Travis K. Sterne
Elizabeth A. Pendleton
Erika E. Lentz
Rachel E. Henderson
20230228
Coastal Change Likelihood in the U.S. Northeast Region: Maine to Virginia - Maximum Change Likelihood
1.0
raster digital data
data release
DOI:10.5066/P96A2Q5X
Woods Hole Coastal and Marine Science Center, Woods Hole, MA
U.S. Geological Survey, Coastal and Marine Hazards and Resources Program
https://doi.org/10.5066/P96A2Q5X
https://www.sciencebase.gov/catalog/item/6197cb8dd34eb622f692ee19
Travis K. Sterne
Elizabeth A. Pendleton
Erika E. Lentz
Rachel E. Henderson
2023
Coastal Change Likelihood in the U.S. Northeast Region: Maine to Virginia
raster digital data
data release
DOI:10.5066/P96A2Q5X
Reston, VA
U.S. Geological Survey, Coastal and Marine Hazards and Resources Program
Suggested citation: Sterne, T.K., Pendleton, E.A., Lentz, E.E., and Henderson, R.E., 2023, Coastal Change Likelihood in the U.S. Northeast Region: Maine to Virginia: U.S. Geological Survey data release, https://doi.org/10.5066/P96A2Q5X.
https://doi.org/10.5066/P96A2Q5X
https://www.sciencebase.gov/catalog/item/6197cb8dd34eb622f692ee19
Coastal resources are increasingly impacted by erosion, extreme weather events, sea-level rise, tidal flooding, and other potential hazards related to climate change. These hazards have varying impacts on coastal landscapes due to the numerous geologic, oceanographic, ecological, and socioeconomic factors that exist at a given location. Here, an assessment framework is introduced that synthesizes existing datasets describing the variability of the landscape and hazards that may act on it to evaluate the likelihood of coastal change along the U.S coastline within the coming decade. The pilot study, conducted in the Northeastern U.S. (Maine to Virginia), is comprised of datasets derived from a variety of federal, state, and local sources. First, a decision tree-based dataset is built that describes the fabric or integrity of the coastal landscape and includes landcover, elevation, slope, long-term (>150 years) shoreline change trends, dune height, and marsh stability data. A second database was generated from coastal hazards, which are divided into event hazards (e.g., flooding, wave power, and probability of storm overwash) and persistent hazards (e.g., relative sea-level rise rate, short-term (about 30 years) shoreline erosion rate, and storm recurrence interval). The fabric dataset is then merged with the coastal hazards databases and a training dataset made up of hundreds of polygons is generated from the merged dataset to support a supervised learning classification. Results from this pilot study are location-specific at 10-meter resolution and are made up of four raster datasets that include (1) quantitative and qualitative information used to determine the resistance of the landscape to change, (2 & 3) the potential coastal hazards that act on it, (4) the machine learning output, or Coastal Change Likelihood (CCL), based on the cumulative effects of both fabric and hazards, and (5) an estimate of the hazard type (event or persistent) that is the likely to influence coastal change. Final outcomes are intended to be used as a first order planning tool to determine which areas of the coast may be more likely to change in response to future potential coastal hazards, and to examine elements and drivers that make change in a location more likely.
CCL is a first order planning tool that estimates the likelihood that an area of coast will experience change based on its inherit resistance to change, metrics associated with specific land cover types, and the hazards that impact a coast. The CCL Maximum Change Likelihood is the combination of supervised learning outcomes from the Fabric, Perpetual, and Event hazards. Each 10 mpp raster cell is assigned a value between 1 and 10 that is an estimate of change likelihood, where 1 is low and 10 is high, based on an ordinal scale. All relevant information pertaining to each grid cell is stored in the associated attribute table. This dataset covers the Northeast US coastline between +/- 10 meters elevation relative to mean high water (MHW) from Maine to Virginia.
2010
2021
ground condition of source data
None planned
-77.5279
-66.8837
45.1930
36.5149
ISO 19115 Topic Category
geoscientificInformation
oceans
None
U.S. Geological Survey
USGS
Coastal and Marine Hazards Mission Area
Woods Hole Coastal and Marine Science Center
Coastal Fabric
Elevation
Interpretation
Bathymetry
Landcover
Land Cover
Topography
UVVR
Unvegetated-Vegetated Ratio
Shoreline Change
Coastal Hazards
High Tide Flooding
Storm Recurrence
Wave Power
Storm Overwash
Sea Level Rise
Coastal Change Hazard Assessment
Coastal Vulnerability Index
Machine Learning
Autoclassification
Automation
Arcpy
ArcGIS Pro
Support Vector Machine
Training Samples
Supervised Classification
Decision Tree Framework
scientific interpretation
land use and land cover
USGS Thesaurus
marine geology
coastal processes
sea-level change
topography
hazards
USGS Metadata Identifier
USGS:6197cb8dd34eb622f692ee19
None
Northeast US
Maine
New Hampshire
Massachusetts
Rhode Island
Connecticut
New York
New Jersey
Delaware
Maryland
Virginia
St. Croix Island International Historic Site
Acadia National Park
Gateway National Recreation Area
George Washington Birthplace National Monument
Cape Cod National Seashore
None. Please see 'Distribution Info' for details.
Not to be used for navigation. Public domain data from the U.S. Government are freely redistributable with proper metadata and source attribution. Please recognize the U.S. Geological Survey (USGS) as the source of this information. Additionally, there are limitations associated with coastal change hazard assessments. Although these data are published at a resolution of 10 mpp and are considered high resolution, the intended scale for use is around 1:24,000. Please read the associated data release (https://doi.org/10.3133/dr1169) for a list of caveats, applications, and use recommendations for these data.
U.S. Geological Survey
Travis K Sterne
Mailing and Physical
384 Woods Hole Rd
Woods Hole
MA
02543
(508) 548 8700 x2219
tsterne@usgs.gov
https://www.sciencebase.gov/catalog/file/get/6197cb8dd34eb622f692ee19?name=MaxCCL_Graphic.jpg
Outer Cape Cod with Maximum CCL data layer
JPEG
Version 6.2 (Build 9200) ; Esri ArcGIS 10.6.0.8321
Thieler, E.R.
Hammar-Klose, E.S.
1999
National assessment of coastal vulnerability to sea-level rise; U.S. Atlantic Coast
vector digital data
Open-File Report
1999-593
Reston, VA
U.S. Geological Survey
https://doi.org/10.3133/ofr99593
https://pubs.usgs.gov/of/1999/of99-593/
Elizabeth A. Pendleton
Erika E. Lentz
Travis K. Sterne
Rachel E. Henderson
2023
Development and application of a coastal change likelihood assessment for the northeast region, Maine to Virginia
Data Report
1169
Reston, VA
U.S. Geological Survey
Suggested citation: Pendleton, E.A., Lentz, E.E., Sterne, T.K., and Henderson, R.E., 2023, Development and application of a coastal change likelihood assessment for the northeast region, Maine to Virginia: U.S. Geological Survey Data Report 1169, 56 p., https://doi.org/10.3133/dr1169. The CCL data release (https://doi.org/10.5066/P96A2Q5X) is associated with the CCL Data Report (https://doi.org/10.3133/dr1169)
https://doi.org/10.3133/dr1169
https://pubs.er.usgs.gov/publication/dr1169
All data values represent a compilation of coastal hazards likely to be present in the coming decade based on previous empirical research and expert opinion. The final output generated is the expected outcome based on this information.
All data were checked for accuracy during processing. Any inconsistencies in the final data product are artifacts of source data.
CCL is a model for coastal landscapes in the Northeast United States. All output is "clipped" to an elevation domain; this dataset represents areas where coastal change in the coming decade may be greatest. Existing gaps in coverage for this dataset are a result of data gaps in source information (fabric and hazards).
Horizontal coordinate information is referenced to the World Geodetic System of 1984 (WGS 1984) in a Geographic Coordinate System or WGS 1984 Web Mercator (auxiliary sphere) in a Projected Coordinate System. Source data were resampled to 10 mpp for use. There may be resampling errors associated with coarsening (e.g. elevation data were resampled from 1 mpp to 10 mpp) and rectilinear conversion of the finite element climatological wave data. Likewise some data, like NOAA’s ESI and the shoreline change data were rasterized from a source vector, and there can be spatial inconsistencies associated with the rasterization of vector data. The horizontal accuracy of this dataset is assumed to be better than +/- 30 meters , but dynamic coastal areas may experience much higher rates of change during storms, and horizontal offset at the shoreline maybe much higher (+/- 100 meters) in certain areas.
This dataset’s domain is defined by the z-values (elevation) domain of the Fabric dataset (of this publication), and as such has a horizontal positional uncertainty of up to 50 cm along the edge of the domain, which corresponds to +/- 10 meters MHW . However, this dataset has no explicit vertical depth values itself, and therefore there is no vertical position accuracy estimate except along the boundary of this dataset domain.
Travis K. Sterne
Elizabeth A. Pendleton
Erika E. Lentz
Rachel E. Henderson
2023
Coastal Change Likelihood in the U.S. Northeast Region: Maine to Virginia - Perpetual Hazard Compilation
raster digital data
https://doi.org/10.5066/P96A2Q5X
https://www.sciencebase.gov/catalog/item/61783250d34e4c6b7fe2a4a2
Digital and/or Hardcopy
2010
2021
ground condition
PerpetualHazardsCompilation
Perpetual Hazards data
Travis K. Sterne
Elizabeth A. Pendleton
Erika E. Lentz
Rachel E. Henderson
2023
Coastal Change Likelihood in the U.S. Northeast Region: Maine to Virginia - Event Hazard Compilation
raster digital data
https://doi.org/10.5066/P96A2Q5X
https://www.sciencebase.gov/catalog/item/61783250d34e4c6b7fe2a4a2
Digital and/or Hardcopy
2010
2021
ground condition
EventHazardsCompilation
Event Hazards data
Travis K. Sterne
Elizabeth A. Pendleton
Erika E. Lentz
Rachel E. Henderson
2023
Coastal Change Likelihood in the U.S. Northeast Region: Maine to Virginia - Fabric Dataset
raster digital data
data release
DOI:10.5066/P96A2Q5X
Woods Hole Coastal and Marine Science Center, Woods Hole, MA
U.S. Geological Survey, Coastal and Marine Geology Program
https://doi.org/10.5066/P96A2Q5X
https://www.sciencebase.gov/catalog/item/61781f88d34e4c6b7fe2a444
Digital and/or Hardcopy
2010
2021
ground condition
Fabric
Fabric dataset
This step and all the subsequent steps were completed by Elizabeth A. Pendleton or Travis K. Sterne using ESRI ArcGIS Pro geospatial software. Any steps that mention the use of “tools” or “functions” refer to geoprocessing tools utilized in ArcGIS Pro. The steps described in detail below are computed on the domain defined by the fabric, event hazards composite, and perpetual hazards composite datasets, which can be found as complimentary data products in this data release. The data processed and included in this dataset has been clipped or modified to fit within the domain of the Northeast CCL study area. Classification schema in this dataset are defined by possible unique spatial combinations of the seven land cover types defined by the Fabric dataset and Hazard (both Event and Perpetual) datasets. Land cover types include rocky shores, hardened shorelines, developed, forest, marsh, unconsolidated shores, and tidal flats. Event hazards include high tide flooding, storm overwash, and wave power . Perpetual hazards include relative sea-level rise projections for 2030, storm recurrence interval, and short-term shoreline erosion rate. Unique combinations between fabric and hazards datasets can be enumerated to over 10,000 classes, and thus were binned according to user-defined criteria based on previous knowledge of landscape and hazard interaction. The final raster dataset (at end of step 4) presented here is the result of a supervised machine learning (Support Vector Machine (SVM)) landscape classification using training samples created by the user.
2021
Step 1: In preparation for supervised image classification, two sets of training samples were created, one for the dataset with fabric and perpetual hazards and one for the dataset with fabric and event hazards. For the fabric and perpetual hazards dataset, 30 classes that represented combinations of landscapes and hazards were created with a total of 607 samples created within the 30 classes. For the fabric and event hazards dataset, 25 classes that represented combinations of landscapes and hazards were created, and a total of 461 samples were created for the 25 classes. For a detailed explanation of classes, refer to the associated Data Report (Pendleton and others, 2023).
PerpetualHazardsCompilation
EventHazardsCompilation
Fabric
2021
TrainingSamples
Travis K Sterne
U.S. Geological Survey, NORTHEAST REGION
Geographer
mailing address
384 Woods Hole Road
Woods Hole
MA
02543
US
(508) 548 8700 x2219
tsterne@usgs.gov
Step 2: The Image Classification Wizard in ArcPro was used to process the SVM classification. Supervised, pixel-based parameters were applied in the configuration step, and the previously compiled training samples (generated in step 1) were used to train the model and generate two classified datasets, one for Event Hazards and one for Perpetual Hazards. The pixel values of the classified raster that was produced for each Hazard type from the execution of the SVM machine learning step were unique integers that indicated 1) which land cover type from the Fabric dataset would be affected (tens place) and 2) the CCL value which would be added to the existing one included in the Fabric dataset in subsequent steps (ones place). For a list of these values see tables in the complimentary CCL Data Report (Pendleton and others, 2023).
TrainingSamples
2021
EventHazardsOutclass
PerpHazardsOutclass
Elizabeth A. Pendleton
U.S. Geological Survey, NORTHEAST REGION
Geologist
mailing address
384 Woods Hole Road
Woods Hole
MA
02543
US
(508) 548 8700 x2259
ependleton@usgs.gov
Step 3: Each of the classified rasters created in the previous step were combined with the Fabric dataset. Using raster calculator, the estimated impact (0-3) of each respective hazard was added to the change likelihood value (CCL) assigned to the Fabric dataset. The resulting raster dataset ranged in value from 0 to 12, which was then normalized to a maximum value of 10 using raster calculator, where any cells with values greater than 10 were reclassified to a value of 10.
EventHazardsOutclass
PerpHazardsOutclass
Fabric
2021
EventHazardsOutputReclass
PerpHazardsOutputReclass
Step 4: Finally, the two outputs from step 3 were mosaicked to a new raster (Mosaic to New Raster tool) using 'maximum' as the mosaic operator, in order to retain the maximum CCL value from the event and the perpetual raster outputs. This produced the Maximum Coastal Change Likelihood raster layer – the maximum predicted change likelihood between the Event and Perpetual Hazards scenarios created in the previous step.
EventHazardsOutputReclass
PerpHazardsOutputReclass
2021
MaxCCL
Step 5: Raster calculator was used to identify areas where event and perpetual hazards are predicted to have a high (CCL of greater than 7) or low (CCL of 7 or lower) degree of impact. Areas where event hazards resulted in a CCL of greater than or equal to 8, but perpetual hazards resulted in a CCL lower than 8 were assigned a value of 1, areas where both event and perpetual hazards resulted in a CCL of greater than or equal to 6 were assigned a value of 2, areas where both event and perpetual hazard resulted in a CCL of less than 6 were assigned a values of 3, and areas where event hazards resulted in a CCL less than 6, but perpetual resulted in a CCL of greater than or equal to 6 were assigned a value of 4. The resulting geotiff raster included values ranging 1 to 4, each value indicating which type of hazard is most likely to cause significant change to the coastal landscape, if any.
EventHazardsOutputReclass
PerpHazardsOutputReclass
2021
CCLFourSquare
Raster
Grid Cell
128025
118491
1
Mercator_1SP
0.0
0.0
0.0
0.0
0.0
0.0
row and column
10.0
10.0
meters
WGS_1984
WGS 84
6378137.0
298.257223563
USGS_CCL_MaximumCCL_2022.tif
Raster geospatial data file.
Producer defined
OID
Internal object identifier.
Producer defined
Sequential unique whole numbers that are automatically generated.
Value
Unique numeric values contained in each raster cell that represents change likelihood. Definitions for change likelihood can be found in the associated Data Report - Section 1.2.1.
U.S. Geological Survey
0
No CCL value assigned
U.S. Geological Survey
1
Extremely unlikely to change
Producer defined
2
Very unlikely to change
U.S. Geological Survey
3
Unlikely to change
U.S. Geological Survey
4
Somewhat unlikely to change
U.S. Geological Survey
5
Mostly uncertain to slightly unlikely to change
U.S. Geological Survey
6
Mostly uncertain to slightly likely to change
U.S. Geological Survey
7
Somewhat likely to change
U.S. Geological Survey
8
Likely to change
Producer defined
9
Very likely to change
U.S. Geological Survey
10
Extremely likely to change
U.S. Geological Survey
Count
Number of raster cells with this value.
Producer defined
146803.0
365003781.0
U.S. Geological Survey - ScienceBase
mailing and physical address
Denver Federal Center, Building 810, Mail Stop 302
Denver
CO
80225
US
1-888-275-8747
sciencebase@usgs.gov
This dataset contains the raster data layer (.tif) and associated files (.sld, .ovr, .cpg, and .dbf) needed to view and edit the information it contains, as well as the FGDC CSDGM metadata in XML format. The .sld is a Service Layer Definition file used by ScienceBase to display the data, the .ovr file contains the pyramids used by a GIS to display the data at different scales the .cpg file is for charactersets, and the .dbf is a dBASE table file used to store data attributes.
Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data on any other system or for general or scientific purposes, nor shall the act of distribution constitute any such warranty.
GeoTIFF
ESRI ArcGIS Pro v2.6.3
259
https://doi.org/10.5066/P96A2Q5X
https://www.sciencebase.gov/catalog/file/get/6197cb8dd34eb622f692ee19
https://www.sciencebase.gov/catalog/item/6197cb8dd34eb622f692ee19
None
20230228
U.S. Geological Survey
Elizabeth A. Pendleton
Geologist
Mailing and Physical
384 Woods Hole Rd
Woods Hole
MA
02543
(508) 457 2259
ependleton@usgs.gov
Content Standard for Digital Geospatial Metadata, FGDC-STD-001-1998
FGDC-STD-001.1-1998