USGS Perspective on The Ele-Hydro Concept This is a proposal and concept by the USGS to drive the GIS community to think about how elevation and hydrography data can coexist in a GIS just as they do in nature. We will not solve all the issues, but rather point out that we need to solve these issues. Adapted from a presentation by Jeff Simley USGS National Hydrography Data Manager

Erosion and Resistance Elevation and Hydrography data are inherently linked. The gravitational-erosional force of water sculpts the landscape and that sculpting creates incisions in the terrain in which the water flows. The type of soil, rock and vegetation and the arrangement of these landscape characteristics via geologic processes, determines how water moves over the surface. Thus, Elevation and Hydro form each other in a symbiotic relationship.

Integration and registration Elevation and Hydrography GIS data have a chicken and egg type relationship. There is considerable debate whether the elevation comes first or the hydrography comes first. You might think the elevation comes first, but in cartography, and photogrammetry, such as the USGS topographic mapping program, the water always came first. 1- We mapped in 3D and plotted in 2D. I think this is an important distinction for us to consider today. One of the first things plotted in the stereo model were the streams.

Contours integrated with Hydro Heads-up Digitizing http://letters-sal.blogspot.com/2011/05/digitizing-wacom-way.html Contours integrated with Hydro Then the contours were molded and shaped around those streams. With the advent of orthoimagery and GIS technology we went from mapping in 3D photogrammetrically to mapping entirely in 2D and 1- so for many years now we have been mapping in 2 Dimensional space, i.e. heads-up digitizing directly from the computer screen. If elevation and hydrography are in a symbiotic relationship, why does GIS collect, manage, and package the two as discrete products? Well, because it’s a convenient way to manage these data.

How we think and collect now. . . Separate funding mechanisms $$$ Indiana LiDAR $3.3million Indiana Hydro $2.6million Managed separately Hydro National Hydro Data 100k 24k Local Data Watersheds Vector Elevation Stovepiped Empires Turfs National Elevation Data PointCloud DEM Raster Elevation is typically in a raster format and hydrography is typically in a vector format. 1- Because of these differences they use 2- totally separate production processes. 3- It’s easier to have two separate programs, 4- an elevation program and a hydrography program. And that might not change.   5- In GIS we create and perpetuate the stovepiping with the GIS layer concept. And there is a lot of merit to that, but maybe it’s not entirely justifiable. Managing data by individual layers is a fundamental 2D concept with GIS, but with high-resolution elevation data we can consider the 3D landscape as a holistic feature, similarly as we did when we collected photogrammetrically. Collecting in 2D layers can work, but as expected, 6- eventually we encounter data that just doesn’t integrate and we wonder why. 7- Usually it’s because we have isolated our programs in stovepipes and they are out of synchronization. 8- In addition, separately managed programs create increased costs. Separate data standards Data housed separately (raster vs vector) Separate collection methods

High Resolution lidar DEM However, the world of elevation data is quickly changing with the advent of highly accurate LiDAR data. The LiDAR data is of recent vintage and thus captures the most current forms on the landscape. And it also has a very fine resolution and can represent more detail of the earth’s surface.

Indiana’s National Hydrography Data Update Project Illinois - Indiana High-resolution lidar has caused the hydrography community to re-think their requirements regarding the density of the drainage network. Their long-desired need for more resolution in the stream network, and better temporal accuracy, is now at hand. We now have the necessary discrimination in the landscape via the elevation data to detect an entirely new density and pattern of streams. But we are still stuck with this problem - we have the elevation department and the hydrography department - the elevation layer and the hydrography layer.

LiDAR Point Cloud Drainage features Wetlands DEM Catchments Subbasins Hydro Drainage features Wetlands DEM Hydro-Flatten Hydro-Enforced Watershed Boundaries LiDAR Point Cloud Catchments Subbasins Floodplains Infra- structure Building footprints Utilities Roof analysis Transportation signage Maybe we need to think in terms of having elevation and hydrography data as a vertically integrated program. Managed, funded and collected as components of the landscape and not as individual layers. And not only should we be collecting elevation and hydrography as an integrated feature, but all landscape features could be collected as well, such as watershed boundaries, stream catchments, vegetation, geologic rock units, wetlands, soils, floodplains, and the list goes on.   As an example, combining the elevation, the drainage boundaries, and hydrography you have a start to a truly integrated landscape product. One that doesn’t stop here, because the drainage boundaries can also be intersected with other landscape factors such as precipitation, temperature, and landcover, to produce an integrated landscape data set that can generate information related to streamflow and provide information for many other science-based problems. Veg Forest Canopy Forest Health Crop Health Grasslands Soils

Some integration occurring. . . How we think now - Some integration occurring. . . Hydro Elevation Combined USGS Apps: StreamStats NHDPlus U.S. Topo This not to say that no one is doing integration of elevation and hydro data. In fact, the USGS and many other organizations are doing post integration of elevation and hydrography data. However, it’s integrating data that’s out-of-sync, but in order to produce a more accurate product we are forced to integrate the elevation and hydro. Forcing integration by “burning” the Hydro data into the Elevation data

U.S. Topo This slide, shows an application called StreamStats which is a Web-based (GIS) that allows users to easily obtain streamflow statistics, drainage-basin characteristics, and other information for user-selected sites on streams. However, this application was created by a post integration of the elevation and hydrography layers. This next slide is a zoomed-in view of the USGS Digital Topographic Map, named U.S.Topo. It too has post integration of the elevation and hydrography layers so that contour turnbacks in stream valleys align with the hydro from the National Hydrography streams. In order to produce these applications we have to force the elevation layer, the DEM, to agree with the hydrography layer by “burning-in” the streams into the DEM. However, when either of these layers is updated, they immediately become out-of-sync with each other and so each time we create a revised U.S.Topo or a revised streamstats application, it will again be necessary to force the integration of the elevation and hydrography layers. In addition, because these layers are managed separately, they are temporally disconnected, i.e. the hydro layer may be 10-30 years older than the elevation layer, or vise-versa. Our streamstats and U.S.Topo applications are what I call post-integration. This is not what I’m talking about in this presentation. I’m not talking about forced integration of two separate data themes, but rather the integration from the very beginning of acquisition, an integration of GIS layers from the source.

Packaging the Data NHD Wetlands Catchments Landcover Vegetation Elevation Vegetation Landcover Catchments Wetlands NHD An inconvenient problem is that hydro and elevation are vector and raster datasets. So the packaging might keep the two data layers separate. But the total package could contain many other landscape factors for a more complete and integrated landscape dataset.

Derive hydrography from the terrain Ultimately, we want to be able to derive the hydrography from the terrain. If the elevation and hydrography are in a symbiotic relationship, with excellent elevation data such as that derived from LiDAR, we can do this. Currently, there is work being done to derive hydrography from LiDAR, but it can be an extremely difficult process. In high frequency terrain it’s rather a straight-forward process, but it’s adversely affected by man-made features such as road embankments. In flat terrain it can be almost impossible to derive.

Indiana’s Local-Resolution NHD streams and lidar DEM We already have good vector hydrography as with the USGS National Hydrography Data, especially with Indiana’s NHD Update Program, including a fairly rich set of attributes. That set of rich attributes can be greatly expanded by integrating it with elevation data. Customers are demanding that we need to go one or two more steps to get desired drainage densities, with better accuracy, better precision, and better integration, and we are at the doorstep with our technology that can make this happen.

Integration – What does it mean? Alignment of elevation and hydrography such that streams flow in channels Interoperability such that a levee can be connected to a river Data Model that links elements and features Synthesis such that streamflow can be estimated Derivatives such that gradient, channel dimensions, etc. can be calculated Program where a DEM and Hydro can be jointly produced and funded Synchronization so that data are temporally coincident Product in a way that both elevation and hydro can be packaged together Delivery of data to be accessed in unison So what does Integration mean? In the case of integrating elevation and hydrography it can involve nine components: (1) The alignment of elevation and hydrography such that streams flow in channels. (2) Interoperability such that a levee can be connected to a river. (3) A Data model that links the elements. (4) Synthesis such that streamflow can be estimated, (5) Derivatives such that gradient and channel dimensions can be calculated. (6) An integrated Program such that a DEM and hydrography can be jointly produced and funded. (7) Synchronization such that the data are spatially and temporally coincident. (8) An integrated Product such that the data can be packaged together. (9) Delivery such that the data can be accessed in unison.

Integration – Why? Reduce cost Provide improved credibility Produce accurate data Increase the value of the data Empower science-based decisions Challenge and drive research We need to integrate elevation and hydrography for six reasons: (1) To reduce cost, (2) To provide improved credibility, (3) To produce more accurate data, (4) To produce more value added data, (5) To better empower decisions for science-based problems, (6) To challenge and drive research in this area.

An Integrated Landscape In conclusion. We need to be thinking in terms of the “next-generation” data package that can delivere “ready-to-use” datasets of the landscape that stand on top of a foundation of elevation and hydrography data; truly integrated from the source, and working in the same symbiotic relationship as represented in nature. “All geospatial software is moving to 3D. It is now an expectation.” Steve Snow, ESRI Imagingnotes.com, http://www.imagingnotes.com/go/article_freeJ.php?mp_id=292