Groundwater depletion and Global Food Security: An interview with Marc Bierkens
Groundwater depletion is often seen as a regional problem, the people who will suffer are those who will be directly affected by reduced agricultural yields and drinking water shortages. When looked at through a whole-system lens, however, a water shortage in one of the world’s major agricultural regions could drive up food prices, reduce exports, and otherwise mess up the whole world’s food supply.
I had an opportunity to interview two very knowledgeable people for a story on groundwater depletion I wrote for the American Geophysical Union. (The story also appeared, albeit with a much narrower focus, as a press release).
The story of food and water security is an extremely important one, and I personally believe that water scarcity issues will be one of the dominant controversies in my lifetime. ***
Colin: I was wondering if you looked at that other study that I was talking about?
Marc: Yeah I just looked at it. It’s a study by Paolo D’Odorico showing that if you allow food globalization and a virtual water trade, in the end, you are more prone to drought. I didn’t know that study, so for me it’s a new insight.
C: What I’m sort of hoping to do is – because the two studies seem to go side-by-side really well.
M: I think partly, in the sense that – well of course there is connection. If you look at, for instance, water table drops in the central US what you see there is that a lot of the crops grown there are either fuel crops these days, but they used to be cash crops like corn, wheat, and soybeans which were exported to other parts of the world, for instance Europe. Which means that you have a virtual water flux from, in this case the central US to Europe, that causes over-exploitation of groundwater locally. So that’s the connection, basically. So I can see that there is connection. I know a little about virtual water trade, because there is a professor in the Netherlands called Arjen Hoekstra who actually did a lot of work on that topic, calculated a lot of virtual water flows between countries. And I know him pretty well, so I know his work. I know what the idea of virtual water trade is, and in this sense, it can both enhance as well as mitigate groundwater depletion. The example I gave you here is basically that if food is exported, and you grow that food with an over-exploitation of the groundwater to irrigate, then basically it worsens groundwater depletion because you’re not only using it to feed local people but also export it in terms of virtual water. But for instance in India and Pakistan, which are the hot spots of groundwater depletion in the world, what you can see there is that, actually, this is used to feed the local population because the enormous population growth. So, in that case, if they were the import food, suppose if India generates money by other stuff, by making electronics or so on, and they spend that money to buy food for instance from Brazil, then that import of food will make the growing of food locally less necessary. So this then mitigates groundwater depletion. So yes, by these two examples you can see that globalization of trade and groundwater depletion are in fact very much connected.
Okay, long story, but I think this is true. That’s the way these two papers are connected basically. But to be honest when we did this study we were not contemplating relations with global water trade yet, at this point. We were just curious about what’s the rate of depletion of groundwater in most parts of the world. We knew only of local studies up until now, and we wanted to have a global perspective.
C: Right. So how did groundwater depletion, because the study was looking at a big span of time. So how did the depletion rate change over the study period?
M: Well, the rate itself, the rate of depletion, changes almost linearly from the 1960s to the early 1990s, or half of the 1990s. And then you see a sharp increase which is related to the increase of upcoming economies and population numbers in India, mainly. Also a little bit Brazil, but mostly India and China. Although population numbers are not growing, there is an increase in the rate of industrialization that also uses water. And also, as the development level goes up then people use more water. So both China and India, and also Pakistan actually, are huge groundwater users. And the US. So you see, especially in those Asian countries you see a sharper increase in the last decade. But, apart from that we come up with a steady increase in the groundwater depletion rate from the 1960’s on to now, except for this increase in rate in the last decade or so. So that’s basically what comes out of the analysis.
C: So you were mentioning that it was population growth and industrialization. What do you think was the biggest reason for this increase in depletion?
M: The reason is that once population starts to grow people start to live in locations where there is less precipitation, and if you want to feed those people you need to find water somewhere else. And if you have aquifers that are filled with groundwater that are currently not being recharged in these dry climates, but were recharged in previous periods, sometimes thousands of years ago when these regions had a wetter climate, then people will use the groundwater. I think groundwater… first they use it for drinking water, and later on they discover that they can irrigate with it. The level of development goes up, so they get some money to buy pumps and they start pumping. So you get thousands of little pumps pumping up groundwater for irrigation, which again enhances development and population numbers, and population numbers needs more food so it’s sort of a run away train. But of course a very dangerous one because in the end it cannot be sustainable. The groundwater starts to drop. It will eventually be at a level so low that a regular farmer with his technology cannot reach it anymore, and it’s not being recharged in these areas because recharge rates are quite low on average, so this is basically an accident waiting to happen. But right now it’s basically mitigating water stress because you’re living on loaned money, in this case loaned water. It’s a water debt that you build up because these aquifers are not being recharged and this enables you to raise your standard of living. I don’t want to insult you, but it sounds a little bit like how some of the people in the US and Europe live when it comes to money.
C: I know that there is another study, I think it was in 2008, the GRACE satellites (Press release, Nature letter), when NASA was suggesting that groundwater was depleting as well. But they weren’t able to give any sort of clues of how much was in the aquifer. Like if we know the depletion rate, but we don’t know how much was there to start…
M: It’s the same. It’s the same in our case. We do know how much is being depleted because we can calculate how much is recharging. We have a database of country statistics that tells us how much groundwater each country is abstracting, we can basically downscale that using smaller scale data on where the people live and where the industry is and so on. We can use these sorts of things to downscale the abstractions, then we subtract until we get a depletion rate. This depletion rate, we know that it comes from the aquifers, but we do not know exactly how much is in there. We only know at what rate it is being depleted. The reason is that there are only a few places in the world where you have extensive hydrogeological surveys. The US is one of them, where we know the thickness of the aquifer and also – if you go down deep enough, often, especially if you’re on the coast or in [??] eventually you hit brackish or salt ground water. So there is a limited store of course, but how much? One does not know. But of course it’s not always necessary to know exactly groundwater there is because you can calculate easily how long it takes for this groundwater to be so deep that a common farmer with a day-to-day pump cannot get it up anymore, and run into trouble. And you can also calculate what the effect will be on rivers and wetlands if it falls below the lowest point in the landscape at a certain distance where the wetlands are situated you know this wetland is going to dry up.
C: So do we have any sort of clue when we might start seeing some of those effects?
M: No. Well, that’s very locally determined. I mean, I could not say something about that globally yet. And actually, we’ve just submitted a proposal to our own National Science Foundation, the Dutch one, basically trying to figure out these two questions; At the rate that we are depleting now, when do we hit the boundaries? When do we see that the groundwater tables are too deep to be of any use for a common farmer? So they run into trouble of getting there crops irrigated. And at what place, and when will rivers run dry during the summertime because you don’t have any groundwater discharge anymore? Or when wetlands fall dry? These types of questions are ones we want to answer at a global scale. We’re planning to do research on that, but we haven’t yet.
C: Oh okay. I was mostly thinking about – because it’s sort of the Midwest United States, and the Central Valley in California, and then Northwest India were, it seemed like, the heaviest hit regions.
M: Yes, and also Northeastern China, in the Yangtze region. In northeastern China, northwest of Beijing, that region is also quite extensively hit by this problem.
C: Yeah, but with the Midwest US, I don’t know about China specifically, but the other three are all really big agricultural centres.
C: So is there any sort of indication of when… Because the agriculture will eventually have to cut back if they run out of the groundwater for irrigation. But do we have any clue of when that might start happening? Or might have to happen?
M: Again that is a difficult question because, for instance in the Central Valley of California they have other means of getting water. They get water from, well they have the California aqueduct, which basically brings water all the way from Lake Shasta down to Los Angeles in the end. And also Los Angeles gets water from the Colorado river, so there is an aqueduct and surface water system. These are highly developed countries, so they have the ability to mitigate a lot by technology. They can stretch the problem much longer than a country like India or Pakistan where you have exploding numbers of people, in terms of population growth, and less technology to mitigate it by either looking at other water sources or even looking at technologies of getting more crop per drop, as they say. So increasing water use efficiency.
So in that sense it really depends on where you are, it’s basically socially and economically determined how much you can cope with this problem. I think in the end if people need to be fed in the US, and the country is still rich enough, they would do it like in Saudi Arabia, which means you build big desalinization plants and you use sea water. It costs of lot of energy of course, and it’s expensive, but if people need to be fed and there is money, then they will do it. But of course India and Pakistan cannot do that, also, they are too far from the sea, by the way. So again, I cannot give a definite answer ‘Okay, it’s within 10 years, or 20 years or so.’ I hope to give that answer basically when we finish this study that we just applied for at our own National Science Foundation. So it’s definitely a question that we hope to answer.
C: Perfect. I was just hoping to talk a little bit about the sea level rise finding, too. Because it said in the paper that this was about 25% of the sea level rise rate right now. Will that hold steady? Because it was saying that the sea level rise from this might go up.
M: Yes I think so, because what you see if you look at the trend that we reconstructed, in the last decade you see an increasing rate. On the other hand, the sea level rise has been, I think, counteracted by the storage of water behind big damns like the Three Gorges Dam in China, and so on. But if you look at the rate of dam building, it’s been tapering off over the last decade or so. It’s like an S-curve, right? It’s a sigmoid curve, and it’s on its last part of that, so you see it tapering off. But this is basically increasing the groundwater depletion. So, I believe that the abstraction of groundwater that is now stored in aquifers on land, most of it will eventually end up in the ocean, that effect on sea level rise will increase. But of course it will never lead to sea level rise that will endanger cities and so on, comparable to the melting of Greenland or Antarctica, it’s not of that level. But it is still 25% of the current rate, which means that you have to take account of it if you want to observe sea level rise as a sign of what happened to the glaciers and to the ice caps. You have to account for that groundwater depletion signal as well, and it’s actually as large as the signal due to the ice caps or the glaciers individually.
C: Just the way that I was looking at this study, when taken with the other study, is you were talking about how it’s mostly going to see local effects. But when you start thinking about sea level rise and the virtual water trade, it seems like what was a local problem has the potential to affect everybody?
M: Yeah but not the sea level rise. We’re still talking about very little sea level rise right. It’s about 0.8 millimetres per year, so in a century that’s…
C: Not much.
M: Right, not much. It’s about 10 centimetres, so that’s not something that would keep people awake. So it’s not that. But the global water trade, or the global food trade, of course, will be affected because if you have very important trade of food between one country and the other – for instance look at [the Netherlands]. One of the important export products of the Netherlands when it comes to agriculture is pork meat. We have a lot of pig farms, and they are being fed on power fodder, very condensed, nutrient-rich stuff that comes from soybeans, and also sometimes corn. This is important in the Netherlands, but it is also important for countries like the US and Brazil. Now if you have groundwater depletion occurring in Brazil or the US, then that is connected to the export of pork meat from the Netherlands. So the two are connected. If by the depletion the export of food for pigs is hampered, then either we find another source of that food for the pigs, or the whole industry will also collapse in the Netherlands. So there is a global connection to water shortages in one part of the planet, to economic activity and agriculture on another part of the planet. So that’s a very interesting set of connections that are in place right now. And of course this groundwater depletion is influenced by these connections like I told you in the beginning, but it also can have a detrimental influence on current trade. So it does not only have a local effect, but in that sense a global effect through the virtual water trade. That’s of course not only the groundwater depletion, that’s also using surface water, for instance. And it of course has the same effect.
M: It also accounts for that. So yeah, that’s an interesting observation.
C: Yeah, and one of the things that I was thinking about is, basically once this groundwater is used it’s essentially… it’s sort of lost, because it’s no longer fresh water, it becomes salt water if it goes out to the oceans. And then it’s not available.
M: Yep. It can only be regained by stopping the abstraction from the aquifer and letting it be replenished by rainfall. But in certain areas like in Northeastern India but also in Central US, the rate of recharge is quite low, which means that it can take several decades before it is recharged. So it will take more than a generation to bring levels to normal again. It may be even longer depending on where you are. That’s why it’s unsustainable in the end.
C: Do you think that maybe, with the rapid population growth in some of these places, if they’re using… because you were talking about them basically being on loaned water… is the population going to grow to a point, using all of this groundwater that’s not sustainable. Will there be big problems in the future when the groundwater starts to run out? And they’re relying on this virtual water?
M: Yes I think so. It will eventually affect the local economies and even the local food security. When we talk about the US we’ll talk about the detrimental economic effects, which means that this will, in the end, be bad news for rural areas. There will be, when people leave a certain area, depopulate it. I don’t know the English word for it. Anyway, that’s what will happen. But when it comes to upcoming economies like India the effect may even be food stress, hunger, and all the social unrest that comes with it. You can see it coming from a long distance in time. We had a sentence like that in the paper, but it was speculative. It was an assumption based on what we found that in the end, that in India and Pakistan this would in the end lead to – you know if you extend the population, if you let the population grow by extending the irrigated areas using groundwater that is not being recharged then you will run into a wall at a certain point in time, and you will have hunger and social unrest to go with it. So that’s something that you can see coming for miles, but of course we cannot say exactly when that will happen, but I certainly believe that it will. But anyway, it’s not very scientific in that sense because we don’t have any hard data or analysis to corroborate that, it’s just an assumption.
C: Do you know of any programs in place to recognize this and deal with it? And try to slow down the groundwater extraction?
M: Yes, I know that, well I’m not a real local expert, but I do know that in the Central US, in the Olgallala Aquifer, that they have been looking at trying to recharge certain areas by applying artificial recharge from streams, trying to make people use surface water more, and increase water use efficiency. That’s a huge one, that means you are basically trying to get better crop production for less water using all kinds of technological innovations. For instance using drip irrigation instead of flooding, from which you get a lot of evaporation losses. Things like that. Only irrigate when absolutely necessary. By doing soil moisture sensing. These types of techniques of course will decrease water use, and if you don’t increase the acreage then eventually that will increase groundwater recharge and I think, one of the reviewers of our paper actually mentioned that there are certain areas in the Central US that were able to get recharge going again, and they see the groundwater tables rising because of that. So it’s basically artificial irrigation from streams and a higher water use efficiency.
In California they probably have to also start looking to even further increase those things, but also maybe using desalinization as an additional water source. And then of course in California they also have surface water from the northern part of California from the snow that accumulates there in winter. That water is caught behind the Shasta dam, and is used in the Central Valley later on in their summer season. And they can try to optimize that even further than they have already. So in developed countries there are certain things that one can do, but again, this is easier to do in more developed countries than in developing countries because it requires technology.
So I think that’s one, and in the end of course it’s… I’m not a vegetarian myself, but I think if we all became vegetarians then of it would also help because you know that if you grow one kilogram of meat you need ten kilograms of wheat to support that, and the amount of water therefore for a kilogram of meat is much higher than for vegetables or grains. So our diet also has an effect on it, in the end.
C: That was actually all the questions that I had. I was wondering if there was anything that you wanted to add?
M: I think one thing that could be mentioned is that we are actually, well this is the first time this has been done at such a global scale because we had this database of groundwater abstractions that is almost covering all countries. It took quite some time to get that database in order. I think the authors that are from the International Groundwater Resources Assessment Centre, they are the ones who did this. It’s a lot of difficult and tedious work to do that, but in the end it enables you to do studies like this, and we could not have done this study without such a database. So it also stresses the importance of having all of these water use data collectors.
***This story in its original formulation had a bit too much throat clearing at the beginning, which I’ve moved here:
“Given that it was such a ripe topic, the story got picked up widely. If you’re interested (or if you want to see how different outlets cover the same piece of news), see takes by the CBC, Discovery News, National Geographic, Treehugger, the Montreal Gazette, and Daily Climate.
As is always the case when talking to knowledgeable folks, you cover a lot more in an interview than can ever make it into a little story. On that note, I’ve put a transcript of my interview with Marc Bierkens, the lead author of the new study, below. (I can’t link to the study, it’s still a paper in press).
But first, I want to bring attention to something that made me a little bit nervous. The bolded section of this quote ran in the story,
It was an assumption based on what we found that in the end, that in India and Pakistan this would in the end lead to – you know if you extend the population, if you let the population grow by extending the irrigated areas using groundwater that is not being recharged then you will run into a wall at a certain point in time, and you will have hunger and social unrest to go with it. So that’s something that you can see coming for miles, but of course we cannot say exactly when that will happen, but I certainly believe that it will. But anyway, it’s not very scientific in that sense because we don’t have any hard data or analysis to corroborate that, it’s just an assumption.
You can probably see why it made me nervous, I am quoting an assumption the researcher had. But that led me to consider it as part of a bigger issue – Why do journalists interview researchers? Without a doubt, one major aspect is to make sure we don’t misrepresent the science. But another big piece is to try to get the social relevance, the “why should people care?” I think that letting subject-matter experts think out loud a little bit (even if they don’t have a peer-reviewed paper on that exact point) is fine, as long as it is related to their area of expertise. If we weren’t willing to take these tangential points, then the value of the interview is, I think, severely diminished.
Anyway, enough of that. On with the interview……”