National Liaison Committee Meeting for the NWQP — Part 3

Video Transcript
Download Video
Right-click and save to download

Detailed Description

In these videos, speakers discuss upcoming changes to the USGS National Water Quality Program (NAWQA Project) and three new priority areas for the USGS Water Mission Area. Gary Rowe discusses plans for transitioning from current NAWQA Project activities to the new priority areas. Chad Wagner discusses plans for the Next Generation Water Observing System (NGWOS), which will provide high-fidelity, real-time data on water quantity and quality. Katie Skalak presents information on the Water Prediction Work Program, or 2WP, an ambitious federal partnership for developing a national, interagency new capacity for water prediction. And Mindi Dalton shares plans for the Integrated Water Availability Assessments (IWAAs), which will predict water availability for human and ecological uses at regional and national scales.


Date Taken:

Length: 00:27:20

Location Taken: Washington, DC, US


- [Man] Katie is actually a research hydrologist, with our former national research program, a geomorphologist by training. But now she's the chief science advisor for the Water Prediction Work Program.

- Okay, thank you. So here is my contact information for myself and the new program manager, David Lesmes. The big picture for Water Prediction Work Program, or 2WP, ultimately it's sort of within under the wedge on a much broader capacity for integrated hydro-terrestrial modeling. This'll be a national asset that supports the earth and biological system prediction capability. So, Gary mentioned our strategic science plan. That's organized our science, or broke it into component parts, thinking about our observations, our process understanding, models, and data delivery. What we're going to do is essentially integrate all of that to drive a predictive capacity. We're focused in the near term on the national water model, in a partnership with NOAA, but this integrated Hydro-terrestrial Modeling Portfolio will be much broader, and will include leading edge developments of modeling across multiple fronts. We're ultimately driving the prediction of stream temperature, surficial processes, and here we're really thinking about hillslope processes or entrainment of sediment and particles on the landscape, and then in-stream transport of those particles, sediment, constituents in these waterways. This is sort of a stream-biased perspective. We're also gonna be thinking about reservoirs, lakes, other inputs of water. Ground water, it's not just streams. We're thinking about this in sort of a program space on multiple tracks and boundary conditions. We'll have three program timescales operating simultaneously. In the very short-term, we're focused on the National Water Model, in a collaboration with the National Weather Service. In the mid-term that is driving our efforts towards what a next major capability upgrade will be to hydrologic prediction, and that may be the National Water Model. It could be broader. And then in the long-term, what we're really looking for is this cutting-edge... ecosystem of modular models that can couple together at a user-specified need. They're all gonna be wrapped and... integrable at a user-defined space. The explicit goals of the program. First, we're going to improve and augment our predictive water modeling capabilities over a range of spatial scales. We're thinking from the watershed all the way out to Kona's, and then a broad range of temporal scales. So this is short-term forecasting all the way out to management timescales, including a capacity for improved, hydrologic processes, water temperature, surficial processes as I mentioned, and then in-channel transport of materials. But our second goal is to cultivate a collaborative community around high-level research questions and modeling efforts. This includes establishment of a shared computing environment. To get these models to couple and work together, we all have to define the space as a community. So the goal here, this secondary goal, is to forge this shared community space. Gary mentioned we're thinking about our research now in an integrated way, and the first thing I went and did was chop everything into silos. It's just a simple way to start thinking about the problem. The science teams that I devised are focused around these areas. Instead of thinking about entrainment and transport. As a geomorphologist, I certainly think about the life cycle of the sediment particle, and so I thought about it from its origin all the way out to the outlet of the basin. I just think about as sediment, and so that led to okay well let's think about constituents in the same way, and then hydrology is an obvious team need. Similarly with temperature. And then we have what we're calling Topo/Bathymetry, and that's sort of the geospatial fabric pulling in the remote sensing data. Sort of the shape of the container that everything's flowing through. Moving out, we have an Observation Network Team. This is really gonna be ultimately not superseded by NGWOS, but it's a direct line into NGWOS. And then our Interoperability Core Team. We're thinking about the data interoperability, model interoperability. Finally, the Ecology Core Team, which will involve a lot of partnerships, both internally at USGS and beyond. So this leads into some of the partnerships. So where we're at right now, we're sort of in this space of 2WP is right at the crux of the Water Missionary within USGS and this collaboration with NOAA. And as we move out, our priorities and our directions will be determined by our partnerships. So what this looks like will depend on how we prioritize our partnerships. At the same time, USGS is a bureau. It's thinking about integrative predictive capacity, so this would involve pulling in ecosystems, core system science, all of our other mission areas around this issue of integrated prediction. The water missionary is leading the edge of this for the bureau effort, but we're also at a seat at the table for this Community Integrated Hydro-terrestrial Modeling Portfolio. The sub-committee on water availability and quality that Gary referenced, partnership with other agencies. We've seen this quite a bit. I'll just use it to point out the Water Prediction Work Program sits kind of in the middle of this pipeline and flow. So I've tended to think about it in that way. It's not just coupling models. We need data delivery. We need this Next Generation Water Observing System, and then we need decision support tools to drive the questions that we're going to answer. We're a conduit along this flow. And the information goes both ways, and I think that's the way we're all thinking about it. So the bureau is thinking about prediction in this space of fundamental data... and this is sort of just a level one product. This is at the bare minimum if we're going to move out on prediction, we need this to be integrated and shared. At this next level, it was posed in terms of fundamental models, and what that really means is a singular model that captures a single process. The National Water Model is an example of, it captures more than one process, but it's really just focused on flow. That's a fundamental model. But if you couple the National Water Model with the National Groundwater Model, now you have a higher-level model. I tend to think about this space as fundamental research. And then here we're getting into model integration. We have a suite of spatiotemporal data services. Some of this is included in NWIS modernization, but it's much broader than that as well. And then finally, we have this decision support assessments, IWAS, as an example. I sort of use this to organize my thinking about 2WP and how we will move forward. At the first level, we have to think about data integration. I wanna quickly point out... In the previous slide, we have this lateral component of integration but then also vertical. We need a full integration from data all the way out to decision support. There's a lot of work to do in this lateral space of how we couple data that's going to feed our models, and what are the research questions we're going to answer to drive model integration. The way that I've been thinking about data integration on a national scale... Here's some USGS data from NWIS. How would we integrate that data from the NSF sites? Is it even possible to do that? There's a huge opportunity there to pull in other networks. Will it work? Can we do it? Down here is an example. The neon ecological data, if we move out on ecological forecasting, and similarly data collected at EPA Ecoregion levels. We're going to need to get that information, to speak to each another and be useful to one another in order get to a predictive framework. We've heard a little bit about these National Academy of Science recommendations for the future water priorities for the USGS. The high level questions that they posed for the USGS to consider, that are extremely societally important for us to move forward on. This is something that requires integration, so what is the quality and quantity of atmospheric, surface, and subsurface water, and how do these vary spatially and temporally? How do human activities affect water quantity and water quality? These are very high level questions that no single discipline can move out on to address. It requires integration and teams to solve these problems. We've tended to think about some of those high level questions at a watershed scale. This is conceptually very simplistic. We've thought about land use. We've thought about how it impacts sediment transport, surface runoffs, suite of hydrologic processes. But we tend to do it in a discipline-specific way, and there's more that can be done on integration. NAWQA has done a great job at integrating at the watershed and national scale, but there are still important questions that remain. To sort of get to the question about our data and our models, if our models are doing really well, can we save money on our monitoring? I would say probably not, because we really need to capture nonstationarity. That builds the framework for how we'll be able to predict unsteady conditions and situations like changing climate or forest fires. So I would say we really need a tight coupling there. We're used to thinking about it in this space. We can do better at integration, but what happens when we move beyond that space? What about scaling watershed to region or national? I would argue we're at the infancy there in many ways. A lot needs to be done. So what we're going to do is basically build on the NGWOS approach, and we're all gonna be going into the same basins to sort of focus our efforts. Chad mentioned the approach that they're gonna be using, with these 22 water resource areas. I hope I'm not... Chad's not here to correct me, so I'm just gonna go out on it.

- [Man] That's what happens when you step out.

- Where I think we're gonna go is establish frameworks where these 10 watersheds. It would be pretty clear how they could be extensible to other regions through our conceptual frameworks and our modeling. So, it's not just these 10 watersheds. We're gonna be thinking about pulling in data from other watersheds. We're gonna be thinking about how to extend that, and how it makes sense to do it. I'm arguing for developing frameworks with national data. One example would be physiographic provinces. Those water resource areas are another level to think about. EPA EcoRegions. What is the ecology doing to interface with the hydrology to control the surface and subsurface. So, the way that this simple-minded approach that I'm sort of arguing as a first cut, thinking about dominant processes across the landscape will be very useful. Our model complexity can be reduced by thinking about the number of dominant processes. Here is a output from a hydrologic modeling framework that suggests an important process to consider over most of the country is evapotranspiration. That simplifies things pretty quickly. If we wanna do well here, we're gonna have to think about snowmelt. Similarly, if we thought about the inferior processes, we can use that information as well. We can do that for hydrology. We can also do it for geomorphic processes, so these maps show hillslope, colluvial, fluvial, waterfall processes mapped out on a landscape. You can take that spatial information, and use it to simplify your modeling frameworks. We have to do the same thing for all these anthropogenic processes. If we move out and select these other basins, I would argue we have to think about the natural processes, but then how do anthropogenic processes map onto that. For coupled, human, and environmental. Map urbanization, deforestation, dams, all the way down the line. That's sort of the spatial component. If we can get our models to resolve from watershed to national scale, and have a framework that's consistent, that will be a major upgrade and a huge step forward. But we also have to think about timescales. How do we resolve very short and very long timescales? This is an example from the diagram that I showed with the teams. The Constituent Team generated this, as they've thought about how to model constituents. First, they've identified their priority constituents for forecasting. So this is just at the team level. This isn't something that I'm saying, "Okay this is what USGS is going to model for constituents." This is what the team has decided is important, and this list could be flexible and adjustable. Going through this process of data synthesis that I've described, we lead to short-term forecasting and long-term forecasting. Here are a suite of stakeholder uses from 2WP outcomes that would range from hourly to month, and then seasonal to decades. Similarly, this is going to be, I think, big scientific advancements forward. To think about resolving process timescales with modeling timescales. As a geomorphologist, I think about things in spatial and temporal cascades. We have the spatial scales ranging from greens to the watershed, and here we have timescales ranging from less than a year all the way out to tens of thousands of years. Me and my colleagues mapped on what we thought important sediment processes for Chesapeake Bay streams are. You can see ranging from the glacial to periglacial, at the watershed and tens to hundreds of thousands of years, all the way down to rainsplash on the hillslope that entrains small particles and knocks them in the river. The question is, when we're trying to predict sediment transport, do we need to know about all of those things? I would argue we don't necessarily know the answer to that, and so that drives a lot of other research moving forward. We decided to pull out storage, because models at this point don't even think about storage. That leads to breakdowns in our predictions. Sediment just doesn't go into the water, just go downstream. It gets dropped into portions of the landscape, and that limits our ability to predict how we can manage the landscape. That's an important set of processes that models don't even consider yet. The modeling tools that we typically use for sediment have the similar range in scales. There's a really big gap here on human management timescales at the watershed, and we're trying to plug those in with some of the things that we've suggested, but these are really data intensive, and there are few sites where we can do this. This is just gonna be a quick and dirty fix. There's a lot of work to be done. The process timescales cover this whole space. The available prediction tools cover some of the space, so we have a lot to do to think about resolving spatial and temporal scales. Moving up a level to level three: Integrating Models. The way that this has been posed that I think is the most articulate, my colleague, Joe Hughes said, "the reason you would put two models together is because the stakeholder question would drive the need for that." So, he and Fred Ogden in a collaboration at the National Water Center last summer, put together the National Groundwater Model with the National Water Model cutout to basically forecast a set of base flow of conditions at the Northern High Plains. This is an area where they know the model doesn't do well, and the reason it doesn't do well is because it doesn't capture base flow accurately. They put those models together and they've come up with a set of recommendations. They have a set of key findings. They need an integrated approach to identify losing stream reaches. The model basically only allows for one-way flow of water, and doesn't allow for the water to flow back in, and that is a significant challenge when trying to accurately predict flow. I won't read the rest of those. And then finally getting to the highest level, level four. Thinking about decision support or assessment tools. We're going to work very closely with the Integrated Water Availability Assessments to drive the questions that we are going to answer. We have these high level research questions to organize all of our thinking around it, but we really need to be answerable to stakeholders at the end of our process. I think those are not incompatible, and we can answer subsets of those questions by being directly relevant to the stakeholders. In summary, 2WP is a cross-cutting effort that integrates our observations, understandings, predictions, and decision support. This is a really big effort. It's going to revolutionize the way the water mission area is doing in science. It's collaborative and inclusive. It is going to be and is right now a heavy lift and requires wholesale participation from all of our partners and internally. It's also evolutionary. So we're early in, and this is a long-term effort. We continue to take feedback and information as we put it out. At times that can be frustrating, because you'll see this talk, and then if I ever get a chance to talk to you again, what I say will sound maybe completely different. But that's just the nature of how this is going. I think that's my last slide. If there's time for questions, I'll take them.

- [Man] Yeah, we definitely have time for questions.

- [Woman] My question is what's the annual budget for this component--

- This year, it's zero. Notice I didn't have any slides about staging work, and everything that I have presented is very nebulous and sort of high level, and I think conceptual. Yes, very much. I think that's a result of, we don't have deliverables or a set level of funding driving our program right now, and that's okay. I think it's coming.

- [Man] But I would just add that there are activities that are currently funded that are sort of the foundation of a lot of this. So, as part of the planning activities... the direction that those kind of foundational efforts take, fully driven by what results in all these planning activities.

- [Woman] At this point, you mentioned several stakeholders and partners. At this point, conceptual planning has been primarily USGS expertise or--

- Yeah, so I would say it's USGS and our academic partners, but our Water Science Centers have a keen knowledge of what stakeholders need because of the science that they've been doing. So they're represented on all of the science teams, and so I think we're getting a good sense. It's not fully representative, and I think we have work to do on stakeholder engagement, but I think the challenge that we found with 2WP, if you offer something that's very conceptual and broad, everything looks like it fits in. And that's good in some ways, but it can be bad and lead to a sense of over promising in other ways. We really wanted to hone in on where do our intellectual research needs match our capacity? Where do all of those things match the tools that we have available, and where does the data fit for that? I think we have some work to do before we suggest, "hey we're gonna be able to tell you water quality and quantity at all scales and all timescales." I think we don't wanna over promise. The stakeholder engagement has really been driven by NGWOS, and 2WP has been kind of following along.

- [Man] And I think that Mindy's point, there is a lot of foundational work, particular on the modeling side that the various water programs like NAWQA and like the Groundwater Program and et cetera, a lot of the modeling platforms exist, even our agency and other agencies. Different timescales. The challenge is what questions are we gonna try and answer going forward, and how do we fit the right models together to get those questions.

- [Woman] So, I am not sure if I misheard, but in the science teams, is there any role for social science in there?

- Yeah, most definitely. I think as these teams have been built, this is sort of a snapshot, and I think those silos of teams represents how we would get our minds wrapped around what can USGS bring to the table? What are the questions that still need to be answered through our modeling and research efforts? That sort of scoped it how it appears now, but as the process has been going, we've quickly understood the need for a communications team, for a stakeholder team or a social science team, water use and sort of anthropogenic aspects. I think moving forward, especially when we go out into basin-specific activities, the teams will look very different. What we'd probably have are integrated science teams for all the basins, pulling in facets that I haven't even included here. But that will require partnerships, because we don't have a huge stable of social scientists at USGS.

- [Woman] You have some small network of--

- Yes, yes, and they're being tapped quite a bit because their skills are in high demand.

- [Woman] Like land use planners and economists could use tools like this in amazingly impactful ways, and so to have them at the ground level to talk about these--

- [Man] In some ways, it's deja vu all over again, because when the NAWQA first cycle started out, we had for each for each of those 51 watersheds. We had local study that weighs in on committee teams, as well as this national committee, and those were really the folks that really drove a lot of design decisions and stakeholder needs and how we could address those. It helped us with a lot, so do something similar--

- Yeah, yeah. I think I should credit all of the things that we're getting from the NAWQA program. When I first was tapped to start thinking about this, the first person I reached out to was Gary and Laurie Sprig and people from NAWQA. How did you build a regionally relevant effort that's nationally consistent? I hope we're gonna take the best of what NAWQA has to offer.