Transitioning Michigan to statewide Elevation-Derived Hydrography data
Background
Today, Michigan has a statewide geographic information systems (GIS) hydrography layer depicting rivers, lakes and streams that was digitized at 1:24,000 scale during the early the 2000's from USGS topographic quadrangles. At 1:24,000 scale, the location accuracy of features can be off as much as 40 feet. The Department of Technology, Management and Budget’s (DTMB) Geospatial Services team is coordinating an effort to improve the GIS hydrography data on a statewide basis using an Elevation-Derived Hydrography approach, with an improved accuracy to between 2 and 5 feet and elevations on the order of 4 to 6 inches. DTMB formed a focus group to explore the needs and activities for a GIS hydrography data improvement project. This group of stakeholders from state agencies engaged with the United States Geological Survey (USGS) to review the pathway to improving the State’s data following the new USGS elevation derived hydrography standards.
Example of generated points, lines, and polygons for an area in this project's generated hydrography mapping efforts
Timeline
NV5 Geospatial is contracted to develop new flowlines, water bodies, water area features derived from the LiDAR and then conflate existing attributes, where present, to the new features. This is expected to be a multi-year project and will be completed on a region-by-region approach during that time. To help with the development of new and more accurate hydrography features, the availability of additional source data from stakeholders across the state will be an important factor to the project. More outreach activities will occur throughout the project to stakeholders in each region about additional data such as culverts, drains, existing hydrography data and other data in GIS format. Please see the additional resource information provided on this web page for more information about this project.
As early as the 1880s, the USGS began using topographic maps to track the observable hydrographic landscape in the United States
image of 1890 historical topographic map for the Newburyport quadrangle, a digital map in the USGS Historical Topographic Map Collection
In the 1970's, these topographic maps were digitized - rivers and waterbodies were given an assigned letter and number value at a 1:5,000,000 scale ..
black and white map image of county outline with drawn rivers and waterbodies shown - dated to USGS maps in 1970's digitized from original topographic maps
.. with corresponding line graphs and information on nodes and areas. The direction of flow was arbitrarily assigned based on what areas of the topographic map were digitized first
corresponding line plot for previous map image, displays nodes and lines and how they relate to one another - dated to USGS maps in 1970's digitized from original topographic maps
In the 1990's the USGS and the EPA began working together to combine the EPA's Reach File schema network (which had good attribute data) with USGS hydrography (which had good line and geometry data) in the hopes of developing an National Hydrography Dataset (NHD) at a 1:100,000-scale, the first NHD was released in 2001
Almost immediately after release of this NHD data, work started on NHDPlus which would transition to data with a 1:24,000-scale, this datset was released in 2007
NHDPlus version two was released in 2012, and offered value-added attributes - like stream order, attributes to calculate rapid stream network queries, and other information
In 2023, the NHD was retired with a national goal towards remapping US hydrography with high-quality elevation data, pulling NHD, WBD, and NHDPlus into one 3DHP dataset with information on lakes, streams, catchments, drainage areas, and other hydrologic features derived from 3DEP LiDAR data
Michigan is at the forefront of this 3DHP data effort, with much of the state either underway or planned for the next few years in this remapping effort
Creating EDH Data
For repositioning hydrography data, the project leverages Light Detection and Ranging (LiDAR) data that have been collected through aerial acquisition projects over the past few years and is at a Quality Level 2 (QL2). QL2 level means it can provide an accuracy of between 2 to 5 feet and with elevations on the order of 4 to 6 inches. LiDAR provides the elevation data needed to detect water channels and flow directions resulting in more accurate positioning of river, lake, and stream systems and more complete coverage of water features by generating more stream networks than what existed in the previous National Hydrography Dataset (NHD). The goals of this project are to remap the rivers, lakes, and streams - leveraging LiDAR QL2 data - to a (1) higher accuracy, (2) greater density, and (3) meet the new USGS Elevation-Derived Hydrography Acquisition Specifications adhering to the Elevation-Derived Hydrography Representation, Extraction, Attribution, and Delineation (READ) rules.
For a 1D feature, accuracy is evaluated by a visual inspection against imagery, and measurements at three points: upstream end, middle, and downstream end of the feature
This headwater lines up with the hillshade valley ...
headwater line imposed on black and white hillshade imagery
... and follows contours of the elevation surface
headwater line imposed on contour elevation imagery
Elevation derived hydrography is used for all features in a dataset
For example, this wider river, just less than 15 meters wide, is captured as a single line stream through the same methodology
contour topology and hillshade image with single flowline shown according to elevation
Through this elevation derived data and subsequent manual review of positional accuracy, a hydrography dataset is created
Elevation derived hydrography is used to convert imagery, hillshade, and geomorphic indices (GMI) into points, lines, and polygons
Project Benefits
As the Great Lakes State, and a state with over 11,000 inland lakes, having a robust hydrography data layer is critical for business systems that need to leverage the location-based aspects of GIS hydrography datasets. These GIS datasets are a key component to performing predictive analysis critical to managing water resources and providing safe fresh water to citizens of our state. Mapped information on rivers, lakes, and streams is a must for any watershed management activity and having these accurate GIS datasets allows for huge amounts of additional data about lakes and river systems to be linked to specific locations along the water system. This results in the ability to conduct analysis and both quickly and accurately display the relationship between the state's water resources and any potential impacts to the water systems.
Another major project benefit is the additional line density compared to the existing NHD data. The left/blue lines below show the previous data and the right/purple lines show the newly generated - and denser - flowlines.
Comparison of existing NHD data (left) and newly generated 3DHP data (right)
There is also an improved positional accuracy in the newly generated data. See both datasets below compared with imagery - the new flowlines (purple) and the polygon waterbody (blue with black outline) more closely follow the observed landscape than the older (orange) flowlines.
Example of improved positional accuracy - NHD (orange) compared to 3DHP (purple and blue)
Once generated, there are many examples of additional data and analysis that is dependent on an accurate and well-maintained hydrography dataset. This includes - but is not limited to - data for invasive species tracking, habitat management and modeling, land and forestry management, wetland mitigation, nutrient loading, endangered species monitoring, and fish stocking. Hydrography data is also key for emergency management scenarios such as flood management and response, wildfire protection, and spill response. Critical infrastructure and assets such as dams, culverts, bridges, roads, trails, rail benefit from more accurate hydrography data for better predictive flow modeling to assess impacts to infrastructure based on changes in flow volumes and water levels. Hydrography data is also important for tourism to map out the navigable waterways and water trails, charter fishing programs, and recreational activities. Enhanced hydrography data is critical for water management and associated projects such as watershed management, soil erosion, coastal waters management, pollution discharge management, wastewater management, agriculture nutrient run-off, and wellhead protection.
After generated, the 3DHP data goes through an extensive manual review process to identify and remap any possible errors. Some examples of some common geometry errors in hydrographic data can be found below.
screenshot of error with topology, text caption reading: lines must only touch at their ends and not intersect or overlap
screenshot of error with topology, text caption reading: a line from one layer must touch line from another with no gaps or overlaps, unless it is a headwater or termination point
screenshot of error with topology, caption text reading: a polygon feature from one layer cannot intersect or overlap itself
screenshot of error with topology, caption text reading: artificial path flowlines must be inside polygon water features
screenshot of error with topology, caption text reading: flowline segment must be digitized in the direction of flow
Click on images for full error description
Find more topology and network error examples and descriptions at USGS
For more information about the project or NHD/3DHP hydrography data please contact Michigan's Hydrography Steward to the USGS, Tim Lauxmann, DTMB lauxmannt@michigan.gov
View the Statewide GIS 3DHP Hydrography Improvement Project webinar slides
USGS
The implementation of 3D National Hydrography Program (3DHP) marks the start of a new era of water data - the first systematic remapping of the nation's hydrography since the original USGS 1:24,000 - scale topographic mapping program was active
