Episode 4 – Karst Groundwater: quick and slow at the same time?
We often associate groundwater with large water storage and very slow water movement for instance compared to rivers. But is it possible that groundwater flow can be as quick as stream flow and, at the same aquifer, flow for several months or years before it is reaching the surface again? Of karst, it is possible! When chemical weathering is able dissolve carbonate rock, cracks and fissures may grow to a subsurface channel system that can take vast amounts of water flow (see Of Karst! – episode 2).
The schematic figure below shows how this affects water flow in a karst system. At the surface, water may flow for some distance (external runoff towards the recharge area or internal runoff within the recharge area), before it reaches a dissolution widened vertical crack or fissure. On its way, part of it may slowly infiltrate into the soil but the stronger the rainfall event, the more water will infiltrate quickly into cracks and fissures after being redistributed laterally. Consequently, slow and quick infiltration will be followed by slow and quick vertical flow through the vadose zone. The former through the carbonate rock matrix, the latter through the interconnected system of dissolution caves. Finally, recharge and groundwater flow take place, again quickly through the caves and slowly through the matrix. When passing the system through the cave network, water can enter and leave the system within several hours. When taking the slow and diffuse path, the transit through the system may take months to years.
Because of this behavior, hydrogeologists often speak about the Duality of Karstic Groundwater Flow and storage, although it is known that there is a wide range of dynamics between quick flow through the caves and slow flow through the matrix and that lateral redistribution between the interconnected caves and the matrix takes place at almost every part of the system.
Figure 1: Schematic description of karstic groundwater flow and storage (Hartmann et al., 2014; modified)
A rather uncomfortable lesson on quick flow processes in karst was learned by a group of school students on a trip through a karstic cave in Thailand. Due to the quick recharge processes explained above, the groundwater tables could quickly rise blocking the return path of the group and resulting in a dramatic rescue mission:
In order to predict the impact of interplay of quick and slow karstic groundwater processes on cave water levels or water resources in general, karst-specific simulation models are necessary. If you are interested in those, follow the Water Underground blog’s postings and wait for Of Karst! Episode 5, which will introduce karstic groundwater modelling.
While climate change makes dramatic changes to weather and ecosystems on the surface, the impact on the world’s groundwater is likely to be delayed, representing a challenge for future generations.
Groundwater stores are replenished by rainfall at the surface in a process known as “recharge”. Unless intercepted by human-made pumps, this water eventually flows by gravity to “discharge” in streams, lakes, springs, wetlands and the ocean. A balance is naturally maintained between rates of groundwater recharge and discharge, and the amount of water stored underground.
When changes in climate or land use affect the rate of groundwater recharge, the depths of water tables and rates of groundwater discharge must also change to find a new balance.
The time it takes for this new equilibrium to be found – known as the groundwater response time – ranges from months to tens of thousands of years, depending on the hydraulic properties of the subsurface and how connected groundwater is to changes at the land surface.
Estimates of response times for individual aquifers – the valuable stores of groundwater which humans exploit with pumps – have been made previously, but the global picture of how quickly or directly Earth’s groundwater will respond to climate change in the coming years and decades has been uncertain. To investigate this, we mapped the connection between groundwater and the land surface and how groundwater response time varies across the world.
The long memory of groundwater
We found that below approximately three quarters of the Earth’s surface, groundwater response times last over 100 years. Recharge happens unevenly around the world so this actually represents around half of the active groundwater flow on Earth.
This means that in these areas, any changes to recharge currently occurring due to climate change will only be fully realised in changes to groundwater levels and discharge to surface ecosystems more than 100 years in the future.
We also found that, in general, the driest places on Earth have longer groundwater response times than more humid areas, meaning that groundwater stores beneath deserts take longer to fully respond to changes in recharge.
In wetter areas where the water table is closer to the surface, groundwater tends to intersect the land surface more frequently, discharging to streams or lakes.
This means there are shorter distances between recharge and discharge areas helping groundwater stores come to equilibrium more quickly in wetter landscapes.
Hence, some groundwater systems in desert regions like the Sahara have response times of more than 10,000 years. Groundwater there is still responding to changes in the climate which occurred at the end of the last glacial period, when that region was much wetter.
In contrast, many low lying equatorial regions, such as the Amazon and Congo basins, have very short response times and will re-equilibrate on timescales of less than a decade, largely keeping pace with climate changes to the water cycle.
Despite both being in the UK and existing in the same climate, they have distinctly different hydraulic properties and, therefore, groundwater response times. Chalk responds in months to years while the sandstone aquifers take years to centuries.
In comparison to surface water bodies such as rivers and lakes which respond very quickly and visibly to changes in climate, the hidden nature of groundwater means that these vast lag times are easily forgotten. Nevertheless, the slow pace of groundwater is very important for managing freshwater supplies.
The long response time of the UK’s Permo-Triassic sandstone aquifers means that they may provide excellent buffers during drought in the short term. Relying on groundwater from these aquifers may seem to have little impact on their associated streams and wetlands, but diminishing flows and less water could become more prevalent as time goes on.
This is important to remember when making decisions about what rates of groundwater abstraction are sustainable. Groundwater response times may be much longer than human lifetimes, let alone political and electoral cycles.
Mark Cuthbert is a Research Fellow and Lecturer in the School of Earth and Ocean Sciences, at Cardiff University in the United Kingdom. Mark’s work currently focuses on coupled hydrological-climate process dynamics in order to: understand groundwater sustainability; improve interpretations of terrestrial paleoclimate proxy archives; and understand how Quaternary paleoenvironments influenced human evolution.
Kevin Befus leads the groundwater hydrology group in the Civil and Architectural Engineering Department at the University of Wyoming. With his research group, he studies how groundwater systems respond to hydrologic conditions over glacial timescales and in mountainous and coastal environments.
Tom Gleeson leads the Groundwater Science and Sustainability group in the Civil Engineering Department at the University of Victoria. His research interests include groundwater sustainability, mega-scale groundwater systems, groundwater recharge and discharge and fluid flow around geologic structures. Tom is also the founder of this blog, WaterUnderground.
Drought is in the news here in New South Wales, Australia. But how are rainfall, drought and groundwater related?
First, we need to understand what drought is. Is it a water shortage? Or a lack of rainfall? Or something else? In the USA, the National Climatic Data Center define drought as the ‘absence of water’. They identify four types of drought: 1) meteorological drought (a lack of rainfall), 2) hydrological drought (a loss of surface water or groundwater supply), 3) agricultural drought (a water shortage leading to crop failure), and 4) socioeconomic drought (where demand for water exceeds availability).
Let’s look at New South Wales again. It covers both coastal regions, such as Sydney (where I am writing this), as well as a vast interior (where most of my research is based). The Bureau of Meteorology produces meteorological drought maps based on rainfall amounts over recent months. The current map shows large areas of New South Wales are experiencing rainfall totals that are in the lowest 10 percentile (‘serious’), lowest 5 percentile (‘severe’) and the lowest on record.
How does this deficiency in rainfall affect groundwater? And is there a groundwater drought? Long-term measurement of groundwater levels in boreholes (also called wells, depending on your country) can tell you whether water levels are rising or falling. Wells integrate groundwater recharge that comes from both surface water (e.g. rivers that lose water through their base) and from rainfall (also called diffuse recharge).
Real-time data of water levels from telemetered boreholes can provide timely information on groundwater drought (for example, here for NSW). Satellite products such as GRACE, which can infer groundwater levels from small changes in gravity over time, can provide large scale spatial coverage. Modelling products can calculate water balance from meteorological, soil and land use data.
The current Bureau of Meteorology map shows that deep soil moisture is very much below average across New South Wales. If we assume that deep soil moisture levels are only determined by rainfall recharge, then from this we would expect no rainfall recharge of groundwater to be occurring over large parts of New South Wales. From one location, Wellington, close to the middle of the drought region, we have the measured evidence from inside a cave that shows that rainfall recharge hasn’t occurred for 18 months (and counting).
Drip rates have been on the decline since the winter of 2016. But note the decline temporarily slowed in 2017, starting in early April. That is the response to the last time there was rainfall recharge there – owing to almost 70 mm of rain falling over three days in late March 2017. Eighteen months ago.
In the inland of New South Wales, it is clear that in dryland farming regions, the lack of rainfall has now led to an agricultural drought. In contrast, latest available data from our groundwater monitoring networks shows that there is currently no decline in groundwater levels in the major irrigation districts, which is where river recharge occurs. But for our dryland farmers, and ecosystems that rely on rainfall recharge, the karst drip data show that the groundwater drought has hit. Australia is often called a country of drought and flooding rains. Flooding rains are what we need next so that we also have some river recharge to replenish our groundwater resource.
Wellington, NSW. July 2018. This is the UNSW Research Station, normally stocked and cropped, but not this year.
Post by Elco Luijendijk, a junior lecturer, and David Hindle, lecturer and head of geodynamic modelling, both at the Department of Structural Geology and Geodynamics at the University of Göttingen, in Germany.
As we’ve firmly moved into 2018, we can say happy 50th birthday to one of the most revolutionary scientific theories of the last century: plate tectonics. Here we discuss the birth of plate tectonics and what it means for hydrogeology.
Plate tectonic theory explains the how the Earth’s surface is formed and how it consists of rigid plates on top of a layer that is called the asthenosphere and that behaves like a slow-moving liquid. The plates move around, collide and subduct beneath each other. Plate tectonics successfully explains many features of the surface of the Earth, such as mountain belts at the collision zones of plates, ocean basins at places where plates move apart and the concentration of earthquakes near plate boundaries. For instance it is quite easy to recognize the boundaries of tectonic plates if you look at the earthquake distribution in Figure 2.
Plate tectonics birthday cake, showing one tasty tectonic plate (left) subducting below another (right). Source: http://sara-geologicventures.blogspot.de/2012/05/cake-subduction-zone.html
Actually, depending on your definition either 2017 or 2018 is the 50th birthday of plate tectonics. The story why this is the case is a bit complex. Jason Morgan first presented the theory at meeting of the American Geophysical Union (AGU) in 1967. However, the first paper on the mathematical principle of the movement of tectonic plates was published in the same year by McKenzie and Parker (1967). Jason Morgan’s paper (Morgan 1968) is the first one to clearly demonstrate the global geometry of all the major tectonic plates, but had got delayed by peer-review for over a year. The development of plate tectonics involved many scientist and several earlier theories, such as seafloor spreading (which showed that ocean basins were split in two halves that were moving apart). There are surprisingly few books available on the history of plate tectonics, but one that is definitely an enjoyable read is “Plate Tectonics: An Insider’s History Of The Modern Theory Of The Earth” (Oreskes 2003). It is a fascinating collection of stories by most of the scientist that were involved in the development of the theory.
Figure 2 Plate boundaries on earth, with earthquakes > M6.5, since the year 2000, and with selected relative motion arrows for plate pairs – the motions shown are always those between adjacent plates. Double arrows imply spreading – moving apart of plates, mostly on oceanic ridges, while single arrows imply either strike slip motion (California and the San Andreas fault for instance) or convergence (either subduction of an oceanic plate under a continental one – under the Andes mountains in South America as an example, or collision of two continental plates as between India and Eurasia in the Himalayas for instance). Earthquakes are clearly concentrated on plate boundaries. This map was made using GMT (http://gmt.soest.hawaii.edu/).
Ok, that is all very interesting, but you could ask the question: what does plate tectonics have to do with Water Underground?
In some regards not much. We can often ignore plate tectonics when looking at groundwater flow. Hydrogeologists tend to study groundwater supply and pollution on human time and space scales. Because plates move very slowly (up to tens of mm per year), on short timescales the subsurface can be regarded as static layer of rocks that does not move or deform. However, most of the groundwater on our planet is old, and has infiltrated to the subsurface ten thousand years ago or earlier (Jasechko et al. 2017). The oldest groundwater that we know is 1.5 billion years old and was found at 2 km depth in a mine in near Timmins, Canada (Holland et al. 2013). Over its long history it was part of ancient and long disintegrated continents and the plate that holds this water moved from an area south of the equator to its present position.
Plate tectonics affect groundwater. Especially in deeper (several kilometers) parts of the crust, the groundwater pressure, salinity and composition that we encounter today are often the result of a long geological history. Over time, sediments were added and removed by erosion, layers were compacted, folded and/or faulted, which affected groundwater flow and its interaction with the rocks that contain it.
The reverse is also true: groundwater affects plate tectonics. This is perhaps most important near mid-ocean ridges, where two plates move apart, and new crust is being added to these plates all the time. There is abundant evidence for strong circulation of seawater through the subsurface, which cools the hot new crust, reacts with the rocks around it and changes the chemistry of the crust and the ocean. The most visible evidence are so-called black smokers (Figure 3), where hot (350 ˚C) water discharges into the ocean through fissures in the crust and carries along black plumes full of dissolved minerals. At the opposite end of the plates, the presence of water underground changes how easy or hard it is for one plate to subduct beneath another in a plate collision zone, as was discussed at a recent AGU conference (link to session), 50 years after the AGU conference where Jason Morgan presented his theory. On a smaller scale, faults that enable the stacking of rocks in plate collision zones (mountain belts) or the breaking apart of rocks in rift zones (where plates split up), are dependent on the presence of groundwater. Even before the advent of plate tectonics Hubbert and Rubey (1959), showed that water in fault zones can act as a kind of lubricant that enables two adjacent blocks of rocks to move past each other. Because this movement gives rise to earthquakes, groundwater may also play an important role in the earthquake cycle. This role is still heavily debated and is researched by drilling deep wells in faults at plate boundaries, such as at the San Andreas fault in California (Zoback et al. 2010) or the Nankai through (Hammerschmidt et al. 2013).
Without sufficient groundwater plate tectonics may not exist on our planet. The movement of tectonic plates depends on how easily the rocks below these plates can deform. At these depths, high pressures and temperatures promote the slow deformation of the crystals that make up the rocks at this depth. The mechanisms that cause the deformation of crystals are termed “creep”. Whether or not the rock contains water (in the form of -OH groups) affects creep: generally, “wet” minerals are up to a factor of 10 “softer” than “dry” ones. The actual physics and chemistry of how -OH affects and weakens different minerals is not entirely clear. Creep is also essential for the convection of the earth’s mantle, which controls the escape of heat from our planet’s interior and provides the energy to drive plate tectonics. Without convection, there would be no plate tectonics, so the presence of water throughout the earth’s crust, and its continued reintroduction to the earth’s mantle by the subduction of tectonic plates seems to be a key component driving the system, or at least, helping it to keep moving along.
There are many more links between groundwater and geologic processes, too many to cover in a short blog item like this. However, the current state of our understanding is summarized in a highly recommended book “Groundwater in geologic processes”. Many aspects of groundwater flow and its links with geological processes in newly formed, colliding or subducting plates are still uncertain and studied by hydrogeologists, which means that 50 years after the publication of the theory of plate tectonics, many discoveries still lie ahead.
Figure 3 A black smoker at the mid Atlantic ridge emitting hot groundwater into the ocean from newly formed oceanic crust. Copyright: MARUM – Center for Marine Environmental Sciences, University of Bremen.
David Hindle (L) is a lecturer and the head of geodynamic modelling in the Department of Structural Geology and Geodynamics at the University of Göttingen, and Elco Luijendijk (R) is a junior lecturer also in the Department of Structural Geology and Geodynamics at the University of Göttingen.
The Macquarie River, Wellington, Australia, a major tributary of the Murray-Darling Basin. The river is recharging the groundwater, but how far into the research bore field can we detect elevated concentrations of organic matter?
Post by Andy Baker, Professor researching groundwater, caves, past climate, organic carbon and more at the University of New South Wales, in Australia.
We know a lot about the carbon cycle, right? Increased carbon dioxide emissions since the Industrial Revolution have perturbed the carbon cycle. This has led to rising atmospheric carbon dioxide levels and climate change.
Not all this extra carbon accumulates in the atmosphere as carbon dioxide. Carbon sequestration is also occurring, for example in the oceans and terrestrial biosphere. All the carbon fluxes and stores on the planet must balance. In recent years there has been a hunt within the terrestrial system to quantify some missing carbon, such as the particulate organic carbon in river systems and dissolved organic carbon in glaciers.
So, what about groundwater? Could this be a previously unrecognised source or sink of carbon? We already know that the global volume of groundwater of 1.05 x 1019 litres is the world’s biggest source of freshwater. But groundwater natural organic carbon concentrations are low: typically, 1 part per million (ppm). This means that the global groundwater organic carbon store is just 10.5 x1015 g. For comparison, rivers are estimated to sequester this amount in just four years. Basically, there’s no significant store of organic carbon in groundwater.
But hold, on, this raises another puzzle, which is: where has all the organic carbon gone? Groundwater is recharged from rivers and from rainfall. Rivers have much more dissolved organic carbon than the 1 ppm found in groundwater. And the recharge from rainfall passes through the soil. And soil leachates also have much higher dissolved organic carbon concentrations than groundwater. So, despite the high concentrations of organic matter in the soil and rivers, most of this organic matter is ‘lost’ before reaching the groundwater. Is it biologically processed (and therefore a potential source of carbon dioxide)? Or is it sorbed to mineral surfaces (and therefore a potential sink of carbon)? Most likely, both processes occur in competition.
Groundwater organic matter: a carbon source or sink? We don’t know. But a few groups are working on the puzzle. For example, our group at UNSW Sydney is collecting groundwater samples and measuring organic carbon sorption to minerals, and microbial use. In the USA, groundwater data has been mined to understand the rate of loss of organic carbon in groundwater. This December, river and groundwater experts come together at the AGU Fall Meeting to share our understanding. Not least because surface and groundwater are interconnected systems.
Collecting groundwater samples to understand whether organic matter is a carbon source or sink. Long field days at the UNSW Wellington Research Station mean the final sample is often collected at dusk.
Andy Baker is the Director of Research and UNSW’s School of Biological, Earth and Environmental Sciences. His research interests include hydrology, hydrogeology, cave and karst research, paleoclimatology, and isotope and organic and inorganic geochemistry. You can find out more information about Andy at any of the links below:
Dedicated Water Underground readers know that this blog is not just about water science, but also some of the more cultural impacts of groundwater. Keeping in that tradition, today’s post begins with a joke*:
Your neighbor who?
Your neighbor’s groundwater, here to provide water for your plants!
Figure 1. Typical reaction to joke written by the author.
Perhaps this joke needs a little explanation. As we’ve covered before, groundwater is important not just as a supply of water for humans, rivers, and lakes, but also because it can increase the water available to plants, making ecosystems more drought resistant and productive. However, we also know that groundwater moves from place to place beneath the surface. This means that human actions which affect groundwater in one location, like increasing the amount of paved surface, might have an unexpected impact on ecosystems in nearby areas which depend on that groundwater.
Imagine, for example, two neighboring farmers. Farmer A decides retire and sells his land to a developer to put in a new, concrete-rich shopping center. Farmer B continues farming her land next door. How will the changes next door affect the groundwater beneath Farmer B’s land, and will this help or hurt crop production on her farm?
In a new study, my colleagues and I explored these questions using a series of computer simulations. We converted different percentages of a watershed from corn to concrete to see what would happen. Our results showed that the response of crops to urbanization depended on where the land use change occurred.
Figure 2. Conceptual diagram showing how urbanization might impact crop yield elsewhere in a watershed. From Zipper et al. (2017).
In upland areas where the water table was deep, replacing crops with concrete caused a reduction in groundwater recharge, lowering the water table everywhere in the watershed – not just beneath the places where urbanization occurred. This meant that places where the ecosystems used to be reliant on groundwater could no longer tap into this resources, making them more vulnerable to drought. However, places where the water table used to be too shallow saw boosts in productivity, as the lower water table was closer to the optimum water table depth.
In contrast, urbanization happening in lowland areas had a much more localized effect, with changes to the water table and yield occurring primarily only in the location where land use changed, because the changes in groundwater recharge were accounted for by increased inflows from the stream into the groundwater system.
So, what does this mean for the neighboring farmers we met earlier?
For Farmer A, it means the neighborly thing to do is work with the developers to minimize the effects of the land use change on groundwater recharge. This can include green infrastructure practices such as rain gardens or permeable pavement to try and mimic predevelopment groundwater recharge.
For Farmer B, the impacts depend on the groundwater depth beneath her farm. If the groundwater beneath her farm is shallow enough that her crops tap into that water supply, she should expect changes in the productivity of her crops, especially during dry periods, and plan accordingly.
*Joke written by scientist, rather than actual comedian.
Zipper SC, ME Soylu, CJ Kucharik, SP Loheide II. Indirect groundwater-mediated effects of urbanization on agroecosystem productivity: Introducing MODFLOW-AgroIBIS (MAGI), a complete critical zone model. Ecological Modelling, 359: 201-219. DOI: 10.1016/j.ecolmodel.2017.06.002
Sam Zipper is an ecohydrologist. His main research focuses broadly on interactions between vegetation and the water cycle, with a particular interest in unintended or indirect impacts of land use change on ecosystems resulting from altered surface and subsurface hydrological flowpaths. You can find out more about Sam by going to his webpage at: samzipper.weebly.com.
The process of receiving or giving systematic instruction, especially at a school or university.
from Latin educatio(n-), from the verb educare
Educare is a combination of the words e (out) and ducare (lead, drawing), or drawing out.
Based on this definition, I should change the title of this post to: Drawing out groundwater (from the well). This is actually the main occupation of groundwater scientists, isn’t it? Not only are we always withdrawing groundwater from a well or a borehole while sampling, but we also often have to “draw it out” when dealing with managers and policy makers, as sometimes they seem to forget about this hidden (but very important) component of the water cycle. Therefore, we are quite used to these forms of “drawing out” – but what about education? Are we really that effective in “drawing out” groundwater in explaining its peculiarities, issues, and connections within the whole water cycle and, more generally, with the environment?
Indeed, the effort of shedding light on something that is not so visible nor easily studied has the side effect of forcing us to focus solely on it, with a resulting tendency of developing sectorial approaches to water management.
In the preface of a UNESCO Technical paper, I found the following excerpt: “Water resources schemes are now increasingly considered as integrated systems and consequently, civil engineers, geologists, agricultural engineers and hydraulic engineers engaged in planning and design no longer work in isolation”. The document is dated 1974 but, still in 2017, we are somehow struggling to fitting groundwater into Integrated Water Resources Management (IWRM) and to connecting mental and structural “silos”. Quoting Daly (2017), the latter is particularly relevant (especially when education is at stake): if on the one hand, specialization can be the driver for a sound knowledge; on the other hand, this can encourage people to get stuck in their own individual disciplines (or said in other words, their “silos”). Indeed, “silos” exist in their structures, but can also exist as a state of mind that can go hand in hand with tunnel vision (Tett, 2015).
Therefore, in my opinion, the new generation of groundwater scientists (and teachers) should have a new mission: to work (and therefore, to teach) coherently with the integrated and complex nature of the water cycle. In fact, the role of hydrogeologists and groundwater scientists in times of increasing freshwater demand, exacerbated by population growth and climate change effects, requires a serious shift towards a more holistic approach targeting sound groundwater assessment and long-term management.
Arguably, if we are still discussing possible ways of practically implementing this integration, we should definitely start asking ourselves if the the “business as usual” way of working and teaching is effective. If it is not, we must begin investigating how we can go beyond classical approaches to draw groundwater out of the well.
Playing with kids while sampling … can we call it capacity building?!
My humanities colleagues and friends are always talking about postmodernism or pomo for short (see this funny satire). I’ve been thinking a lot lately about ‘modern groundwater’ (stay tuned for a cool paper), so I started wondering if there is ‘postmodern groundwater’.
Modern groundwater is groundwater recharged since the huge spike it tritium in the early 1960’s due to above ground thermonuclear testing. But tritium has a short half-life, so atmospheric tritium concentrations have largely decayed back to pre-bomb spike concentrations (see graph below). So does this mean that we are in the age of postmodern groundwater or pomo gw?
Tritium concentrations in precipitation through time (from USGS)
Googling ‘postmodern groundwater’ comes up with nothing, so maybe I’m on to something new. The only thing online that is close seems to be Michael Campana’s more political and very interesting ‘Postmodern water cycle” shown here:
The Postmodern Water Cycle by Kate Ely, Umatilla Basin hydrologist extraordinaire for the Confederated Tribes of the Umatilla Indian Reservation, as posted on waterwired.
As more people move to urban areas, cities around the world are experiencing increased water stress and looking for additional water supplies to support their continued grow.
The first global database of urban water sources and stress, published online this week in Global Environmental Change, estimates that cities move 504 billion litres of water a distance of 27,000 kilometers every day. Laid end to end, all those canals and pipes would stretch halfway around the world. While large cities occupy only 1% of the Earth’s land surface, their source watersheds cover 41% of that surface, so the raw water quality of large cities depends on the land use in this much larger area.
An international team of researchers from nine institutions, including McGill University in Montreal, surveyed and mapped the water sources of more than 500 cities, providing the first global look at the water infrastructure that serves the world’s large cities. The study was led by Rob McDonald, senior scientist with the Nature Conservancy in Arlington, Va.
Prof. Bernhard Lehner and PhD student Günther Grill of McGill’s Department of Geography contributed a detailed global map of rivers, lakes and watersheds to help map the water sources of each city, while Prof. Tom Gleeson of McGill’s Department of Civil Engineering conducted analysis for groundwater sources.
The research team used computer models to estimate the water use based on population and types of industry for each city, and defined water-stressed cities as those using at least 40% of the water they have available. Previous estimates of urban water stress were based only on the watershed in which each city was located, but many cities draw heavily on watersheds well beyond their boundaries. In fact, the 20 largest inter-basin transfers in 2010 totaled over 42 billion liters of water per day, enough water to fill 16,800 Olympic-size pools.
There is good news in the findings: Many cities are not as water-stressed as previously thought. Earlier estimates put approximately 40% of cities into the water-stressed category. This analysis has the number at 25%.
The study finds that the 10 largest cities under water stress are Tokyo, Delhi, Mexico City, Shanghai, Beijing, Kolkata, Karachi, Los Angeles, Rio de Janeiro and Moscow. (Neither of the two Canadian cities analyzed — Toronto and Montreal — was water-stressed, according to the definition used in the study.)
The study also makes clear the extent to which financial resources and water resources are intertwined. It is possible for a city to build itself out of water scarcity — either by piping in water from greater and greater distances or by investing in technologies such as desalinization — but many of the fastest growing cities are also economically stressed and will find it difficult to deliver adequate water to residents without international aid and investment.
“Cities, like deep rooted plants, can reach a quite a long distance to acquire the water they need,” says McDonald. “However, the poorest cities find themselves in a real race to build water infrastructure to keep up with the demands of their rapidly growing citizenry.”
The study also reveals that:
Four in five (78%) urbanites in large cities, some 1.21 billion people, primarily depend on surface water sources. The remainder depend on groundwater (20%) or, rarely, desalination (2%).
The urban water infrastructure of large cities cumulatively supplies 668 billion liters daily. Of this, 504 billion liters daily comes from surface sources, and that water is conveyed over a total distance of 27,000 km.
Land use in upstream contributing areas affects the raw water quality and quantity of surface water sources.
An estimated one-quarter of large cities in water stress contain $4.8 trillion of economic activity, or 22% of all global economic activity in large cities. This large amount of economic activity in large cities with insecure sources of water emphasizes the importance of sustainable management of these sources, not just for the viability of individual cities but for the global economy.
The research was supported by a grant from the Gordon and Betty Moore Foundation.
Historic 1977 photo of Dr. Joseph Poland, USGS, considered the pioneer of scientific subsidence studies. Dates on telephone pole indicate previous land elevations in an area SW of Mendota. Photo credit: U.S. Geological Survey
Next time you eat food grown in the San Joaquin Valley of California, think about this: the water used for growing them probably came from under ground. Farmers do not really have a choice because the amount of water from rain and snow can’t keep up with the needs for growing food. Every year more water is drawn out of the ground for irrigation. Because of this, the floor of the San Joaquin Valley goes down as the sediments compact once the water is out (see picture on right).
In the latest work from our team, we find a surprising side effect of groundwater pumping: the mountains surrounding the valley (the Sierra Nevada and California Coast Ranges) are moving up a few millimeters each year, as shown by a large number of GPS instruments. This may seem very small to humans, but for hard rocks it is quite fast. We find that this uplift can be explained by the loss of water out of the ground, as shown by gravity data from the GRACE satellite. The water lost through irrigation lowers the weight on the Earth’s crust, which responds by bouncing back up like a spring.
One interesting implication of this study is the impact on earthquakes on the San Andreas Fault. Uplift of the crust (and mountains) decreases the grip on the fault, making it easier to slip and cause small earthquakes during busier times of groundwater pumping. Perhaps more important, our study shows that humans can really move mountains through industrial agriculture. In California, this effect may get worse because more droughts, earlier snowmelt and different rainfall patterns are expected due to climate change.
This article is the second in a series of plain language summaries on Water Underground (link to first). The 5upgoer word processor showed that ~80% of the words in this post are in the 1000 most common words in the English language. For recent news coverage of this article check out this.