Introduction

A point database of land-surface elevation and depth to bedrock spanning New York City and Long Island, New York, was used to interpolate bedrock-surface elevation and overburden thickness for the five boroughs of New York City, which includes Manhattan (New York County), the Bronx (Bronx County), Brooklyn (Kings County), Queens (Queens County), and Staten Island (Richmond County). The interpolated surfaces were used to create digital geospatial data and maps in cooperation with the New York City Department of Design and Construction (NYCDDC) in support of city construction projects, notably those involving construction of geothermal heat pump systems, which require knowledge of bedrock and groundwater depths to determine feasibility (Paino and others, 2012). These bedrock elevation and overburden thickness maps also have application to U.S. Geological Survey (USGS) groundwater studies across the New York metropolitan area as part of a larger investigation on the groundwater sustainability of the Long Island aquifer system (Breault and others, 2021).

Maps presented here supersede previous USGS interpretations of bedrock surface elevation and overburden thickness for Manhattan, Bronx, and parts of Queens (Baskerville, 1994), Staten Island (Soren, 1988), and parts of Long Island (Stumm, 2001; Stumm and others, 2002). Improvements in these maps result from the inclusion of thousands of additional data points, including newly acquired geophysical data, and updated GIS-based interpolation tools and techniques. For the purposes of this study, bedrock is defined as consolidated rock that crops out at the surface or lies beneath soil, artificial fill, postglacial and glacial sediments, and (or) weathered bedrock (saprolite). These unconsolidated deposits that overlie bedrock are collectively referred to as overburden.

Database and Methods

The bedrock elevation database consists of over 14,000 points across Bronx, Kings, New York, Queens, Richmond, Suffolk, and Nassau Counties (DeMott and others, 2023b). Bedrock elevation at each point was adjusted to the North American Vertical Datum of 1988 (NAVD 88), and depth to bedrock was obtained by subtracting the bedrock elevation from the surface elevation at each point. Raster surfaces for bedrock elevation and depth to bedrock were then generated from these data and from additional interpretive sources by using the Spatial Analyst tools available as an ArcGIS Pro (Esri) extension.

Sources of Data

Bedrock elevation was obtained from test boreholes for construction of bridges, tunnels, and buildings; water and monitoring wells; surface outcrops; marine seismic-reflection profiles; and HVSR soundings. Data were compiled from numerous sources, both published and previously unpublished; these sources and specific references are provided in an accompanying USGS data release (DeMott and others, 2023b). For some data points, sources providing the information requested that, for security purposes, the originator of the data be kept confidential; the source for these data points is listed in the data release as “NYC DDC” with an accompanying reference of “Proprietary data.” All data were converted from the reported original elevation datum of the borough to NAVD 88 where necessary. For marine seismic-reflection profiles collected primarily in the East River, the top of bedrock was interpreted based on seismic reflection characteristics in two-way travel time, and individual points from this bedrock horizon were extracted from the seismic profiles. Depth to bedrock was then obtained by converting two-way travel time (an average of 5,000 feet per second for glacial sediments [Stumm and others, 2002] was used and verified at several control points) to depth below sea level, and the resulting depth was then subtracted from the surface elevation to obtain bedrock elevation. Overburden thickness was estimated from the HVSR soundings by using an empirical relation established between measured resonance frequencies and depths to bedrock at nearby wells in Queens, Suffolk, and Nassau Counties (DeMott and others, 2023a). Measurements for 15 wells with known depths to bedrock (fig. 1) were used to develop a power-law regression equation (R² = 0.8373) that relates the resonance frequency (f_r) to the depth of bedrock (Z), in feet, as follows:

Z

= 358.73*

f_r^−1.192

.

(1)

For each site, resonance frequencies from 1 to 3 HVSR measurements were averaged, then subtracted from the surface elevation to obtain bedrock elevation. Borehole, well, and outcrop data were all provided as bedrock elevation data.

Surface elevation for the area was obtained by mosaicking three topobathymetric models: the 2016 USGS Coastal National Elevation Database (CoNED) topobathymetric model: New England; the 2015 USGS CoNED topobathymetric model: New Jersey and Delaware; and the 2017 New York City topobathy light detection and ranging (lidar) digital elevation model (DEM): New York. All three DEMs have an original cell size of 1 meter (m), with vertical accuracy ranging from 3.8 centimeters (cm) (approximately [~] 1.5 inch) to 50 cm (~1.64 foot [ft]). The final DEM was resampled to a cell size of 100 ft (30.48 m) to correspond with the cell size of the bedrock elevation raster. Overburden thickness was determined for each point in the database by extracting the surface elevation from the DEM, then subtracting the bedrock elevation from the surface elevation. An overburden thickness raster was then calculated by subtracting the bedrock-surface elevation raster from the lidar elevation.

Bedrock-Surface Elevation and Overburden Thickness

Bedrock elevation and overburden thickness rasters can be used to examine the nature of the bedrock surface across New York City (figs. 2, 8). Broadly, bedrock elevation tends to be higher, and overburden thinner, in Manhattan and the Bronx. The bedrock elevation surface dips to the southeast, with lower elevation and thicker overburden across Brooklyn and Queens.

Bedrock elevation is highest in north-central Staten Island (maximum of about 360 ft above sea level), in northern Manhattan (265 ft above sea level in Bennett Park), and in the northwestern region of the Bronx, corresponding to the areas where bedrock is also at or near the surface (fig. 2). Bedrock is also relatively shallow (50 ft deep or less) throughout most of the Bronx, Manhattan, and Staten Island, including areas along the East River, where high currents can scour down to bedrock (fig. 8).There is a general spatial trend of lower bedrock elevation and thicker overburden to the southeast and of higher elevation and thinner overburden to the northwest; the lowest bedrock elevation in the five boroughs is on the Rockaway Peninsula in southern Queens (about 1,200 ft below sea level). These spatial trends in elevation mirror trends in the surface elevation.

In the northern boroughs of New York City, bedrock elevation is generally high and overburden thickness is low. In Bronx County, bedrock elevation ranges from ~330 ft below sea level in the Hudson River channel to ~230 ft above sea level in the northwestern region of the county (fig. 3), where an elevational high runs parallel to the Hudson channel. The overburden thickness ranges from 300 ft in the Hudson River to 0 ft in numerous areas where bedrock crops out at the surface, including in the East River, parts of Long Island Sound, and other areas scattered throughout the county (fig. 9). Outside of the deep Hudson River channel fill, overburden is generally thin, with an average thickness of 26 ft. A similar range of elevation and overburden thickness is observed in New York County (Manhattan), where elevation ranges from ~335 ft below sea level in the Hudson River channel to ~260 ft above sea level in the northern arm of the island (fig. 4), where a topographic high runs parallel to the river, extending from Bronx County. Linear, topographic lows are observed along the 125th Street and Dyckman Street faults. Overburden thickness ranges from ~310 ft in the Hudson River channel to 0 ft at bedrock outcrops in the northern arm of the island, particularly along the coasts, and along parts of the East River where rapid currents have scoured the channel down to bedrock. Average overburden thickness is ~64 ft (fig. 10).

On Long Island, bedrock elevation is higher to the northwest, near the East River, and dips lower toward the southeast. In Queens County, bedrock elevation ranges from a high point of ~10 ft above sea level to the northwest and reaches the lowest elevation in the study area at ~1,340 ft below sea level to the southeast and offshore (fig. 5), although the southern offshore area is not well constrained by data. Bedrock outcrops are limited to a few small areas along the northern shoreline and in Long Island Sound and the East River. Overburden thickness increases to the south and eventually reaches a maximum of ~1,280 ft to the south on the Rockaway Peninsula, with an average thickness of ~680 ft across Queens County (fig. 11). Kings County exhibits similar trends, but bedrock elevation is overall lower than in Queens County, with a maximum of ~3 ft below sea level along the northeastern county border and a minimum of ~940 ft below sea level to the southeast in Jamaica Bay (fig. 6). Northwest-southeast-oriented topographic valleys, likely caused by glacial scouring, are visible in the higher elevation region to the northeast (referred to as “Glacial valleys” in fig. 6). Overburden thickness also follows the same trend as for Queens County, with the exception that there are no surface outcrops of bedrock in Kings County. Overburden is thinnest along the northern shoreline and reaches a maximum thickness of ~930 ft to the southeast at Jamaica Bay (fig. 12). Average overburden thickness in Kings County is ~450 ft.

Staten Island (Richmond County) has the highest bedrock elevation in the five boroughs. Bedrock elevation reaches a maximum of ~362 ft above sea level near Todt Hill, in the north-central region of the island (fig. 7). Bedrock elevation is lowest to the southeast, with a minimum elevation of ~680 ft below sea level offshore, in Raritan Bay, although this value is not well constrained by data. Overburden thickness on Staten Island ranges from ~650 ft to 0 ft; overburden is thinnest where serpentinite bedrock crops out in the Todt Hill area and surrounding park lands, as well as in the William T. Davis Wildlife Refuge, along the Palisades Sill, in the west-central part of the island (fig. 13). Overburden becomes thicker in the southern and eastern parts of the island, with the highest values offshore in Raritan Bay.

A comparison of the interpolated surfaces to the original elevation point data inputs was made to assess the accuracy of the surface with respect to the point elevation input. Values for each point were extracted from the bedrock elevation and overburden thickness maps (Extract Multi Values to Points tool in the Spatial Analyst extension, ArcGIS Pro). The original values were then subtracted from the interpolated values to represent the error, and the mean, median, and standard deviation for the interpolation surfaces were calculated. For the bedrock elevation surface, the mean error is −0.97 ft (median error = −0.31 ft, standard deviation = 12.08 ft). For the overburden thickness surface, the mean error is 0.57 ft (median error = 0.10 ft, standard deviation = 13.82 ft).

Limitations

A subset (n=146) of the data points used in the bedrock surface interpolation, originally compiled from Smolensky and others (1990), were used as a proxy for assessment of the uncertainty of the interpolation input data. The subset data were checked against the National Water Information System (NWIS) well database (USGS, 2021), and bedrock elevations were adjusted if necessary and verified based on the original well logs. Differences between the values reported in Smolensky and others (1990) and the verified values are primarily due to the well driller's interpretation of “bedrock.” Some wells have thick layers of saprolite (weathered, noncohesive bedrock) above solid unweathered bedrock; these layers are sometimes referred to as bedrock in drilling reports but do not meet the definition of bedrock used in this study. The median difference between the verified well elevation and the original value was 3 ft (mean difference = 9.7 ft). Of the 146 verified wells, 119 wells had elevations within 10 ft of the value in Smolensky and others (1990).

Although all the points in the dataset could not be independently verified as described above, some data points were adjusted for specific criteria. In some instances, data points obtained from outcrops had bedrock elevations higher than that of the land surface; these data points were adjusted to match the surface elevation from the DEM after examination of the location on satellite imagery to confirm the presence of exposed bedrock. Elevations of other points from borehole or seismic sources that were similarly inconsistent with DEM surface elevations were evaluated independently and either were adjusted to a new elevation or location on the basis of available information (n=311) or were deleted from the database (n=152); duplicate points were also deleted. These differences in elevation are likely due to initial inaccuracies in measurement of the land surface, higher accuracy of the modern DEM than elevation measurements at the time of data collection, and changing land-surface elevation over time due to construction and (or) erosion.

Additional checking and postprocessing of the interpolated bedrock elevation surface was done because of spatial data limitations. In areas where bedrock is probably near the surface (less than 5 ft below land surface) and data points are sparse or absent, interpolated results indicated that the bedrock surface was at a higher elevation than the DEM land surface. These data gaps limit the control on the interpolation performed by the Topo to Raster tool. Although it is possible to set minimum and maximum constraints in the tool, it is not possible to limit the interpolation output so it should always be at or lower than land-surface elevation. Locations where these discrepancies were observed were adjusted through use of the Raster Calculator tool. Pixels with values higher than surface elevation were identified by using Raster Functions (Analysis menu, ArcGIS Pro), first by subtracting the bedrock elevation raster from the surface elevation raster (Minus function). Cells with negative values, indicating bedrock surface higher than land surface, were identified by using the Less Than function. These cells were then set to null (Set Null function), then recalculated to be equal to the surface elevation by mosaicking the surface elevation raster to the bedrock surface (Mosaic to New Raster tool in the Data Management toolbox, ArcGIS Pro), since bedrock is likely to be less than 25 ft below the surface in these areas, but true bedrock elevation is unknown.

Summary

Bedrock elevation and overburden thickness maps presented in this report were created by the U.S. Geological Survey, in cooperation with the New York City Department of Design and Construction, from an extensive database of historical construction test borings, recent exploration drilling boreholes, U.S. Geological Survey observation wells, outcrops, and seismic measurements, combined with published contour data and interpretive features. These map products can assist with future evaluations of hydrogeologic conditions in the city, particularly in evaluating prospective geothermal heat pump installations.

The study area includes a varied topography of exposed bedrock highlands in northern Manhattan, central Staten Island, and parts of the Bronx. On Long Island, bedrock crops out in a few isolated areas in northwestern Queens County. Most of the study area’s bedrock is covered in unconsolidated sedimentary overburden of Pleistocene and (or) Cretaceous age with a thickness ranging from a few feet to over a thousand feet.

The bedrock underlying the study area was previously mapped as a structurally complex assemblage of high-grade metamorphic rocks, which includes gneiss, schist, amphibolite, marble, granite, and serpentinite. These rocks include the Fordham Gneiss and correlative Queens tunnel complex, Manhattan Schist, Hartland Formation, Inwood Marble, Ravenswood Granodiorite, and the Walloomsac Formation.

The bedrock underlying the study area ranges from easily weathered (Inwood Marble) to very resistant to weathering (Manhattan Schist, Fordham Gneiss). This differential susceptibility to erosion along with the numerous north-northwest-trending faults is thought to control the shape of the bedrock surface underlying the study area. Glacial scouring of the bedrock during the Pleistocene is the most recent control on the bedrock surfaces mapped.

The bedrock elevation database consists of over 14,000 points across Bronx, Kings, New York, Queens, Richmond, Suffolk, and Nassau Counties. Bedrock elevation at each point was adjusted to the North American Vertical Datum of 1988, and depth to bedrock was obtained by subtracting the bedrock elevation from the surface elevation at each point. Rasters for bedrock elevation and depth to bedrock were then generated from these data and additional interpretive sources by using ArcGIS Pro spatial analysis tools.

Bedrock elevation was obtained from test boreholes for construction of bridges, tunnels, and buildings; water and monitoring wells; surface outcrops; marine seismic reflection profiles; and horizontal-to-vertical spectral ratio soundings.

The bedrock elevation data were interpolated by using an iterative finite difference interpolation technique developed for elevation surfaces in ArcGIS Pro. The interpolation tool allows for multiple types of input data and additional geological constraints to customize sinks and drainages to create a hydrologically correct digital elevation model.

Surface elevation for the area was obtained by merging three topobathymetric models. Depth to bedrock was acquired for each point in the database by extracting the surface elevation from the digital elevation model, then subtracting the bedrock elevation from the surface elevation. The overburden thickness raster was then calculated by subtracting the bedrock-surface elevation raster from the land-surface lidar elevation. The output rasters were interpolated to a cell size of 100 feet, based on spatial resolution of the dataset. Rasters were generated for the full dataset, then clipped to each individual borough. Contours at 100-foot intervals were generated from the output rasters and smoothed for the final geospatial and map products.

Bedrock elevation is highest in north-central Staten Island (maximum of about 360 feet above sea level), northern Manhattan, and the northwestern region of the Bronx. There is a general spatial trend of lower bedrock elevation to the southeast and higher elevation to the northwest, with the lowest bedrock elevation in the five boroughs in the Rockaway Peninsula in southern Queens (about 1,200 feet below sea level).