Data Management Plan

To support the modeling of storm-induced flooding, the USGS Coastal National Elevation Database (CoNED) Applications Project has created an integrated 1-meter topobathymetric digital elevation model (TBDEM) for the Northern California Coast. High-resolution coastal elevation data is required to identify flood, hurricane, and sea-level rise inundation hazard zones and other earth science applications, such as the development of sediment transport and storm surge models. This TBDEM consists of the best available multi-source topographic and bathymetric elevation data for the onshore and offshore areas in Northern California. The Northern California TBDEM integrates 26 different topographic and bathymetric data sources including LiDAR point clouds, hydrographic surveys, single-beam acoustic surveys, and multi-beam acoustic surveys obtained from USGS, NOAA, and USACE. The topographic and bathymetric surveys were sorted and prioritized based on survey date, accuracy, spatial distribution, and point density to develop a model based on the best available elevation data. Because bathymetric data are typically referenced to tidal datums (such as Mean High Water or Mean Low Water), all tidally-referenced heights were transformed into orthometric height based on a common geoid (Geoid09) that are normally used for mapping elevation on land based on the North American Vertical Datum of 1988. Every input data source in the TBDEM has been horizontally referenced to UTM Zone 10, NSRS2007. The spatial resolution is 1-meter with the general location ranging from the Golden Gate Bridge to the Oregon border, and extending inland to an elevation of at least +10 m and offshore to the 3-nautical-mile limit of California's State Waters. The overall temporal range of the input topography and bathymetry is 1986 to 2019. The topography surveys are from 2009-2019. The bathymetry is from 1986-2019. The nearshore void zone (not covered by lidar or sonar) was filled by interpolating points from adjacent datasets.

Process Steps:

• 2020-04-28 00:00:00 - The principal methodology for developing the integrated topobathymetric elevation model can be organized into three main components. The "topographic component" consists of the land-based elevation data, which is primarily comprised from high-resolution LiDAR data. The topographic source data will include LiDAR data from different sensors (Topographic, Bathymetric) with distinct spectral wavelengths (NIR-1064nm, Green-532nm). The "bathymetric component" consists of hydrographic sounding (acoustic) data collected using boats rather than bathymetry acquired from LiDAR. The most common forms of bathymetry that are used include: multi-beam, single-beam, and swath. The final component, "Integration", encompasses the assimilation of the topographic and bathymetric data along the near-shore based on a predefined set of priorities. The land/water interface (+1 m- -1.5 m) is the most critical area, and green laser systems, such as the Experimental Advanced Airborne Research LiDAR (EAARL-B) and the Coastal Zone Mapping and Imaging LiDAR (CZMIL) that cross the near-shore interface are valuable in developing a seamless transition. The end product from the topography and bathymetry components is a raster with associated spatial masks and metadata that can be passed to the integration component for final model incorporation. Topo/Bathy Creation Steps: Topography Processing Component: a) Quality control check the vertical and horizontal datum and projection information of the input lidar source to ensure the data is referenced to NAVD88 and NAD83, UTM. If the source data is not NAVD88, transform the input LiDAR data to NAVD88 reference frame using current National Geodetic Survey (NGS) geoid models and VDatum. Likewise, if required, convert the input source data to NAD83 and reproject to UTM. b) Check the classification of the topographic LiDAR data to verify the data are classified with the appropriate classes. If the data have not been classified, then classify the raw point cloud data to non-ground (class 1) ground (class 2), and water (class 9) classes using LP360-Classify. c) Derive associated breaklines from the classified LiDAR to capture internal water bodies, such as lakes and ponds and inland waterways. Inland waterways and water bodies will be hydro-flattened where no bathymetry is present. d) Extract the ground returns from the classified LiDAR data and randomly spatial subset the points into two point sets based on the criteria of 95 percent of the points for the "Actual Selected" set and the remaining 5 percent for the "Test Control" set. The "Actual Selected" points will be gridded in the terrain model along with associated breaklines and masks to generate the topographic surface, while the "Test Control" points will be used to compute the interpolation accuracy (Root Mean Square Error) from the derived surface. e) Generate the minimum convex hull boundary from the classified ground LiDAR points that creates a mask that extracts the perimeter of the exterior LiDAR points. The mask is then applied in the terrain to remove extraneous terrain artifacts outside of the extent of the ground LiDAR points. f) Using a terrain model based on triangulated irregular networks (TINs), grid the "Actual Selected" ground points using breaklines and the minimum convex hull boundary mask at a 1-meter spatial resolution using a natural neighbor interpolation algorithm. g) Compute the interpolation accuracy by comparing elevation values in the "Test Control" points to values extracted from the derived gridded surface; report the results in terms of Root Mean Square Error (RMSE).
• 2020-04-28 00:00:00 - Bathymetry Processing Component: a) Quality control check the vertical and horizontal datum and projection information of the input bathymetric source to ensure the data is referenced to NAVD88 and NAD83, UTM. If the source data is not NAVD88, transform the input bathymetric data to NAVD88 reference frame using VDatum. Likewise, if required, convert the input source data to NAD83 and reproject to UTM. b) Prioritize and spatially sort the bathymetry based on date of acquisition, spatial distribution, accuracy, and point density to eliminate any outdated or erroneous points and to minimize interpolation artifacts. c) Randomly spatial subset the bathymetric points into two point sets based on the criteria of 95 percent of the points for the "Actual Selected" set and the remaining 5 percent for the "Test Control" set. The "Actual Selected" points will be gridded in the empirical bayesian kriging model along with associated masks to generate the bathymetric surface, while the "Test Control" points will be used to compute the interpolation accuracy (Root Mean Square Error) from the derived surface. d) Spatially interpolate bathymetric single-beam, multi-beam, and hydrographic survey source data using an empirical bayesian krigging gridding algorithm. This approach uses a geostatistical interpolation method that accounts for the error in estimating the underlying semivariogram (data structure - variance) through repeated simulations. e) Cross validation - Compare the predicted value in the geostatistical model to the actual observed value to assess the accuracy and effectiveness of model parameters by removing each data location one at a time and predicting the associated data value. The results will be reported in terms of RMSE. f) Compute the interpolation accuracy by comparing elevation values in the "Test Control" points to values extracted from the derived gridded surface; report the results in terms of RMSE.
• 2020-04-28 00:00:00 - Mosaic Dataset Processing (Integration) Component: a) Determined priority of input data based on project characteristics, including acquisition dates, cell size, retention of features, water surface treatment, visual inspection and presence of artifacts. b) Develop an ArcGIS geodatabase (Mosaic Dataset) and spatial seamlines for each individual topographic (minimum convex hull boundary) and bathymetric raster layer included in the integrated elevation model. c) Generalize seamline edges to smooth transition boundaries between neighboring raster layers and split complex raster datasets with isolated regions into individual unique raster groups. d) Develop an integrated shoreline transition zone from the best available topographic and bathymetric data to blend the topographic and bathymetric elevation sources. Where feasible, use the minimum convex hull boundary, create a buffer to logically mask input topography/bathymetry data. Then, through the use of TINs, interpolate the selected topographic and bathymetric points, if required, to gap-fill any near-shore holes in the bathymetric coverage. Topobathymetric LiDAR data sources such as the EAARL-B or CZMIL systems provide up-to-date, high-resolution data along the critical land/water interface within inter-tidal zone. e) Prioritize and spatially sort the input topographic and bathymetric raster layers based on date of acquisition and accuracy to sequence the raster data in the integrated elevation model. f) Based on the prioritization, spatially mosaic the input raster data sources to create a seamless topobathymetric composite at a cell size of 1 meter using blending (spatial weighting). g) Performed a visual quality assurance (Q/A) assessment on the output composite to review the mosaic seams for artifacts. h) Generate spatially referenced metadata for each unique data source. The spatially referenced metadata consists of a group of geospatial polygons that represent the spatial footprint of each data source used in the generation of the topobathymetric dataset. Each polygon is to be populated with attributes that describe the source data, such as, resolution, acquisition date, source name, source organization, source contact, source project, source URL, and data type (topographic LiDAR, bathymetric LiDAR, multi-beam bathymetry, single-beam bathymetry, etc.).
• NOAA OCM retrieved the Topobathy DEM file from the USGS at the following link: https://www.sciencebase.gov/catalog/item/5ebc4ba682ce25b51365d660 That data were in UTM 10N, NAD83(NSRS2007) and NAVD88 heights using geoid09, with all units in meters. The data were in a single tiff file. To process to the Digital Coast, NOAA OCM used the gdal_retile python script to break the file into smaller tiles. Additionally, the gdal cloud_optimize python script was used to assign the geokeys and add compression.

Data is available online for bulk or custom downloads