Kathryn E.L. Smith Davina L. Passeri Nathaniel G. Plant 20170421 Tabular digital data U.S. Geological Survey Data Release doi:10.5066/F7CZ35BC St. Petersburg, FL U.S. Geological Survey https://doi.org/10.5066/F7CZ35BC The Barrier Island and Estuarine Wetland Physical Change Assessment was created to calibrate and test probability models of barrier island estuarine shoreline (backshore) and beach sandline change for study areas in Virginia, Maryland, and New Jersey. The models examined the influence of hydrologic and physical variables related to long-term and storm-derived overwash and back-barrier shoreline change. Input variables were constructed into a Bayesian Network (BN) using Netica, a computer program created by NORSYS Software Corporation that allows users to work with belief networks and influence diagrams. Each model is tested on its ability to predict changes in long-term and event-driven (i.e., Hurricane Sandy-induced) backshore and sandline change based on learned correlations from the input variables across the domain. Using the input hydrodynamic and geomorphic data, the BN is constrained to produce a prediction of an updated conditional probability of backshore or sandline change at each location. To evaluate the ability of the BN to reproduce the observations used to train the model, the skill, log likelihood ratio and probability predictions were utilized. These data are the probability and skill metrics for the long-term estuarine back-barrier shoreline change model. The estuarine shoreline change model will be used to examine how geophysical and hydrodynamic variables influence both long-term and storm-driven dynamics of barrier islands overwash and erosion. This information is critical to understanding and predicting the implications of future sea-level and storm impacts on barrier islands and estuaries. A free version of the Netica application is available for download at http://www.norsys.com/download.html. 20160606 dataset creation Complete None planned -75.382739 -73.974687 40.479022 37.862809 USGS Metadata Identifier USGS:391ab886-b013-45cd-8da1-f70eb551d5ed ISO 19115 Topic Category geoscientificInformation oceans environment None Long-term shoreline change Storm-driven shoreline change Coastal Bayesian models Barrier Islands Storms Hurricanes USGS Thesaurus geomorphology ecology geology None New Jersey USA Mid-Atlantic Ocean VA Assateague Island Maryland Virginia MD NJ None The U.S. Geological Survey requests to be acknowledged as originator of the data in future products or derivative research. U.S. Geological Survey Coastal and Marine Science Center Kathryn E.L. Smith Ecologist mailing and physical

600 4th Street South

St. Petersburg Florida 33701 US (727) 502-8073 (727) 502-8001 kelsmith@usgs.gov Monday through Friday, 9:00 a.m. to 5:00 p.m., Eastern Standard Time This project was funded by the USGS Coastal and Marine Geology Program. Acknowledgment of the U.S. Geological Survey, St. Petersburg Coastal and Marine Science Center, as a data source would be appreciated in products developed from these data, and such acknowledgment as is standard for citation and legal practices. Sharing of new data layers developed directly from these data would also be appreciated by the U.S. Geological Survey staff. Users should be aware that comparisons with other datasets for the same area from other time periods may be inaccurate due to inconsistencies resulting from changes in photointerpretation, mapping conventions, and digital processes over time. These data are not legal documents and are not to be used as such. Microsoft Windows 7 Version 6.1 (Build 7601) Service Pack 1; Microsoft Excel 2010 Version 14.0.7173.5000 (32-bit) Norsys 20000101 5.12 for MS Windows (2000 to 7) software http://www.norsys.com/netica.html MathWorks 20030215 R2013a Version 8.1.0.604 (64-bit) software https://www.mathworks.com/ Plant, N. G., E. R. Thieler, and D. L. Passeri 20160502 Earth's Future Volume 4, Issue 5, pages 143-158 Hoboken, NJ Wiley Periodicals Inc. https://doi.org/10.1002/2015EF000331 Data values were checked for logical consistency. Predicted probabilities range from 0 to 1, where 0 is least likely and 1 is most likely to fall within the identified bin. The log-transform of the likelihood ratio can be negative or positive; however, positive values show an improvement over the prior. The Bayesian mean and predicted most likely value should fall within a reasonable estimate of backshore change parameters of the case (calibration) dataset. Data are complete. In some cases, all modeled variables were not available for each observation; however, Bayesian Networks can learn from a file of cases, relying on complete observations to create a maximum liklihood Bayes net. For this model, the expectation-maximization (EM) learning algorithm in Netica was used. At locations where some or all of the data are missing (e.g., due to differences in the spatial resolution of the datasets) predictions can still be made since missing data are taken into account with prediction uncertainty. For example, at worst, with no specific inputs, the BN prediction returns the prior distribution of the output variable, which reflects the spatial variability of that variable over the entire study area. Terrano, J.F. and Smith, K.E.L. 2015 tabular digital data http://coastal.er.usgs.gov/data-release/doi-F71Z42HN/ online 20150101 publication calibration data This data served as the calibration data for the Bayes Net. The Bayesian network (BN) for back-barrier shoreline (backshore) change was constructed in Netica using the following variables: barrier island height, barrier island width, estuary depth, estuary width, maximum storm-driven wave height, long-term and storm-driven (Hurricane Sandy) ocean shoreline change, and long-term and storm-driven backshore change. Calibration (case) data were generated using a transect method in Esri ArcGIS (Version 10.3.1.4959) from pre-Hurricane Sandy lidar, imagery, and hydrodynamic models. The BN was constrained using the case data to produce a prediction of a conditional probability of back-barrier shoreline (backshore) change. These case data and methods were previously published in Terrano, J.F. and Smith, K.E.L., 2015, Estuarine Shoreline and Barrier-Island Sandline Change Assessment: U.S. Geological Survey data release, https://doi.org/10.5066/F71Z42HN. The BN predictions are produced at the same spatial scale as the input variables. At locations where some or all of the data are missing (e.g., due to differences in the spatial resolution of the datasets) predictions can still be made since missing data are taken into account with prediction uncertainty. For example, at worst, with no specific inputs, the BN prediction returns the prior distribution of the output variable, which reflects the spatial variability of that variable over the entire study area. To evaluate the ability of the BN to reproduce the observations used to train the model, the skill, and log likelihood ratio are utilized. The BN was loaded into MatLab and hindcast to predict the best possible skill for back-barrier shoreline or beach sandline change. The equation used to compute the skill and likelihood ratios are given in equations 4 and 5 in Plant and others (2016). Predicted probabilities and skill metrics were output from Matlab into tabular form to produce this summary dataset. calibration data 20160101 Added keywords section with USGS persistent identifier as theme keyword. 20201013 U.S. Geological Survey VeeAnn A. Cross Marine Geologist Mailing and Physical

384 Woods Hole Road

Woods Hole MA 02543-1598 508-548-8700 x2251 508-457-2310 vatnipp@usgs.gov Point Entity point 3458 0.0197519519 0.0253592358 Decimal degrees D North American 1983 GRS 1980 6378137.0 298.257222101 ASI_Predict_LT_Backshore, NJ_Predict_LT_Backshore Model prediction probabilities and skill metrics in CSV format USGS objectid Internal feature number. Esri Sequential unique whole numbers that are automatically generated. X_long Longitude location in geographic coordinate system (World Geodetic System 1984) decimal degrees USGS -75.3827 -73.9769 decimal degrees Y_lat Latitude location in geographic coordinate system (World Geodetic System 1984) decimal degrees USGS 37.86545 40.47719 decimal degrees P-32.0 Predicted probability of the net long-term backshore shoreline change falling between (-32.0) and (-10 m) USGS 0.03 0.29 P-10.0 Predicted probability of the net long-term backshore shoreline change falling between (-10.0 m) and (-0.5 m) USGS 0.08 0.52 P-0.5 Predicted probability of the net long-term backshore shoreline change falling between (-0.5 m) and (+0.5 m) USGS 0.12 0.54 P0.5 Predicted probability of the net long-term backshore shoreline change falling between (0.5 m) and (+10 m) USGS 0.02 0.62 P10.0 Predicted probability of the net long-term backshore shoreline change falling between (+10 m) and (+308 m) USGS 0.01 0.15 LR Log-transform of the ratio of predicted probability in the bin corresponding to the observed net backshore change to the prior probability in that bin, where LR>0 means the prediction is an improvement over the prior USGS -0.826 0.556 mean Bayesian-mean value (sum of p(x)*x over all bins) of the predicted backshore shoreline change (meters/year) USGS -5.7 1.8 mostProb Predicted-most-likely value (center of bin with highest probability) of the predicted backshore shoreline change (meters/year) USGS -21 5.3 PmostProb Probability value at the predicted-most likely value USGS 0.3 0.6 U.S. Geological Survey Kathryn E.L. Smith mailing and physical

600 4th Street South

St. Petersburg Florida 33701 US (727) 502-8073 (727) 502-8001 kelsmith@usgs.gov Monday through Friday, 9:00 a.m. to 5:00 p.m., Eastern Standard Time All of this report is available online. ASI_Predict_LT_Backshore.csv, NJ_Predict_LT_Backshore.csv This digital publication was prepared by an agency of the United States Government. Although these data have been processed successfully on a computer system at the U.S. Geological Survey, no warranty expressed or implied is made regarding the display or utility of the data on any other system, nor shall the act of distribution imply any such warranty. The U.S. Geological Survey shall not be held liable for improper or incorrect use of the data described and (or) contained herein. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. Comma Separated Values (CSV) CSV files containing prediction probabilities and skill metrics for Bayes Net models. http://coastal.er.usgs.gov/data-release/doi-F7CZ35BC/data/Predict_LT_Backshore.zip None 20201013 U.S. Geological Survey Kathryn E.L. Smith mailing and physical

600 4th Street South

St. Petersburg Florida 33701 US (727) 502-8073 (727) 502-8001 kelsmith@usgs.gov Monday through Friday, 9:00 a.m. to 5:00 p.m., Eastern Standard Time Content Standard for Digital Geospatial Metadata FGDC-STD-001-1998 local time Public domain data from the U.S. Government are freely redistributable with proper metadata and source attribution. The U.S. Geological Survey requests to be acknowledged as originator of the data in future products or derivative research.