2011). The impact of induction and mentoring programs for beginning teachers: A critical review of the research. Review of Educational Research, 81 (2), 201–233.

Karakuş, M., Toprak, M., & Gürpınar, M. (2014). Structural Equation Modelling on the Relationships between Teachers' Trust in Manager, Commitment to Manager, Satisfaction with Manager and Intent to Leave. Croatian Journal Educational/Hrvatski Casopis za Odgoj I Obrazovanje, 16 (1).

Klassen, R. M., & Chiu, M. M. (2011). The occupational commitment and intention to quit of practicing and pre-service teachers: Influence of self-efficacy, job stress, and teaching context. Contemporary Educational Psychology, 36 (2), 114–129.

Kyriacou, C. (2001). Teacher stress: Directions for future research. Educational Review, 53 (1), 27–35.

LaMontagne, A. D., Martin, A., Page, K. M., Reavley, N. J., Noblet, A. J., Milner, A. J., Keegel, T., & Smith, P. M. (2014). Workplace mental health: Developing an integrated intervention approach. BMC Psychiatry, 14 , 131.

Lawrence, J. H., Celis, S., Kim, H. S., Lipson, S. K., & Tong, X. (2014). To stay or not to stay: Retention of Asian international faculty in STEM fields. Higher Education, 67 , 511–531.

Leka, S., Jain, A., World Health Organization. (2010). Health impact of psychosocial hazards at work: An overview . https://iris.who.int/bitstream/handle/10665/44428/?sequence=1

Li, R., & Yao, M. (2022). What promotes teachers’ turnover intention? Evidence from a meta-analysis. Educational Research Review, 37 , 100477.

Madigan, D. J., & Kim, L. E. (2021). Towards an understanding of teacher attrition: A meta-analysis of burnout, job satisfaction, and teachers’ intentions to quit. Teaching and Teacher Education, 105 , 103425.

Nguyen, T. D., Pham, L., Springer, M. G., Crouch, M. (2019). The factors of teacher attrition and retention: An updated and expanded meta-analysis of the literature.  Annenberg Institute at Brown University , 19–149.

Perryman, J., & Calvert, G. (2020). What motivates people to teach, and why do they leave? Accountability, performativity and teacher retention. British Journal of Educational Studies, 68 (1), 3–23.

Rahimi, M., Arnold, B., & Horwood, M. (2023). A national study of New Zealand primary-school teaching professionals’ experiences of offensive behavior at the workplace. Humanities and Social Sciences Communications, 10 (1), 1–9. https://doi.org/10.1057/s41599-023-01583-4

Rajendran, N., Watt, H. M. G., & Richardson, P. W. (2020). Teacher burnout and turnover intention. Australian Education Researcher, 47 , 477–500. https://doi.org/10.1007/s13384-019-00371-x

Riley, P., See, S. M., Marsh, H., & Dicke, T. (2020). The Australian principal occupational health, safety and wellbeing survey (IPPE report) . Australian Catholic University.

Rugulies, R. (2019). What is a psychosocial work environment? Scandinavian Journal of Work, Environment & Health, 45 (1), 1–6. https://doi.org/10.5271/sjweh.3792

Sanetti, L. M., Pierce, A. M., Gammie, L., Dugan, A. G., & Cavallari, J. M. (2022). Scale-out of a Total Worker Health® approach for designing interventions to reduce teacher stress: Pilot implementation evaluation. BMC Public Health, 22 (1), 814.

Skaalvik, E. M., & Skaalvik, S. (2011). Teacher job satisfaction and motivation to leave the teaching profession: Relations with school context, feeling of belonging, and emotional exhaustion. Teaching and Teacher Education, 27 (6), 1029–1038.

Skaalvik, E. M., & Skaalvik, S. (2017). Still motivated to teach? A study of school context variables, stress and job satisfaction among teachers in senior high school. Social Psychology of Education, 20 , 15–37.

Stacey, M., McGrath-Champ, S., & Wilson, R. (2023). Teacher attributions of workload increase in public sector schools: Reflections on change and policy development. Journal of Educational Change . https://doi.org/10.1007/s10833-022-09476-0

Stansfeld, S., & Candy, B. (2006). Psychosocial work environment and mental health—a meta-analytic review.  Scandinavian Journal of Work, Environment & Health , 443–462.

Tuxford, L. M., & Bradley, G. L. (2015). Emotional job demands and emotional exhaustion in teachers. Educational Psychology, 35 (8), 1006–1024.

Watt, H. M. G., & Richardson, P. W. (2008). Motivations, perceptions, and aspirations concerning teaching as a career for different types of beginning teachers. Learning and Instruction, 18 (5), 408–428.

Ybema, J. F., Geuskens, G. A., van den Heuvel, S. G., de Wind, A., Leijten, F. R., Joling, C. I., Blatter, B. M., Burdorf, A., Beek, A. V. D., & Bongers, P. M. (2014). Study on transitions in employment, ability and motivation (STREAM): The design of a four-year longitudinal cohort study among 15,118 persons aged 45 to 64 years. British Journal of Medicine and Medical Research, 4 (6), 1383–1399.

Yin, H., Huang, S., & Chen, G. (2019). The relationships between teachers’ emotional labor and their burnout and satisfaction: A meta-analytic review. Educational Research Review, 28 , 100283.

Download references

Open Access funding enabled and organized by CAUL and its Member Institutions. Authors gratefully acknowledge the Research for Educational Impact (REDI) at Deakin University for providing financial support to this study.

Author information

Authors and affiliations.

School of Education, Faculty of Arts and Education, Deakin University, Burwood, VIC, Australia

Mark Rahimi & Ben Arnold

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Mark Rahimi .

Ethics declarations

Conflict of interest.

The authors declare no competing interests.

Ethical approval

Authors confirm that ethical approval for this research project was obtained from The Human Research Ethics Office, Deakin University (Deakin University research: 2019–341), and all research were performed in accordance with relevant guidelines/regulations.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Rahimi, M., Arnold, B. Understanding Australia’s teacher shortage: the importance of psychosocial working conditions to turnover intentions. Aust. Educ. Res. (2024). https://doi.org/10.1007/s13384-024-00720-5

Download citation

Received : 04 October 2023

Accepted : 05 April 2024

Published : 08 May 2024

DOI : https://doi.org/10.1007/s13384-024-00720-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Teachers’ shortage
• Psychosocial
• Work environments
• Mental health and wellbeing
• Find a journal
• Publish with us
• Track your research