Data from: Racial composition and homeownership influence the distribution of coastal armoring in South Carolina, United States
Data files
Apr 23, 2025 version files 9.64 MB
-
Armoring_Data.zip
9.63 MB
-
README.md
7.89 KB
Abstract
The desire to stabilize coastlines has led to widespread use of hard armoring infrastructure across the globe, however, ecologists and coastal managers have increasingly documented the deleterious effects of armoring on ecological communities. Although many studies have assessed economic and landscape correlates of armoring, few studies incorporate race as a predictor of armoring. Race may be an important force structuring the placement of armoring due to the long history of Black land loss in the US Southeast. Here, we assessed the distribution of armoring in the US state of South Carolina with respect to demographic and housing characteristics using a high spatial resolution data set and a combination of spatial statistics and generalized linear mixed models. We found clusters of high armoring counts in the more urbanized Beaufort and Charleston counties, with these clusters frequently occurring in large-scale, planned communities. We found a positive correlation between armor count and the percentage of White residents and homeowners in a census block group. Both terms showed a similar magnitude of effect, with the number of armoring structures predicted to increase from 1.61 to 7.77 and from 1.14 to 8.97 between CBGs that are 0 to 100% White and homeowners, respectively. These results highlight that racial composition and homeownership are strong predictors of armoring count on private, personal property, which provides critical context for how these structures are distributed and underscores that socioeconomic factors can control where their associated environmental impacts may be concentrated.
This text document contains a brief description of all the files in the “Armoring_Data” folder. All R scripts used to run the analysis are hosted on the associated Zenodo link and contain a separate README therein (“0_README.R”). If you have any questions or requests, please email me (Jeffrey Beauvais, he/him, beauvais.work@gmail.com).
Description of the Data and file structure
Parent folder: Armoring_ArcPro_Files
Some folders (“BeauvaisByers_Armoring_Index”, “GpMessages”, and “ImportLog”) are empty and automatically generated by ArcGIS Pro when loading the program.
Sub-folder: BeauvaisByers_Armoring.gdb
Contains a geodatabase (.gdb file) with final point layers used in the analysis for dock counts, lengths, and geographic boundaries. Intermediate files were redundant and excluded, but are available upon request. These are necessary for loading the ArcGIS Pro files. The functional files used to generate the raw data are located in the “BeauvaisByers_Armoring.gdb” (.gdb = geodatabase file), which can only be loaded within ArcGIS Pro:
“Armor_Parallel_Raw”: Armoring data filtered to only contain shore-parallel armoring structures. Columns in this file include:
OBJECTID: ID value automatically generated by ArcGIS Pro (unused).
Shape: Automatically generated by ArcGIS Pro to designate the type of shape present (all values are "Polyline").
Field_Chec: Column originally included in the data (unused).
County: The South Carolina county the structure was located.
Imagesourc: Likely the name of the imagery file used by Jackson (2017) to digitize the armoring structure (unused).
ImageDate: Likely the year of the "Imagesourc" file (unused).
Resolution: Likely the resolution of the "Imagesourc" file (unused).
Method: Unused.
Quad: Contains more local geographic names for the locations of armoring structures.
QA_QCPhoto & QA_QCGround: Unused, likely just a reference to any quality assurance/control steps taken by Jackson (2017).
StrucType: ID number given to assign the type of armoring structure. A full breakdown of this hierarchical numbering system is provided on p. 20 of Jackson (2017). A PDF of Jackson (2017) is provided in the "Armoring_Supplemental" folder on Zenodo.
StrucName: The corresponding name for each StrucType.
Shape_Leng & Length & Shape_Length: Identical columns showing the length (in m) of the structure. The Shape_Length column was recreated to account for the new lengths of structures that span multiple census block groups.
HOR_ACC: Original column included in Jackson (2017), unsure of its original use, and not used.
INBlockID: The ID of the census block group the armoring structure was located. Corresponds to the BlockID column in the "Boundaries_CBGs_All" and "Boundaries_CBGs_Analysis" layers. A value of 0 means that the structure crossed block group boundaries.
“Armor_Parallel_Raw”: A copy of the “Armor_All_Raw” file, but filtered to only contain shore-parallel armoring structures.
“Armor_Parallel_Edited_GCS”: A copy of the “Armor_Parallel_Raw” layer,. Contains two new columns, “MidX” and “MidY”, which give the X and Y coordinates of the midpoint of the structure.
“Armor_Parallel_Points_GCS”: Point file containing points placed at the MidX and MidY locations labeled in the “Armor_Parallel_Edited_GCS” file.
“Armor_Parallel_Edited_UTM”: “Armor_Parallel_Edited_GCS” that has been projected to UTM Zone 17N.
“Armor_Parallel_Points_UTM”: “Armor_Parallel_Points_GCS” that has been projected to UTM Zone 17N.
“Boundaries_CBGs_All”: Boundaries for all census block groups downloaded from the TIGER/Line files from the US Census Bureau (https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html). Columns 1-14 are native to the original downloaded files and documentation can be found on the provided URL. Also contains the following fields:
BlockID & JoinField: Duplicate columns providing a simpler, numeric ID provided to each block group for ease of joins and quick identification.
XCoord & YCoord: X and Y coordinates of the centroid of each polygon.
Analysis: String indicating whether that census block group was included in the analysis based on the length of the shoreline.
"Boundaries_CBGs_Analysis": Copy of "Boundaries_CBGs_All" but filtered to only include areas analyzed in the study.
"Boundaires_Counties": As it sounds, outlines of county boundaries.
"Boundaries_SEStates_Background": Boundaries for the states of Georgia, South Carolina, and North Carolina.
"HotSpot_ArmorDensity_6km_SWM": A hot spot analysis of armoring density created using the 'Hot Spot Analysis' tool in ArcGIS Pro and with the "WeightMatrix_CBG_Analysis_6km.swm" file.
"HotSpot_RawCount_6km_SWM":A hot spot analysis of armoring count created using the 'Hot Spot Analysis' tool in ArcGIS Pro and with the "WeightMatrix_CBG_Analysis_6km.swm" file.
"HotSpot_Comparison_Count_Density": A comparison of the count and density hot spot analyses.
"Near_Table_50m": The table created by the 'Generate Near Table' tool.
"PointDen": A point density raster layer of armoring structures.
"WAI_Dock_Points_BOnly_2011_UTM": Contains only the base ("B") points of docks used in the analysis of dock and armoring co-occurrence. This data comes from Beauvais, J., Markley, S. N., & Byers, J. E. (2024). Exponential growth of private coastal infrastructure is influenced by geography and race in South Carolina, USA. Scientific Reports, 14(1), 9114. The attribute table contains the following columns:
OBJECTID & OBJECTID_1: ID value automatically generated by ArcGIS Pro (unused).
Shape: Automatically generated by ArcGIS Pro to designate the type of shape present (all values are "Point").
Unique ID: As it sounds, a unique identifier is given to each dock. Concatenation of the County, Structure_ID, and Point_Position columns.
County: The South Carolina county the structure was located.
Structure_Type: The kind of coastal infrastructure listed.
Structurte_ID: Unique number assigned to each structure.
Point_Position: The location of the point placed on the structure. Entries are all "B" (base, where the structure connects to land).
Notes: Reference notes for the structure.
Year_Added: The year of imagery the structure was first identified in.
Year_Defunct: The year in imagery a structure was noted to have disappeared.
Decade_Added: As with year added, but the entry reports the corresponding census decade used in the analysis.
Decade_Defunct: As with year defunct, but the entry reports the corresponding census decade used in the analysis.
GeoLength: The length of the structure.
X_B & X_E: The longitude of the base (X_B) and end (X_E) point (NAD 1983 geographic coordinate system). Used in tandem with Y_B & Y_E to calculate the length of the structure.
Y_B & Y_E: The latitude of the base (Y_B) and end (Y_E) point (NAD 1983 geographic coordinate system). Used in tandem with X_B & X_E to calculate the length of the structure.
Loose files
“BeauvaisByers_Armoring.atbx”: An ArcGIS “toolbox” file automatically created for each project. Read at the following link for more:
“WeightMatrix_CBG_Analysis_6km.swm”: A spatial weights matrix file used in the creation of the hot spot maps.
Sharing/Access Information
This data is open to use under a CC0 license.
The data analyzed in this study were originally collected by Jackson (2017). Armoring data was associated with census block groups in ArcPro and then analyzed with respect to demographic data gathered from the Census Bureau via GLMMs in R. A copy of the Jackson 2017 is provided in the "Armoring_Supplemental" folder.
Jackson, C. 2017. Mapping Coastal Erosion Hazards Along Sheltered Coastlines in South Carolina 1849 to 2015. South Carolina Department of Health and Environmental Control Office of Ocean and Coastal Resource Management.
ArcGIS Pro and R were used for analyses.