Making Headland Classes from a Geospatial Mash-Up

Douglas A. George, Bodega Marine Laboratory, UC Davis/Applied Marine Sciences, Inc.

Elena E. vandebroek, Department of Hydraulic Engineering, Delft University of Technology, The Netherlands

 

A headland-sized knowledge gap

Rocky headlands are prominent coastal morphological features that may focus wave energy, shed eddies, deflect or block alongshore currents and sediment transport, or create down-current retention zones (e.g., Davies et al., 1995; Alaee et al., 2004; Winant, 2006). Presently, we make assumptions about the relationships between sediment flux and sediment reservoirs, which are important when analyzing sand mining in the Bay or beach nourishment on the outer coast. Whereas extensive studies of nearshore physical processes have resulted in a good understanding of alongshore and cross-shore transport at beaches, flow and sediment transport along rocky shores and around headlands remains poorly understood.

Research gaps and societal needs prompted us to explore how headlands affect circulation and transport rates of sediment or biota. Unlike other physical marine features that have existing classification structures (e.g., beaches [Wright and Short, 1984 and Scott et al., 2011], coral reefs [Freeman et al., 2012], and submarine canyons [Harris and Whitney, 2011]), headlands are not systematically categorized. Developing a classification for headlands will open new avenues for research, both in explaining these headland types (e.g., geological framework or rock types) and in determining the effect of different headland types on flow, sediment transport, and associated geomorphology and ecology. This article describes how we used GIS to create a database of headland shape, size, complexity, and nearshore bathymetry to categorize different types of headlands along the California coast. George et al., 2015 (submitted to Marine Geology, January 2015) provides a more complete description of our study and findings.

headlands along the ocean, headlands in the bay

The planform details of headlands are identifiable from satellite imagery, maps, and geodatabases that include digitized shorelines. Headlands are frequently associated with cliff-backed shorelines, which comprise approximately 80% of the ocean’s coasts (Emery and Kuhn, 1982). Inside San Francisco Bay, shown in figure 1, some headlands create dynamic locations for navigation and habitats – e.g., Point San Pablo, Point Richmond, and Coyote Point. Other in-bay headlands are well known to the surfing, kayaking, and sailing communities for their turbulence – e.g., Fort Point at the Golden Gate, Yellow Bluff in Richardson Bay, and Point Blunt on Angel Island. 

Figure 1. Aerial image of San Francisco Bay region showing flow and sediment transport patterns around selected headlands. The headlands vary in size and shape from Yellow Bluff (flat, small) to Fort Point and Pt. San Pablo (sharp, large). Satellite photograph in natural color from NASA Operational Land Imager (OLI) on Landsat 8 on April 16, 2013 (http://earthobservatory.nasa.gov/IOTD/view.php?id=81238).

Figure 1. Aerial image of San Francisco Bay region showing flow and sediment transport patterns around selected headlands. The headlands vary in size and shape from Yellow Bluff (flat, small) to Fort Point and Pt. San Pablo (sharp, large). Satellite photograph in natural color from NASA Operational Land Imager (OLI) on Landsat 8 on April 16, 2013 (http://earthobservatory.nasa.gov/IOTD/view.php?id=81238).

This study spanned the 1,800-km outer California coast, which comprises a variety of beaches and geological features: approximately 28% pocket beaches, 32% sandy beaches, and 39% rocky shoreline (Scholar and Griggs, 1998). Headlands are associated with all the coastline types with particular prominence in creating pocket beaches and defining rocky shorelines. We identified 78 outer-coast headlands using USGS geological maps, remote-sensing imagery, NOAA navigational charts, and shoreline characterization geospatial databases from the California Coastal Sediment Management Workgroup (CSMW). Several potential entries were excluded for not satisfying these criteria; adjustments to them may increase the number of headlands in the database. For the present, we chose to bypass the Bay to maintain the ocean focus of the study.

We developed the following selection process for headlands: 1) identify a perturbation in the coastline in the remotely-sensed imagery; 2) confirm a named headland in the navigational charts; 3) cross-confirm with similar geology units in the geological maps; and 4) identify similarities in shoreline characterization in the CSMW geospatial database. See figure 2. Criterion #2 preferentially selects headlands that are substantial in relative size, with the smallest headland two hectares in area. At each headland a baseline was obtained by creating the straight coastline that would exist without that headland. 

 

Figure 2. Selection process for identifying a headland in ArcGIS using four sources of information, including layers from the National Atmospheric and Oceanic Administration (NOAA), United States Geological Survey (USGS), and the California Sediment Management Workgroup (CSMW). The example is Bodega Head on the Sonoma County coastline north of San Francisco. Using these layers, 78 headlands were selected between Point St. George (Crescent City) and Point Loma (San Diego). No headlands were identified inside San Francisco Bay because of the outer-coast focus of the study.

Figure 2. Selection process for identifying a headland in ArcGIS using four sources of information, including layers from the National Atmospheric and Oceanic Administration (NOAA), United States Geological Survey (USGS), and the California Sediment Management Workgroup (CSMW). The example is Bodega Head on the Sonoma County coastline north of San Francisco. Using these layers, 78 headlands were selected between Point St. George (Crescent City) and Point Loma (San Diego). No headlands were identified inside San Francisco Bay because of the outer-coast focus of the study.

gis for geometry and bathymetry

From the geospatial database, we extracted several geometric parameters using a variety of ESRI tools and third-party extensions (XTools Pro and ET Geowizards). See figure 3. These parameters quantify the size, symmetry, and complexity of each headland and its relationship to the general trend of the coastline. For some parameters, it was appropriate to take an “upcoast” (‘up’) and a “downcoast” (‘dn’) measurement. A few examples of these parameters include: 1) “perimeter length of a headland” derived by subtracting the baseline length (L) from the overall perimeter of each headland polygon; 2) “apex angle,” defined as the angle of the ocean-facing front of a headland, determined by summing the up and down angles between the cross-headland transect and along-headland transects (αup and αdn, respectively); and 3) “bathymetric slope” described in more detail below. Additional parameters were derived from these measured parameters, including aspect ratio (headland width/length), and rugosity (baseline length/perimeter length).

Figure 3. Schematic and Bodega Head example of geometric parameters gathered in ArcGIS to generate the database for classification of California headlands. Various tools were used to automate the calculations of length, angles, and size (perimeter and area). The difference between up and down angles emerged as an additional metric of the asymmetry of a headland (phi) and the pointedness of the apex angle (alpha).

Figure 3. Schematic and Bodega Head example of geometric parameters gathered in ArcGIS to generate the database for classification of California headlands. Various tools were used to automate the calculations of length, angles, and size (perimeter and area). The difference between up and down angles emerged as an additional metric of the asymmetry of a headland (phi) and the pointedness of the apex angle (alpha).

The underwater expression of a headland was determined by extracting bathymetry from merged topographic and bathymetric digital elevation model (DEM) of the California Seafloor Mapping Program and California Shoreline Mapping Project, a joint study by the California Ocean Protection Council, USGS, and NOAA. This database is enormous – on the scale of two terrabytes – with the high-resolution acoustic surveys in LAS v1.2 format and DEM data in ERDAS IMG format (1,500 m x 1,500 m tiles). The full project area includes four production blocks spanning approximately 16,000 km2 and 17 counties. NOAA did not generate contours, however, which led us to develop an efficient contouring protocol. We only required data along five transects for each of the 78 headlands, shown in figure 4. Three transects perpendicular to the headland and a shore-normal reference transect on either side of the headland, where the shoreline is approximately straight and aligns with the baseline of the headland. Rather than generate bathymetry contours for the entire dataset, we built three models to cull through the database’s 8,040 tiles and generated contours only in the relevant tiles. An abbreviated description of the process is: 1) using a transect, identify relevant DEM tiles; 2) create contours at 0, 5, 10, 15, 20, 25, 30, 40, 50, and 75 m NAVD88 within those tiles; 3) find intersections of the transects with the contours; 4) in the case of multiple intersections (caused by artifacts in the contouring process or offshore linear bars), average the distances to reduce to single intersections for each contour (for alongshore bars and shoals, the first intersection was used); and 5) calculate the distance between the shoreline and each contour. Outside of GIS, we derived a suite of ratios and bathymetric slopes from these distances for further analysis in conjunction with the geometric values described earlier.

Figure 4. Bathymetry extraction at Bodega Head from the merged California Seafloor Mapping Program and California Shoreline Mapping Project database. Distance to specific contour depths was determined by identifying which raster tiles intersected the transects and generating contours prior to distance calculations.

Figure 4. Bathymetry extraction at Bodega Head from the merged California Seafloor Mapping Program and California Shoreline Mapping Project database. Distance to specific contour depths was determined by identifying which raster tiles intersected the transects and generating contours prior to distance calculations.

conclusion

From merged shoreline and bathymetric headlands parameters, we used cluster analysis to group features in a way that maximizes the difference between the groups and minimizes the difference within a group. Several of the basic parameters (e.g., size, shape, shoreline complexity) suggested clusters, but ultimately three parameters – perimeter length, bathymetric slope ratio, and apex angle – were found to jointly classify the headlands into eight groups, as seen in table 1. For more details about the cluster analysis and conclusions, see George et al., 2015 (Marine Geology, submitted January 2015). A similar approach to classifying headlands inside San Francisco Bay eventually would be possible, because the California Ocean Protection Council, USGS, and NOAA consortium is mapping the bathymetry to the same high resolution as was done on the outer coast through the California Seafloor Mapping Program.

Acknowledgements

This work was supported in part by the Hydrologic Sciences Graduate Group at the University of California, Davis, and the US Geological Survey’s Coastal and Marine Geology program. The authors wish to thank John Largier (Bodega Marine Laboratory), Patrick Barnard and Curt Storlazzi (USGS), Dennis Hall (NOAA), and Joshua Novac and Erin Musgrave (Dewberry).

Douglas George can be reached at dgeorge@ucdavis.edu.

References

Alaee, M.J., Ivey, G., Pattiaratchi, C., 2004. Secondary circulation induced by flow curvature and     Coriolis effects around headlands and islands. Ocean Dynamics 54, 27-38.

California Sediment Management Workgroup Shoreline Characterization Database                         (http://www.dbw.ca.gov/csmw/SpatialData.aspx)

California Seafloor Mapping Program and California Shoreline Mapping Project                               (http://walrus.wr.usgs.gov/mapping/csmp/index.html)

Davies, P.A., Dakin, J.M., Falconer, R.A., 1995. Eddy Formation Behind A Coastal Headland.             Journal of Coastal Research 11, 154-167.

Emery, K.O., Kuhn, G.G., 1982. Sea Cliffs: Their processes, profiles and classification. Geological     Society of America Bulletin 93, 644-654.

Freeman, L.A., Miller, A.J., Norris, R.D., Smith, J.E., 2012. Classification of remote Pacific coral         reefs by physical oceanographic environment. Journal of Geophysical Research-Oceans 117,       C02007, doi:10.1029/2011JC007099.

George, D.A., Largier, J.L., Storlazzi, C.D., Barnard P.L. Classification of rocky headlands in             California with relevance to littoral cell boundary delineation. Submitted to Marine Geology,     January 2015.

Harris, P.T., Whiteway, T., 2011. Global distribution of large submarine canyons: Geomorphic           differences between active and passive continental margins. Marine Geology 285, 69-86.

Scholar, D.C., Griggs, G.B, 1998. Pocket Beaches of California: Sediment Transport Along a             Rocky Coastline. California's Coastal Natural Hazards. University of Southern California Sea       Grant Program.

Scott, T., Masselink, G., Russell, P., 2011. Morphodynamic characteristics and classification of         beaches in England and Wales. Marine Geology 286, 1-20.

Winant, C.D., 2006. Three-dimensional wind-driven coastal circulation past a headland. Journal     of Physical Oceanography 36, 1430-1438.

Wright, L.D., Short, A., 1984. Morphodynamic variability of surf zones and beaches - a                  synthesis. Marine Geology 56, 93-118. 

Summer 2015 Volume 8 Issue 1