Hydrography Webinar Series - Session 3

Video Transcript
Download Video
Right-click and save to download

Detailed Description

 

USGS Hydrography Webinar series that shares success stories from users, provide information on NHD and other related products, and provide a forum for users.

Details

Date Taken:

Length: 00:53:10

Location Taken: US

Transcript

0:06
Al Rea: Well, thanks to everyone for joining us today.
1:00
This is the second session of the USGS Hydrography Webinar Series.
1:06
I'm Al Rea and I work with Jeff Simley.
1:09
Together, we're CoManagers of the NHD.
1:15
I'd also like to acknowledge Steve Aichele who'll be helping with the questions and answers
1:19
later on, and Alison Jason who's helped us with the WebEx training center technology
1:26
and a lot of other things.
1:28
There are also many other people who've been helping make this webinar series possible,
1:34
but there are too many to name here, so we'll move on.
1:41
Looks like we've got a little bit over a hundred people online now.
1:50
I'll go over here, this is the agenda for today's seminar.
1:54
I'll have some brief introductory comments about the Hydrography Webinar Series and then
2:00
I'll give a short overview of how the flows are computed using NHDPlus.
2:05
Then Ed Clark will give us the feature presentation for today, about the National Flood Interoperability
2:12
Experiment.
2:13
We'll have some time for questions and discussion after Ed's presentation.
2:33
I
2:47
want to give a little bit of introduction to the webinar series.
2:50
This is our second in the series, and we have three basic purposes for the webinar series
2:57
that we're showing here.
3:00
First we'd like to share success stories from users who solved realworld problems using
3:06
hydrography data.
3:08
We're also going to provide some information on the NHD, the WBD, and other related products.
3:15
Finally, we want to provide a forum to our user community, similar to a conference setting.
3:22
In these days of shrinking budgets and travel restrictions, we felt there's a place for
3:27
a virtual conference of this sort.
3:35
[silence] Al: These are some of the topics we hope to
3:40
cover during the seminar series.
3:42
The list isn't intended to be exclusive though, so please let us know if there are other topics
3:48
that you'd like to see covered.
3:55
We plan the webinars to be a mix of presentation types.
4:02
We'll have users describing how they've used hydrography data to solve a particular problem.
4:07
We'll have presentations on the underlying technology and the data.
4:13
We also plan to have some "Rapid Fire" sessions in which several presenters will give very
4:19
short, fiveminute presentations, just to give us a quick overview of what they're doing.
4:25
We'll also try to collect your feedback on what you want to see in future seminars.
4:36
These are some of the ways which you can get more information about the webinar series.
4:46
The main resource is the web page shown on the top, here.
4:50
Note that the URL is casesensitive.
4:54
We also have a newsletter that Jeff Simley sends out each month.
4:58
If you'd like to be on the mailing list for the newsletter, send us an email.
5:05
We will send out announcements for future webinars through the AWRA and some of the
5:10
other organizations.
5:13
Also, when you sign up for one of the webinars, we'll put you on a list for future webinar
5:19
announcements.
5:21
We promise not to spam you with other stuff from this list.
5:27
We plan to have webinars every six to eight weeks.
5:32
We've placed all the phones except for the presenters' on mute to minimize the background
5:36
noise during the webinar.
5:38
If you have a question, please find the Q&A tab, which has a questionmark icon in the
5:44
upper right of the WebEx window.
5:48
You can type in your questions there.
5:51
If you've made your screen fullscreen, the presentation fullscreen, then the Q&A
5:59
options are on a pulldown menu that appears when you put your cursor near the top of the
6:06
screen.
6:07
We'll monitor the Q&A list for questions from participants and we'll ask the presenter
6:11
to answer some of them during the discussion session.
6:14
We'll post written answers to all the questions later, on the seminar series' website.
6:22
We'd appreciate it also if you'd take some time to answer a few very short questions
6:28
at the end of the seminar to help us improve future webinars.
6:39
Now I'll start with a very short presentation on the NHDPlus, because last month's presentations
6:46
by Bill Samuels on RiverSpill and ICWater and today's presentation on flooding by Ed
6:52
Clark, both those use the NHDPlus for flow computations.
6:59
I want to give you some background on how the flows can be computed using NHDPlus, and
7:05
specifically how the mean annual and the mean monthly flows that are included in NHDPlus
7:11
were computed.
7:13
I'm going to cover this pretty quickly and at a very high level.
7:16
There's a lot more information about this in the NHDPlus User Guide, if you'd like to
7:21
know more information.
7:27
I'd like to take a moment here to acknowledge the really very small team that's been working
7:57
on the NHDPlus for the last decade or so.
8:01
The team you see here led by Tommy Dewald at the EPA conceived and developed the NHDPlus.
8:08
An interesting thing about this team is that they've only met altogether in person just
8:13
a few times over all these years, but they've spent probably close to a thousand hours on
8:19
the phone discussing every aspect of NHDPlus.
8:23
I'm not kidding, a thousand hours is actually a pretty good estimate of the time.
8:28
Version 2 of NHDPlus was actually produced entirely virtually.
8:34
The team didn't meet at all in person during the production of NHDPlusV2.
8:38
I think it's testament to how much a small team can accomplish given the right technology.
8:46
Personally, I'm very proud to have been a part of all of this.
8:54
Here's an overview of what I'll cover in this presentation.
8:57
First I'll talk about the major components of the NHDPlus for use in flow computations.
9:02
These include the surfacewater network, the NHDPlusnetwork attributes, catchments, catchment
9:10
and watershed attributes, and points of interest that are linked to the network such as gages,
9:15
modelforecast points and dams.
9:18
Then I'll talk about the specific example of flow analysis that's included in the NHDPlus
9:26
which is called the EROM, or the Enhanced Runoff Method, for computing flows and velocities.
9:34
These are the major components of the NHDPlus, with the vector data on the left, and the
9:46
raster data on the right.
9:48
NHDPlus includes a snapshot version of each of the three major datasets that were integrated
9:54
to create NHDPlus, and this includes the complete 1100,000scale NHD, the watershedboundaries
10:02
dataset, and the national elevation dataset on the raster side.
10:08
NHDPlus also includes the actual features that were used to do hydroenforcement of elevation
10:14
data.
10:15
The catchment polygons are one of the major products.
10:19
We also include the hydroenforced DEM, the flowdirection and flowaccumulation grids,
10:25
a raster version of the catchments, and lots and lots of attributes.
10:38
The NHDPlus has a set of attributes that are known as the Value Added Attributes or VAAs.
10:45
The VAAs encode lots of information about the stream network, and they enable many different
10:51
kinds of analyses.
10:54
Shown here are a few of the attributes that are used for analysis in navigation.
10:59
There are several other attributes as well.
11:02
Today, we're going to focus on the second and third navigation attributes, the from
11:08
and tonodes, and the hydrologic sequence number.
11:13
The NHDPlus has a nationally unique set of node IDs.
11:23
These can be used for nodetonode networkwalking models.
11:29
Lots of the traditional flow models fit into this category.
11:34
An interesting thing about these node numbers is that they're entirely conceptual.
11:39
They exist in a flow table that identifies the streamflow lines that flow to each node,
11:45
but there is no actual feature class of nodes in the dataset.
11:56
Probably the most important of the Value Added Attributes is the hydrologic sequence number.
12:02
The hydro sequence number is designed so that from any flow line in the network, all the
12:07
flow lines that are upstream have a higher number, and all the flow lines that are downstream
12:12
have a lower number.
12:14
This is pretty amazing, since the national stream network is a whole lot more complex
12:19
than the one that's shown here on this slide.
12:21
It includes many thousands of divergences.
12:24
In other words, it's not a simple dendritic network.
12:30
The one really useful aspect of hydro sequence numbers is that if you sort them in descending
12:35
order, you can walk through the entire network and do some process, like computing drainage
12:41
area.
12:42
Wherever you are in the network, you will already have processed everything upstream
12:47
of you.
12:48
All you have to do is add up the upstream areas.
12:51
This means you can do processing like this in a single pass through the entire network,
12:56
and that makes the processing very efficient.
13:00
[silence] Al: So we can use attributes like hydro sequence
13:09
number, along with several others, to do network navigation using database queries.
13:16
If the network was simple and dendritic it would be pretty easy to explain what the queries
13:20
are, but because we have divergences in the network the queries are a bit more complex.
13:26
They're still queries though, and databases are really good at doing queries very fast.
13:31
The NHDPlus has a tool that can do network navigation using queries.
13:37
It can do these four types of navigation upstream main stem, upstream with tributaries, downstream
13:46
main stem, and downstream with divergences.
13:50
The tools supports a variety of stop conditions, too, like stopping the trace after going 50
13:56
miles upstream, for example.
13:58
Finally, because these are all database queries, they're really fast and they can work efficiently
14:05
on really large navigations.
14:08
[silence] Al: The VAA navigation tool is available as
14:16
a tool bar in our map and it's also available as a .dll that you can call from your own
14:21
custom program like a Python script.
14:24
With this tool it would be pretty easy to make a script that can do some very sophisticated
14:29
networkbased analyses.
14:33
Here's a simple example using the tools showing an upstreammainstem and an upstreamwithtributaries
14:45
navigation for the Susquehanna River.
14:48
Both these navigations happen very fast.
14:52
[silence ] Al: Aside from the streamnetwork VAAs, probably
15:01
the next most important feature in the NHDPlus is the catchments.
15:06
Catchments are the immediate areas that drain to each segment of the stream network.
15:11
I'll show a picture of some catchments in a moment.
15:14
On this slide I have some details about how the catchments are computed, but the most
15:19
important thing that I want you to know about catchments is the last point.
15:22
They really are the key to understanding and connecting the landscape with the stream network.
15:30
That's important because overland flow is quite different than flow in channels.
15:35
In this process, a lot of things can happen such as pesticides being washed off of fields
15:40
and into a nearby stream and so on.
15:45
Catchments really are the key to understanding how landscape attributes like land cover,
15:51
or geology, or soils, affect stream flow.
16:02
The way that we compute catchments is by combining the NHD streams and water bodies, the watershed
16:10
boundaries, and elevation data.
16:12
Exactly how we do that is a topic for another day.
16:17
For now, just understand that this process is the key to getting catchments in NHDPlus.
16:24
[silence] Al: Here's a picture of some catchments.
16:31
You can see there's a catchment for each line segment in the NHD network.
16:35
There are catchments also for each coastal line segment.
16:41
The catchments in the NHDPlus cover the entire continental US.
16:45
If you drop a hypothetical drop of water anywhere in the 48 states, you can know instantly where
16:52
in the stream network that drop of water would go, simply by intersecting the point where
16:57
it falls with the catchmentdata set.
17:01
[silence] Al: The NHDPlusV2 includes the catchment attributes
17:10
that you see here.
17:12
There are mean annual and mean monthly precip and temperature, as well as runoff from a
17:18
simple waterbalance model that was developed by Dave Wolock and Greg McCabe.
17:25
There are also all the categories of the "2011 National Land Cover Data" set.
17:31
For all of these we have both the attributes for each individual catchment, but also the
17:37
accumulated attribute for each catchment, and that includes all the catchments upstream
17:44
of it.
17:45
There is also a mean latitude for each catchment, but this one isn't accumulated for upstream
17:51
catchments.
17:54
All these attributes except for the NLCD were used to compute the flow estimates that are
18:00
included in the NHDPlus.
18:02
[silence] Al: Now that we have a stream network in catchments,
18:12
we need to tie some observations to the structure.
18:15
Here's some examples of the kinds of observations in points and lines that we can link to the
18:20
network.
18:21
There's really no limit to the kinds of things that we could link to the networks.
18:25
We do this in GIS using something called "linear referencing," which is similar to mile markers
18:31
along the highway.
18:32
We don't have time to get into those details right now, though.
18:35
For the flow analysis we're going to be looking at the stream gages.
18:39
They are the things that we're most interested in.
18:43
[silence] Al: Now I'm going to give you a really brief
18:49
overview of a specific example of streamflow modeling and that's the one that's used in
18:54
the NHDPlus itself to compute the mean annual and the mean monthly flow.
18:59
It's called the "Enhanced Runoff Method" or EROM, for short.
19:02
EROM goes through these steps to compute the flows.
19:08
The flows are computed for each of the first five steps here in NHDPlus and then each step
19:17
gives us an improvement of the estimate.
19:20
The results of the fifth step are considered to be the best estimate of the flows.
19:27
I'll walk through each of these steps very, very quickly, but I just want you to get a
19:32
feel for what the process is like.
19:36
Al: This is the first step, which is a waterbalancerunoff model by Dave Wolock and Greg McCabe.
19:45
It's developed from the PRISM precip and temperature data for the US and from similar grids that
19:51
were developed by the Canadian Forest Service for Canada and for Mexico.
19:58
Al: This graph shows the results of the waterbalance model.
20:03
The points are gages.
20:04
If the estimates were perfect they would all fall right along that red line.
20:12
Al: The second step is an adjustment for what's called "excess evapotranspiration" from the
20:23
streams and from water bodies.
20:26
This graph shows how that adjustment worked for the Colorado River basin and you can see
20:31
it improves the estimate a lot in this region.
20:35
It has less effect in more humid regions.
20:39
Al: This graph shows how the steptwo adjustment moved the points closer to the red line.
20:51
This particular graph is for the lower part of Region 10, which is the Missouri River
20:57
basin.
20:58
Al: Step three was a regression of basin characteristics against flows for the gages.
21:09
The regression improves the fit of the data quite a bit.
21:16
Al: Step four lets us account for additions and removals from the flow network.
21:24
There's a table called Plus Flow AR that contains records that tell us where flows are added
21:30
or removed from the network.
21:31
Currently, this table only includes a few of the largest water transfers.
21:35
But the structures are there, so that we could represent lots more detail if we had the data
21:40
to support it.
21:44
Al: Here are the results of additions or removals for the New England region.
21:52
You can see that for most of the gages it had no effect at all, but it improved the
21:56
estimate considerably for that one outlier.
21:58
Al: Step five is an adjustment using the flows from the gages.
22:07
Basically, for every flow line that has a gauge we adjust the flow and match it to what
22:13
the gage tells us it should be.
22:19
Then we adjust the flows on upstreamflow lines, and then we reaccumulate those flows downstream.
22:30
Al: Then, finally, there's a process that's done that randomly renews 20 percent of the
22:40
gages just to see how that affects the results.
22:42
That gives us an idea of how good the results are in step five.
22:47
This is a picture of the final results.
22:54
There's a flow estimate for each of the 2.7 million flow lines in the NHDPlus, and here
22:59
you can see we've set the line width to...represent the flows, there.
23:08
I just want to briefly mention that we're developing NHDPlus now for the high resolution
23:13
NHD using 10meterelevation data.
23:17
But we'll have to cover that another time because we need to move on to our main presentation
23:24
today, which is going to be Ed Clark's presentation on the National Flood Interoperability Experiment.
23:32
Alison, if you could switch over to Ed, I'll do a quick introduction.
23:38
Ed is the National Flash Flood Service Leader in the National Weather Service Headquarters,
23:46
Forecast Service Division.
23:48
Prior to joining headquarters he was a Senior Hydrologist at the Colorado River Forecast
23:53
Center in Salt Lake City.
23:55
Since joining the hydrology team at the National Weather Service Headquarters, he has led efforts
24:01
on behalf of the Office of Hydrologic Development to scope and design the innovative waterresourcesscience
24:08
and services interoperability and datasynchronization capabilities, as well as planning for the
24:14
NOAA National Water Center.
24:16
Ed's also a CoChair, along with me, of the Subcommittee on Spatial Water Data.
24:21
Ed, go ahead and take it away.
24:23
Ed Clark: OK, thanks Al.
24:25
Can you see my screen?
24:27
Al: Yes.
24:28
Ed: All right.
24:29
Well, thanks Al.
24:30
This is a really exciting opportunity to share how we within the National Weather Service
24:35
and actually across the broad consortium of university and academic partners are using
24:40
the National Hydrography data set to really explore and push the future of flood forecasting.
24:46
The National Flood Interoperability Experiment is a little bit of a spontaneous and somewhat
24:54
serendipitous effort.
24:55
This is really the brainchild of Dr. David Maidment at the University of Texas, Austin.
24:59
It's been championed by him among other members of the CUAHSI community.
25:03
The goal is to explore the future of flood forecasting done at very, very high spatial
25:09
resolutions.
25:10
The 2.67 million catchments and streamflow lines are the NHDPlus.
25:15
Explore realtime floodinformation services, the weather naval services that will share
25:20
information seamlessly, transparently across the operational forecasting and development
25:27
and then decision management processes.
25:29
Most excitingly, this is an opportunity for us to engage the academic community and explore
25:35
emerging capabilities through the National Water Center.
25:38
For those of you who don't know NOAA, through a directed appropriation I was fortunate enough
25:43
to build a National Water Center on the campus of the University of Alabama in Tuscaloosa.
25:47
It was completed in 2014, and we're actually cutting the ribbon on that building next week
25:53
in Alabama.
25:56
We really hope this is only the catalyst for emerging waterresources capabilities within
25:59
the National Weather Service and NOAA, but it serves as a center point for engagement
26:04
with other federal agencies.
26:05
Case in point, the USGS is a close colleague with us, not only in Subcommittee on Spatial
26:11
Water Data, but it's been a partner throughout the planning and preparations for the National
26:18
Water Center, and will continue to do so.
26:21
Why flooding, why national experiment dedicated to exploring alternatives and new ways for
26:27
doing flood forecasting?
26:28
I don't think I need to tell this audience that flooding is our most severe form of weathernatural
26:33
disasters.
26:34
On average we lose more fatalities to flooding than any other severe weather disaster.
26:41
The exception, the outlier here is heat, but we don't classify that as severe.
26:47
In fact, if we look at the last hundred or so years of floodloss data and extrapolate
26:52
that forward, we can anticipate that flooding will cost the nation.
26:57
Direct impacts of losses due to flooding will cost the nation a delta of $300 billion and
27:02
we run the risk of losing an additional 2,500 lives.
27:07
To put that in perspective, $300 billion is greater than the cost of the National Highway
27:13
System throughout its 60year development.
27:16
2,500 fatalities are more than the fatalities of service men and women since the onset of
27:23
9/11.
27:25
Why is the National Flood Interoperability a use case for the hydrographydata set?
27:33
The hydrographydata set is the language that links together flood forecasting, flood predictions
27:39
in time and space with the expected impacts.
27:42
It is this framework that allows us to seamlessly communicate between where a point is modeled
27:50
and forecasted, and where those impacts may be conflated.
27:54
To facilitate this nextgenerationhydrologic modeling I'll describe little bit about the
27:58
basis for its use within the modeling components of the National Flood Interoperability Experiment.
28:02
As I said, it is the framework for data conflation, and not just between Weather Service data
28:08
sets, but with other federal data sets such as the National Flood Hazard Layer from the
28:13
Federal Emergency Management Agency.
28:16
It allows us to intersect projections and forecasts with impacts or data sets from the
28:24
Environmental Prediction Agency.
28:25
The list is really limitless, and I think Al alluded to that in one of his earlier slides.
28:31
Any type of data can be mapped to the NHD and in doing so, it really becomes a very
28:37
powerful tool for describing not only the impacts of flooding but other waterresourcerelated
28:41
variables.
28:43
Finally the hydrography, using a common hydrographydata set or a common language, a common mapping
28:48
convention projects a framework for the research in academic communities to better tie with
28:53
federal operational forecasting institutes.
29:00
NFIE is not just a oneyear experiment.
29:02
In fact, as the National Water Center is developing these capabilities, many of the types of systems
29:08
that will be demonstrated in glimpse during the NFIE will have some longevity for operationalization
29:13
at the National Water Center.
29:14
This allows us to go from our current forecasting system of approximately 3,600 locations for
29:21
Advanced Hydrologic Prediction Service AHPS, water.weather.gov, to over 2.6 million locations.
29:28
That means that, when we get to the instantiation of a WRFHydrodriven forecasting system on
29:34
the NHDPlus, the [inaudible] will be on average no more than a mile away from a [inaudible]
29:41
forecasting location.
29:42
Most importantly, this new data set will inform emergency managers on how they can do a better...how
29:48
they can improve their decision about...they can improve their processes for the protection
29:52
of lives and property.
29:54
That's NOAA's and the National Weather Service's main organic mission.
30:01
We asked three specific questions within the National Flood Interoperability Experiment.
30:05
How can your realtime hydrologic simulations, at very high spatial resolutions covering
30:09
the Nation, be carried out using the NHDPlus?
30:12
How can this lead to a more improved emergency response and community resiliency?
30:16
In order to test this, we are partnering with members of the emergency management community
30:23
as well as social scientists.
30:25
Finally, how can an improved interoperability framework as I said earlier, web services
30:30
support these first two goals that lead to success and sustained innovation in the research
30:35
to operations process, and more importantly, the operations to research process?
30:39
There are five major components of the National Flood Interoperability Experiment.
30:45
There's NFIEGeo, or the National Geospatial Framework for Hydrology.
30:50
There's the NFIEHydro, this is a coupling of highresolution hydrologic forecasting,
30:54
numerical weather prediction models, and downstream routing models.
30:59
There's a research track within the Flood Interoperability Experiment to explore how
31:03
riverchannel information and dynamically generated flood invasion mapping could be carried out
31:07
or coupled to a national hydrologic forecasting system.
31:12
As I just said, there's a component of this that will explore how new data sets, emerging
31:18
data sets, enabled through web services, can be more probably made available to the emergency
31:23
response community.
31:25
I will pause here and say that all this is done within the confines of a true experiment.
31:29
None of the information, none of the services from the NFIE will be misconstrued or translated
31:35
as operational forecasting.
31:38
We do this because we can't use the public as a guinea pig.
31:40
More importantly, we need to characterize the magnitude, the skill and the uncertainty
31:46
in these types of emerging technologies.
31:48
Nevertheless, it's a very exciting time.
31:50
So, what is NFIEGeo?
31:51
At its heart it's built on nine feature classes from the NHDPlusV2.
31:57
These are the subwatersheds, the HUC12; the catchments, 2.67 million catchments that Al
32:04
spoke to; the flowlines, and the waterbody classifications.
32:07
There had been an attempt to include the dams within this framework, but due to complexities
32:13
and sensitivities with other partner agencies, we decided to eliminate that from this process.
32:19
In order to account for very largescale [inaudible] or regulations, humandriven actions, the National
32:25
Weather Service forecast information basins and points, and for [inaudible] those four
32:30
reservoir locations will be integrated into the NHDPlus.
32:33
And forecast downstreams of major points of diversion or operation will be used to replace
32:39
information coming out of the NFIE hydrologicmodeling framework.
32:43
A major component of this is integrating, not only with the USGS Water Watch and endless
32:50
National Water Information System points, but also with the FEMA National Flood Hazard
32:54
Layer, allowing us to guess at the magnitude and extent of flooding when hydrologic additions
33:01
are higher than normal.
33:06
Next component in NFIE is the hydrologic forecasting system.
33:10
This is very exciting, because it couples some of the emerging capabilities from the
33:16
National Weather Service in the form of this image here you see in the upper left, the
33:20
high resolution rapid refresh atmospheric model.
33:23
This is the onekilometeratmospheric model producing rainfall, temperature, fluxes, wind
33:30
information on over the corners.
33:34
This is then coupled within the WRFHydro framework developed by Dr. David Gochis and his colleagues
33:40
at the National Center for Atmospheric Research to operate the NoahMP land surface model.
33:47
Output from the NoahMP land surface model informs of unrouted runoff is then mapped
33:54
back to the NHDPlus catchment.
33:57
The nodetonode relationships within the NHDPlusV2 that Al spoke to are used by a subalgorithm
34:04
called the repeat model, built by Cedric David while he was studying with Dr. David Maidment
34:09
at the University of Texas, Austin, to route the water using Muskingum equations.
34:14
Finally, the last component of the hydrologic model is the intersections of the impact,
34:19
mapping it back to the catchment scale as well as the FEMA Flood Hazard Layer data.
34:25
I will say that cobbling this system together has been facilitated in no small part to two
34:32
of Dr. Maidment's graduate students, Fernando Salas and Marcelo Somos, both at the University
34:38
of Texas.
34:39
Marcelo is actually postdoc.
34:41
We really would not have been able to keep the place and time we're with NFIE today if
34:47
it hadn't been for their dedicated efforts.
34:50
Like I said, that last step is coupling this with the flood risk zones, so this is using
34:55
the FEMA Flood Hazard Layers within the NHD catchments to map the locations of expected
35:01
flood impacts.
35:03
As we move forward in time, not only generating runoff information is a goal for the National
35:13
Weather Service, but getting to the end point where, as data becomes more and more available
35:17
through efforts led by the USGS such as their 3D Elevation Program, actually acquiring LIDARscale
35:23
data, LIDARdrive data as the scales necessary for dynamically generated [inaudible] mapping.
35:30
This will be explored through some elements of the National Flood Interoperability Experiment.
35:37
Programmatically the NFIE or Interoperability Experiment has been going on for approximately
35:42
10 months.
35:44
Beginning on June 1st, we operate this Summer Institute.
35:46
This is an exciting time where we bring students into the University of Alabama as well as
35:49
into the National Water Center to really refine, develop and push on this system and explore
35:56
these new capabilities.
36:01
We forward out of the modeling component.
36:04
The National Flood Interoperability Experiment will explore emerging web services enabling
36:09
us to better establish a system combining multiple components.
36:12
In this case, we'll leverage the capabilities of ArcGIS online to explore conflation of
36:17
model output with other geospatial data sets.
36:20
We'll leverage CUAHSI and HydroShare projects to explore the issues of publishing and making
36:29
available very large data sets as we prepare for boarding the WRFHydro systems within the
36:35
Weather Service operational supercomputing system at WCOSS.
36:40
We're blowing the doors off of hydrologic data.
36:43
In fact, if we explored all of the data from four or five years old, that runs on the order
36:49
of [inaudible] about three terabytes per day.
36:53
To put that in reference, the outputs from the atmospheric models are on the order of
36:56
three to five gigabytes for operational runs.
37:01
By going to these very high spatial scales, 250meter grids and high temporal scales, one
37:07
hour, we're really creating at a full clock a new, large data issue.
37:15
Finally we're exploring additional components of dissemination.
37:18
This is within the NFIE.
37:20
We will leverage the work done by CIWATER's project principle investigators are Dr. Norm
37:26
Jones and Jim Nelson at Brigham Young University and their Tethys platform, which allows a
37:31
visualization and plotting of streamflow hydrographs at any of the 2.67 million NHDPlus streamlines
37:41
across the country.
37:43
Within the Weather Service, there are 12 Weather Forecast Offices that will, this year, begin
37:50
adding to their Twitter feeds static images of the NHDPlus flowlines and TinyURLs that
37:57
will be disseminated via Twitter.
38:01
We're moving in the right direction.
38:03
We're moving towards using the language of the National Hydrography Data Set to define
38:09
where impacts are, where the flooding is, and communicate that in a more geospatial
38:15
nature to our users.
38:17
I mentioned the Summer Institute a little prematurely.
38:20
This is an effort that is really exciting to me.
38:23
Not only are we looking at the NFIE to explore a new methodology for developing modeling,
38:29
but to establish a new paradigm for working with academia to help foster the development
38:35
and growth of the waterresources engineers, hydrologists and forecasters that will need
38:41
to drive the systems tomorrow.
38:43
The NFIE will start off at the twoweek boot camp on June 1st.
38:44
It will run for approximately seven weeks.
38:46
It will culminate with the Capstone event July 15th through 17th.
38:52
This Capstone event will also coincide with CUAHSI HydroInformatics "Model and Data Interoperability:
38:57
From Theory to Practice" conference.
39:00
This is open to the public.
39:02
Registration is required.
39:04
If you'd like to know more, if you'd like to get a handson view of the next generation
39:09
or glimpsing the forecast of the next generation with modeling, I'd encourage you to attend.
39:15
And who wouldn't want go to Tuscaloosa in the middle of July?
39:18
I hear it's a wet heat.
39:20
People often ask us, "What is NFIE for the Weather Service?"
39:24
I use this slide as it's numbering.
39:28
It is a community initiative.
39:29
It is the brainchild of Dr. Maidment, but has been supported by [inaudible] of other
39:35
federal agencies, private partners and academic sectors.
39:40
It demonstrates an initial set of waterrelateddata services using community standards.
39:45
It will demonstrate the future of realtimeflood simulation and mapping using cuttingedge tools
39:50
and technologies.
39:51
It is also a prototype for a projectbase interdisciplinary educational model for merging scientists and
40:00
technologists.
40:01
The semester curricula and the NWC Summer Institute are two really exciting opportunities
40:06
for us to engage with the next generation of forecasters and hydrologists.
40:10
These are just some of the organizations that have been involved the University of Alabama,
40:16
the University of Texas, NCAR, and our colleagues at the USGS not only in the core science group
40:22
[inaudible] , but across the water and water information groups.
40:27
All of NFIE though, I will say, in summary is predicated on this ability to have a national
40:33
hydrofabric, or in this case, National Hydrography Data set.
40:38
Finally, this is the first example of leveraging the National Water Center as an interagency,
40:44
multiagency and national asset.
40:47
The mission of the National Water Center is scientific excellence and innovation, driving
40:50
water prediction and decision making for a waterresilient nation.
40:55
What we see, what we hope is that this is a center for collaborative research and development,
40:59
a center for nationally consistent operations that complement the current capabilities of
41:05
the Weather Forecast and River Forecast Centers across the country.
41:08
It will support the River Forecast Centers through robust data services, and will support
41:14
missionoriented research and development.
41:17
[inaudible] it's a catalyst for engaging academia.
41:20
It's a proving ground for exploring merging science and technologies.
41:24
It is an effort not only engaging research and development...I'm sorry, research operational
41:33
information exchanges, but also that operations match the research community exchange, helping
41:39
us inform the development community, academic community, what our missions are, what are
41:43
our gaps are, what our science and services needs are.
41:47
By doing so, it's a very symbiotic relationship.
41:51
I believe that's all I have in terms of the review of the NFIE.
42:02
Al, I believe now we time for some questions?
42:05
Al: Yes, Ed.
42:06
We've got time for questions.
42:07
We have actually plenty of time here for questions.
42:10
Go ahead and type your questions into the Q&A box.
42:14
Male Panelist: Al, our first question is from Danielle [inaudible] from Idaho.
42:18
She asked whether or not land disturbances are going to be incorporated into this, such
42:25
as wildfires and how wildfires might disturb the land and change the dynamics of the flooding.
42:31
Ed: Well, I'll take that question.
42:36
The short answer is that within the NFIE itself the parameterization of the land surface is
42:44
relatively limited to what we have available to us today in the National Land Cover Data
42:48
set.
42:49
The longerterm answer is that, as we transition from NFIE to the operational systems within
42:55
the National Water Center, that substantial amounts of research and dedicated development
43:01
activities will be geared at answering that question.
43:04
Realtime in reparameterization of burned areas is probably at the tip of that.
43:10
That's one of the first things that we can do just in terms of the availabilities of
43:15
burn scars from the BAER teams, about burns of wildland fire delineations from the Interagency
43:23
Fire Center.
43:24
But longer term, this could include longer term changes in land use, land cover as gathered
43:30
since by the USGS EROS data center.
43:33
Data simulation for both the shortterm phenomena and longerterm phenomena will certainly be
43:39
included within the modeling system, but that's going to be glimpsed at here with the NFIE.
43:51
Al: Do we have any more questions from people?
43:54
Now is the time to ask.
44:00
Male Panelist: We're not seeing the others right now, Al.
44:07
If you'd like to ask a question, just type it into the Q&A part of the screen.
44:14
[silence] Female Panelist: There are a couple of questions
44:21
coming into the chat section that you might want to take a look there, as well, since
44:27
they're coming in there.
44:29
[silence] Male Panelist: Here's a question from Charlie
44:35
Palmer.
44:36
I think it's actually for you, Al.
44:38
For states lacking NHDPlus as well as comprehensive stream gage network, will a white paper be
44:43
assembled to help encourage the needs of financing these data networks?
44:51
Al: I think that question is, I guess, have we written a white paper on trying to finance
45:03
this?
45:08
We haven't really written a white paper, but within USGS, we are planning on, and I mentioned
45:15
this earlier, building NHDPlus for the high resolution NHD.
45:21
Now, you say for states that don't have NHDPlus right now.
45:28
The only state that doesn't have it is Alaska.
45:33
Alaska is quite challenging just because of the real lack of data up in Alaska.
45:41
We did have a lot of activity going on in Alaska with building data for NHD and also
45:51
the elevation data.
45:53
There's a very active program of acquiring IfSAR data, which is fivemeterresolution data
46:01
that's being acquired for the state of Alaska.
46:06
I think we're about a third of the way through Alaska right now.
46:13
Maybe one of the other people online knows a little bit better of the status of the IfSAR
46:21
acquisition.
46:22
Once the IfSAR is collected, NHD data is being extracted from that.
46:28
That's quite a complex process, but it is happening.
46:32
Eventually we'll have NHD data that will be good quality NHD data up in Alaska, and we'll
46:42
be looking at building NHDPlus data from that.
46:46
I hope that answers the question.
46:50
Male Panelist: Then, we have a question from David Gilbert who asks, "If we have an area
46:56
that is interested in being a pilot for this initiative, how can we get involved?"
47:00
Ed: I'll answer that.
47:02
Al: Yeah, go ahead.
47:04
Ed: But I'm not sure if this is referring to the Nation Flood Interoperability Experiment
47:11
or the broader term, the broader emerging capabilities coming out of the National Water
47:16
Center.
47:17
If that's the case, for the NFIE this experiment is pretty well set in stone for this year.
47:27
I will say that we endeavor to do interoperability experiments in the future and will probably
47:32
advertise that as we did this year through the American Water Resource Association.
47:35
So, look forward to their meeting in November to learn more about the results for this year's
47:40
work, and then how to get involved next year.
47:42
If it is about the more broader capabilities stemming from National Water Center, that
47:51
will be...We aim to provide those services everywhere.
47:56
Through the evaluation phase, we will certainly work to develop prototypedata sets that facilitate
48:02
evaluation by not just internal Weather Service partners, but the broader federal community.
48:10
Male Panelist: All right.
48:14
We have a question here from Kevin [inaudible] that actually might be appropriate to your
48:22
CoChair roles on the Subcommittee for Spatial Water Data, relative to OMB, FGDC and Advisory
48:30
Committee for Water Information.
48:32
"How can we make this information more available to OMB, for example, and help with funding
48:38
and support?"
48:39
So, maybe just expressing where the NFIE and the OWDI in general fit in relation to some
48:47
of those other acronyms.
48:51
Al: Do you want to take that or do you want me to?
49:00
Ed: If you wanted to jump on it, I don't have a great answer for how we interact with OMD.
49:05
But the short answer is, perhaps the safe answer is, we will work through our respective
49:15
agencies to communicate other necessities/requirements through our budget process, that tie back
49:23
to the NHD, so that they can be considered for appropriation.
49:32
I don't know.
49:33
I know, Al, that from the Noah perspective we have been successful in garnering support
49:39
for FY15.
49:42
We're appropriated for an FY15 activity that directly requires Noah to do this work.
49:48
It's called the centralized demonstration, centralized demonstrationmodeling project.
49:52
You can find that in the president's budget.
49:57
I think the strength here is that Noah and USGS told a story of efficiency's gained through
50:05
interagency collaboration.
50:06
Maybe that's part of how that can be successful.
50:12
Al: Yeah, I think that's good.
50:15
I'll just add that there's a an effort going on called the Open Water Data Initiative that's
50:23
being led through the Subcommittee on Spatial Water Data that Ed and I cochair.
50:27
That's getting us a fair amount of notice up at fairly high levels.
50:35
We've done some briefings.
50:36
We've done a number of briefings about the Open Water Data Initiative.
50:43
In this, NFIE is one of our use cases that we are looking at in Open Water Data Initiative
50:50
along with a couple of others.
50:53
I think it's getting a lot of attention.
50:57
There have been budget initiatives.
50:59
We all know there's money in the president's budget, but we all know that that doesn't
51:05
always translate to money in the final budget after congress gets through with it.
51:15
I think it's getting quite a bit of attention.
51:19
I'm expecting that things are going to start shaping up pretty well, hopefully, into the
51:25
future.
51:26
About as much as we probably should say.
51:30
[laughs] Male Panelist: OK.
51:31
Then we have a question from Kate [inaudible] . She says, "Do you foresee the possibility
51:37
of end users being able to model for a specific type of storm or rainfall event, and then
51:44
exporting these as polygons that can be used in a GIF?"
51:51
Ed: I'm not sure if that's directed towards the modeling efforts demonstrated through
51:56
the NFIE or the flow calculations that Al just spoke to.
52:00
I'll answer the former.
52:02
The nature of the forecast system will be such that it will attempt to do this nationally.
52:08
It can't be operated by an end user, it's not that type of a model.
52:15
It's more analogous to how the atmospheric models are developed.
52:18
So grids are generated on the operational supercomputer, then provided to forecasters.
52:24
But to answer the later part of that question, we hope that as the scale in this type of
52:29
operation modeling system is demonstrated and characterized, that it will greatly influence
52:35
the ability for the end forecaster at the Weather Forecast's office to better delineate
52:39
a [inaudible] polygon with much more [inaudible] not only in spatial scale, but in temporal
52:46
scales.
52:47
Al, I don't know if you want to speak to operating the NHDPlus flow calculations for specific...
52:55
Al: It's interesting.
52:57
We've had some discussions about that just in the last couple of days, about code repository.
53:07
There's this concept within the Open Water Data Initiative for something called a "marketplace
53:14
of tools and code."
53:17
It's definitely something that's on the radar, to be able to make tools and databases available.
53:27
Open Water Data Initiative is really focused on trying to do that with both the tools and
53:33
the data.
53:35
I think there's definitely an effort going on in that direction.
53:41
Male Panelist: There's two more questions up.
53:47
First question is from Steve [inaudible] . "If an agency wanted to build tools that use the
53:55
databases and tools, what is the process to help facilitate the architecture development
54:01
and deployment of the applications?"
54:04
I'm going to speculate that relates to the NFIE.
54:11
Ed: Oh, I wasn't expecting it [laughs] was geared at the NHD.
54:15
That's a good question.
54:18
I don't have any answer of the top of my head, mostly because we don't know what the NFIE
54:23
is going to look like.
54:25
I will say that one of the broader goals of the National Water Center is to cultivate
54:31
something equivalent to the weather enterprise for water resources.
54:36
And say that the sooner that we can begin a dialogue with other agencies and with the
54:41
private sector about how they can add value on top of the underlying data sets developed
54:45
by nextgeneration forecasting tools, the better we'll all be.
54:48
Male Panelist: All right.
54:51
There's actually a question came to my mind as well.
54:56
This one is from Cindy Rachel.
54:58
"Will there be comparisons made between the NFIE and static maps created by the USGS Flood
55:06
Inundation Mapping Group?
55:07
Most, if not all, of those maps are produced from running HECRAS and not NHD base.
55:12
It would be interesting to do, to see how these compare.
55:16
Ed: Yeah, I think that's an interesting question.
55:19
We certainly will put the output from WRFHydro through a very robust evaluation framework
55:25
as we move forward with that initiative.
55:27
In terms of the NFIE, I think that can certainly be done.
55:30
I don't know that it's specifically in the student project planning, but that's something
55:37
that I can talk with...My colleagues Fernando Salas and Marcelo Somas will actually be responsible
55:43
for the experimentation itself.
55:47
The other aspect of the NFIE that I think is exciting is exploring how a hydrologic
55:52
model of the national scale on the NHD can be intersected with local and regional hydrologic
55:59
models, perhaps making interoperable with NHD hydrology, or the NFIE hydrology, and
56:07
the hydrologic analyses from underlying RAS models.
56:12
Al: OK.
56:15
I think we probably need to wrap it up at this point.
56:21
If you have more questions we'll leave the Q&A open here for just a little longer.
56:27
I think it'll be open as long as we keep the meeting open.
56:30
So if you want to ask some more questions, go ahead.
56:34
We'll also get written answers to any of the questions that you ask and we'll post those
56:42
on the website in the next few days, hopefully, after we get this all wrapped up.
56:53
We'll also post the presentations out on the website as well.
56:57
We'll be posting the recording, but it takes a while for us to do that because we have
57:03
to do postcaptioning for Section508 compliance.
57:08
That will take longer before we can get the recording posted.
57:15
I just want to wrap up.
57:20
I'd like to announce that the next session in our Hydrography Webinar Series will take
57:26
place on July 30th.
57:33
Finally, I just want to point out again that the recordings and presentations will be posted
58:09
to this website that you see there on the screen.