 Good morning everyone. We're excited to get started with day two of our workshop and thank you all for being here. I think I can speak on behalf of the planning committee that we were very pleased with day one and learned a lot. So thank you all very much for your presentations and your engaged discussion. We're hoping we can keep everything rolling for day two. I just have a few updates on the schedule. We thought we might begin the day as planned. You'll see on your agendas that we'll have a quick recap from Venkat to kick off the day and then we'll go into our panel three discussion at 9.15. It's possible we may end our panel three a little bit early and then go into our breakout session three and we'll go ahead and shorten the breakout sessions just a little bit. I think we probably don't need as much time as we had allotted yesterday. So we'll have about an hour for breakout session three. Then we'll go into a short break so that you all have a chance to refresh your coffee and then we'll do our report back just before lunch at 11.45. We'll then read and convene at 12.45 and go into panel number four. I think that panel number four will wrap up around 1.30 and then we'll head into our last breakout session of the day that we've shortened to about 45 minutes. After that we'll report back again and then have some closing remarks probably wrap up around three if not a few minutes after three. So I hope that makes sense. Everyone is on board with that plan. I wanted to let you know we're also going to have a slight shift to the breakouts again today. We'll be combining the green and the yellow groups. So if you take a look at your badge and you have a green sticker where it says 628 or a yellow sticker, both of those will be going to room 105 which is just out of here to the left and down the hallway a little. So we'll have Antar who's graciously agreed to help moderate that session and we'll just combine those groups into one. The other two groups, the orange group will continue to stay in this room and the blue group will be in room 106 which is on this floor today also just around the corner to the left once you exit this room. And with that I will turn it over to Venkat. Thank you very much. So I wanted to start off with the recap of day one and this is a good time for people to jump in because I have taken some of the points made by people both in the lenary as well as in the panel and the breakout to try to summarize where we are. And again, I do not want to misstate and misspeak. So this is a chance for a little bit of a discussion. So irrigation, one of the first things we started off was irrigation. And even though in this great state of Kansas as pointed out by Jim Butler, there's a lot of data. Most of the world we do not have data, you know, people water without keeping an accurate count. So why is it needed? Because in order to do accurate water balances, irrigation and amount of data is probably a good variable to know rather than trying to guess work it from indirect measurements. Second thing, human factors. There are other things, human factors. So one of them is industrial use. I mean, I do not know if industrial use again reports water usage to some precision. I mean, aggregated water is OK if even they tell you once a year that the consumption, that's fine. But you don't want daily or hourly values. It's probably overload of information. And of course, many of the areas of the world where water is used for drinking brings about a whole bunch of other things. But the real thing is that when you run models, you have to know what you put in a model. And several people alluded to hydrostratigraphy. And many models have good things. And the exploration industry has lots of hydrostratigraphy or stratigraphy, not I shouldn't say hydrostratigraphy. Maybe it's time for this community to start using that subsurface geophysics and other advanced tools to do it. Because remember, mapping hydrostratigraphy is not the same as putting it in the model. Because in the model, you have to put actual variables for hydraulic conductivity, poor distribution, porosity, and all these things. And that's actually another translation involving uncertainty. So speaking of uncertainty, one of the biggest things I thought was, and I got this message clearly, is mismatch of skills. So when you think about it, you're going from process to a model to observations. And again, you can do it other way around. I could say process, observations, models. But I thought this is an easier way to put the model in the middle of everything. So in the process, we are all trying to estimate groundwater recharge. And it's one thing which you probably don't have too many direct observations of. Again, the limited situations, and mostly in the United States, where you probably can have data to monitor the monitor wells, look at lysimeter data, and to see actually how infiltration is contributing to recharge and replenishment of the aquifer. Very limited cases, but you can do it. Now, models, you take all the stuff and put it all into these boxes. Listen, this is all from the web. So I've generously taken these pictures. It's not mine, so. And then observations, most of the time, we're all used to in-situ observation, whether it's river flow, groundwater observing well, et cetera. But now, we have seen this great new era in satellite remote sensing in the last 20 years or so, where you have grace and you have insert. Now, again, this is the problem, is that you want to correlate the process, which happens at very small spatial scales to observations which could happen in scales of over 100,000 square kilometers and then use indirect observations of things like subsidence, which, again, are complicated by lack of complete knowledge of hydrostatigraphy to figure things out. So in some sense, we are dealing with a lot of uncertainty. But also, the other thing is that we have a lot of tools now, especially statistical tools, as well as observational tools, to reduce this uncertainty to some manner. And again, what reduction means or how much is to be achieved is a big question. And there are a whole bunch of other things which are very important in the case of models, that model parameters and a parameter could be the value of your saturated hydraulic conductivity. How do you define saturated hydraulic conductivity for clay loam versus loamy clay? I mean, there are two actual categories of soil. So it's complicated. It's not that straightforward. And then integration of observation. Many people made this point. And I'm going beyond just integration between downscaling of satellite data, but integration of models, observations, and then finally, analysis. You want to come up with analysis, because in the end of the day, science aside, this also has a valuable societal component, as actually Tony showed in a couple of his slides. So oops. What did I do? OK. And the last slide was other issues. Quality of data and access to data, because when we are talking about NGN, we're talking about other countries of the world, as Tony pointed out. We do not have access to in-situ data in some of these countries. And one of the requests made when one of the questions was NGA resources is like, can NGA facilitate this access? And again, I'm sure it will depend on a case-by-case but certainly any investigation in a different country other than the United States will benefit vastly from access to their in-situ data. Second is water quality data. One of the things which was said, most of the talk has focused on water quantity. And water quality data is hard to come by and is hard to obtain except in small select situation. And again, water quality data is a small select thing. It's not going to be at a great scale. I mean, you're talking about water quality in an aquifer. It's probably one point, which is sampled over a couple of years. And then common global database, as pointed out by some of the keynote speakers, I mean, even groundwater data, there's only eight countries which put their data into this big database. So can something be done to facilitate this? And this is not just NGA, but USGS, USDA, NASA, NOAA. Is there some way to get this awareness out there to say, hey, look, we want to understand global water resources, especially groundwater, so what do we do? And then access to high performance computing, even though some of them have become less of an issue today, but people were very insistent that having access to high performance computers could help it, especially in the era of trying to get high resolution, special resolution. And obviously international collaborations are important. And we want to figure out how to leverage it so that we can get more and more access and places where we can try out our scientific hypothesis and solve scientific and societal problems. So I'll end with that, but I'd really want to hear from the people in the room to see if I missed something or to add something or comments, concerns. Come on, don't be shy. Yes, Kamini? In here. So I think the only thing that struck me that might be missing, sorry, I'm Kamini, saying I school of minds, that we talked a lot about yesterday and little bits and pieces was the human piece. Yes. And so that might be the only thing that wasn't captured there in my mind was that idea that it's just hard to get data on what people do with any systems and models. But that's the only thing I see that we talked a bit about yesterday that didn't show up here. Sankar. Other piece I just would like to highlight is one is looking at the budget issues, the other aspect is looking at the extremes. And because during extremes, it's only the surplus or deficit of one of the variables. So it's important to have some attention to that and it's relevant to even perhaps remote sensing, which may be very helpful there too. So a quantification of hydrological extremes. But with respect to groundwater, good point.