 So now the floor is open to questions for the next 15, 20 minutes. So far away. Khamenei. Khamenei Singh at Colorado School of Mines. I had a question that might relate to anything that any one of you were talking about. But I know that this is a workshop in part about remote sensing. But it seems like soft data could potentially inform on water budgets or some of the systems that any of you are talking about. I'm wondering if that's been anything that you've thought about using, or state of the science on terms of using that sort of information? Part, yes. That's what I meant by using natural language processing to start getting a lot of information on a proxies. In fact, I view the remote sensing activity as a proxy source of information as well. So on demand for the extent of pumping and how groundwater levels might be dropping, an interesting proxy that we used in India was to get sales by district of pumps of different capacities and different types. Because the interesting thing that when Matt was, I don't remember. I think Ali asked the question of Matt as to whether there was data availability in India. 10, 15 years ago when we were able to get the data, that was not a problem. But the problem was most of the benchmark stations that they were monitoring for groundwater were in the upper 15 meters. And farmers were pumping at depth. So how do you actually make any use of that data? And what we found was that in the annual surveys, farmers reported whether or not they had bought pumps and at what depth they had installed them, et cetera. So we were able to reach out to industry association for pump manufacturers. And they were able to give us data on the horsepower of the pumps, whether it was a submersible or a centrifugal and so on. And since then, Lorelyn, who's there, she's been also looking at media reports of people reporting anything to do with groundwater quantity or quality, geocoding that, and then comparing that with whether or not there are academic research papers there. And you can see that there's going to be a divergence between those two kind of pieces of information. But it's the beginning. I think thinking about soft information for groundwater characterization, there's a long history, you know, using fishies and so on for stratigraphic information and then mapping it in, but using a more diverse source of information for converging on budgets, I think is a beginning. Yeah, I was also going to just say that geologic cross sections, geologic maps, that might be a good place to look for. Yeah, Jim Butler, Kansas Geological Survey. Let me just build on Dr. Singh's question. And one of the most important soft data sources would be the distribution of pumping wells in these heavily pumped areas because you'll see there are gaps in terms of a spatial coverage. And there's a reason for that. And the reason is regarding the transmissivity of the aquifer, the water quality, et cetera. So we need to really fully utilize that piece of soft information as well. And let me also just ask the first speaker, Manu, about the deep groundwater in the Midwest. There are no shallow aquifers in those areas. Those shallow aquifers that you were showing were associated with streams, channels, and there's just not much. There used to be a shallow aquifer. The whole system used to be shallow, but we have changed that through time. So just something that I wanted to point out. I considered using an India picture rather than the US one, but I had the US one handy. Any other questions? Please. Hi. Sorry, good morning. I see you go from the World Bank. So we have several projects. It's like the irrigation overused the groundwater. And we are trying to manage the irrigation in agriculture so to reduce the groundwater overuse. So in our case, the original water balance is really important because we lead it for the water resources and location management. But in our case, it's like we don't have data for how much groundwater was overused. So actually there are several questions in practice. We are looking for the data available or the available tools or models can really help us to solve those questions. So there are tools available to estimate groundwater use if you don't have those data, but Manu brought up some really big concerns with some of those tools. So typically they'll look at the water demand. So you'll get evapotranspiration, crop coefficients. You can estimate that with satellite data or with ground-based data if you know what crops are being grown. But, and then how much surface water is available. And I don't know if you have that information, but that's another challenging piece to get oftentimes. But then the big challenges are people over-irrigating their crops or are they under-irrigating their crops and that one is really challenging to know. If you have some estimates of soil moisture, you get a sense of if the soil moisture is not changing much over the growing season, then you get a sense of that. But again, it's a very challenging thing to estimate without having that actual data set. Let me add something to you guys on the table here. How about electricity? Is electricity a good surrogate for pumping? Or is there, can you discern the signal between the electricity used in a pump and in a house? I mean, obviously they're using house. Yeah, so we spend quite a bit of time trying to work our way through that. If you assume that the pump efficiency is the textbook efficiency, it's an exact measure. The problem is that the pump efficiencies were all over the place when we tested pumps. But in response to your question, what I was going to say was there's a major pump manufacturer in India called Kidloscar. And we had extensive discussions with them for them to add a water meter into each pump and then have it recorded and be made available through various ways. That's only a $10 or $15 add-on to each pump. And they're willing to do it. Their reaction was there's no market for this pump because it's the question that was brought up on data earlier. Why would someone want to collect that information? The government should, the world bank should, anybody who's trying to manage should. So if it was mandated by the government that all wells have to have pumps which have a meter built into it, they would totally make it. And then the cost would come down because of volume of sales. And then you have direct data on each use, which if you're trying to manage is actually very important because all the surrogates don't really help you. Go ahead, Jim. Okay, this Jim Butler can to survey again. I agree completely with your comments on water use. And here I'm gonna brag a bit about the state of Kansas because we do have all of the high capacity pumping wells are metered, it reported annually and it's subject to regulatory verification. So that sort of data set, we as a community need to exploit so that we can assess some of these methods and get a feel for how folks are operating in the field. So I think on the electricity, using electricity information as a surrogate, one issue is that you have other aspects like subsidy and other things that really completely mask what is actual consumption of electricity for withdrawing. So that's one issue. The other question that I wanted to pose is we do see, or in other words in academia, we talk about conjunctive management. We do have a detailed, like for example, if it is surface water management, reservoir, rule curves and other things available. But on the other hand, conjunctive management in a detailed way, how this is how it should be operated, there is nothing existing, your closure on the management part is really a big challenge. I mean, even in places where you have significant groundwater extraction, you don't have a detailed information on this is how it's going to operate. So an integrated planning doesn't exist at all because groundwater to some extent is appropriated very locally. And so it's a very big challenge. This is Jessica Lawson, my name, Challenge Corporation. Just to circle back to the electricity question and thinking about, and the question from the World Bank a little earlier, speaker, participant from the World Bank, thinking about the international development context in places where we are most data limited are often agriculturally dependent places in many times. If we're talking about Sub-Saharan Africa, there's increasing exploitation of groundwater in off-grid areas. So electricity information is going to be very challenging there because the pumps are either solar or diesel generator powered. It's just yet another challenge in terms of filling in the picture where the data is most needed. Hi, Jim Doborowski, USDA, NIFA. Mark, so how fresh is fresh? And is there a temporal shift to these volumes? And how about a major use scenario that you think of for these continental shelf sands that hold these large volumes? Okay, how fresh is fresh? The magnetotoleric systems image formation resistivity. So we don't know whether some of the more well, I would say we don't know. The wells, offshore wells, the salinities are in many cases under 1,000 milligrams per liter, but they can get up into brackish conditions of 5,000 milligrams per liter. So some water treatment may be required to produce these waters and use them. Both Holly Michaels and I have done calculations that suggest that you could develop these offshore fresh water for 30 years but eventually it's a non-renewable resource and eventually you'll get seawater intrusion. What was your second question? Whether it had temporal variability. Good question. We think these are quite old ground waters, probably a place to see an age, tens of thousands or hundreds of thousands of years old as a result of emplacement over many sea level cycles. And in terms of a major development scenario, coastal mega cities potentially could use these resources in times of extreme drought, thinking about Cape Town or Saupolo that had experienced these recent droughts. I would pitch this as a backup water supply or even for New York City if their aqueduct system failed. But this is expensive. Offshore wells are $5 million, just the exploratory wells and piping to the land and desalinization facilities to maybe clean the water up. Those are all unknowns as well as whether it's got arsenic in it or not or other contaminants. So this is a new area that I think is worth pursuing but there's a lot of unknowns and so far nobody has developed these resources because of the cost, I think. I think I missed it when you were talking. Is the depth to water a kilometer or so? It's less than a kilometer. It's somewhere between 200 to 400 meters depth offshore. Okay, so then my reaction is this, this is a non-starter, sorry. I'll explain why. The energy requirement for desalination of seawater is 120 meters. So if you could basically take ocean water directly and desalinate it, the energy cost of that is already a third of the energy costs that you're going to have for pumping this up. So, and then if you still have to treat it and the well is deeper and more expensive, then economically it won't be competitive. Well, I'm not an expert on the transmission system and pumping system issues but the cost of desalinating brackish water is a lot lower than desalinating seawater, so did you? But I think it's about 30 meters equivalent head for brackish, 120 meters for fully seawater. But if you're pumping from 500 meters, you've already trumped all that. Okay, that may well be. In some of the wells that have been drilled, like off of Florida, there are artesian conditions. So there's, we don't really know what the head conditions are in a lot of these systems, but there, yeah, and on Nantucket where I have experienced that the deep aquifers are under artesian conditions and higher than the shallow aquifers, but that may be the case that it's not universal. We have two questions from online. This is from Francisco Munoz-Ariola from the University of Nebraska. The context was data and uncertainties in numerical modeling and diagnosis and his questions are, how can we identify and use socio-hydrological variables and processes that could lead to address the challenge of meeting water demands? And how could we compare the uncertainties from geophysical data and socio-hydrologic data? Yes, that was the question to Holly because her talk was on that, would that be reasonable or was it just to the spanner? I think that conflict is always there. On the one hand, when we work with groundwater systems and you put a lot of effort into the characterizing the physical system, the criticism always is that the social factors are poorly constrained and then sometimes it's the other way around. With a lot of metering and everything, you probably get a good sense of the withdrawals but some other aspects, ideally you want both and even if you have both, you probably can't get water quality right as Holly showed because it takes a lot less effort and detail to get just a water quantity model right for groundwater but once it comes to water quality and contaminant concentrations and things like that, you really want meter scale, sometimes characterization. Carly, the second question is one more, right? How could we compare the uncertainties from geophysical data and socio-hydrologic data? That's a very site-specific thing. There are some places that are more data-rich from the human side and others that are more data-rich from the physics side. So let's put an end to this panel and go for lunch. I'm sure everybody's very hungry because we did not have enough food at breakfast today. It's my biggest complaint. I'd like to first ask you all to give a big round of applause to the panel.