Abstract
Here, two large data sets, SMAST drop camera survey (242,949 samples, 2003 to 2019) and the US Geological Survey databases (27,784 samples, 1966 to 2011), are combined to derive sea floor surficial substrate probability maps for the Northeastern US continental shelf from Virginia Beach to the Gulf of Maine to 300 m depth (218,571 km2 ). Geostatistical models were used to estimate the probability of five geologic and one biogenic substrate types being present at a 250 m resolution, and the proportional contribution of each substrate type to the seabed composition at a 500 m resolution.This dataset includes:- Six geotiffs that represent the estimated probability of each substrate type at 250 m resolution- Five geotiffs thats represent the proportional contribution of each substrate type to the seabed composition at 500 m resolution- Tabular data for the samples that were used to build the geostatistical models that produced the geotiffs above. The table includes sample latitude, longitude, data source (USGS or SMAST), and presence/absence data at each location for sand, mud, gravel, rock, cobble, and shell debris