Topography provides information about the structural controls of the Great Basin and therefore information that may be used to identify favorable structural settings for geothermal systems. Specifically, local relative topography gives information about locations of faults and fault intersections relative to mountains, valleys, or at the transitions between. As part of U.S. Geological Survey efforts to engineer features that are useful for predicting geothermal resources, we construct a detrended elevation map that emphasizes local relative topography and highlights features that geologists use for identifying geothermal systems (i.e., providing machine learning algorithms with features that may improve predictive skill by emphasizing the information used by geologists). Herein, we provide the trend and local relative elevation maps documented in DeAngelo and others (2023), describing the process of removal of the regional trend and the resulting detrended elevation maps that emphasize basin-and-range scale structural features. Regional elevation trends were estimated using a local linear regression and subtracted from a 30-m digital elevation model (DEM) of topography to create the detrended elevation (i.e., local relative topography) map; therefore one could add the detrended surface to the corresponding trend surface to construct the original DEM. In an effort to optimize the detrended surface, alternate versions were produced with different rates of smoothness resulting in three detrended elevation maps. The resulting detrended elevation maps emphasize geologic structure and relative displacement, and these products may be useful for other geologic research including mineral exploration, hydrologic research, and defining geologic provinces. References DeAngelo, J., Burns, E.R., Lindsey, C.R., and Mordensky, S.P., (2023), Detrending Great Basin elevation to identify structural patterns for identifying geothermal favorability, Geothermal Rising Conference Transactions, 47, Reno, Nevada, October 1-5, 2023.

Dunes with a high relative topography can often be easily distinguished in high-resolution lidar-based digital elevation models (DEMs). Thus, researchers have begun using relative topography metrics, such as the topographic position index (TPI; Weiss, 2001), to identify ridges and upper slopes for extracting dunes from lidar-based DEMs (Wernette et al., 2016; Halls et al. 2018). DEMs are often used for automated delineations of intertidal and supratidal habitats in coastal applications despite issues related to vertical uncertainty. However, the level of vertical uncertainty from data collected with conventional aerial topographic lidar systems has been found to be as high as 60 cm in densely vegetated emergent wetlands throughout the United States (Medeiros et al., 2015; Buffington et al., 2016; Enwright et al., 2018). This uncertainty can also impact elevations in other habitats such as dunes due to vegetation cover and slope (Su and Bork, 2006). Another challenge when mapping geomorphology-based habitats (e.g., dune, beach, intertidal marsh, forest) on dynamic barrier islands is the need for standardized methods that are efficient and repeatable. In response, we developed an approach that builds on recent efforts using relative topography to identify ridges and upper slopes for dune delineation (Wernette et al. 2016; Halls et al. 2018) by also applying Monte Carlo simulations to treat elevation uncertainty in coastal settings when extracting elevation-dependent habitats from a DEM (Liu et al. 2007; Enwright et al. 2018) for a case study on Dauphin Island, Alabama. Beyond just the application of uncertainty, we refined ridges and upper slopes extracted from a DEM by removing small noisy polygons and using manual refinement. This data release contains each of these iterations to show the importance of uncertainty analyses and manual refinement when using automated extraction of elevation-dependent habitats from a DEM. This data release includes a TPI directory, which contains four polygon shapefiles that represent each step in the TPI-based dune delineation process, which includes: 1) step1_raw_ridges_upper_slopes.shp; 2) step2_refinement_extreme_water_level.shp; 3) step3_refinement_via_noise removal.shp; and 4) step4_final_refinement_from_visual_inspection.shp. Since this a step-wise process, each step includes the prior steps. A second component of this data release is a raster named “Prob_Abv_Storm” that estimates the probability of a pixel being above the extreme water level with a 10-percent annual exceedance probability for National Oceanic and Atmospheric Administration’s Dauphin Island tide gauge (station ID: 8735180).

A bare-earth, hydro-flattened, digital-elevation surface model derived from 2010 Light Detection and Ranging (LiDAR) data. Surface models are raster representations derived by interpolating the LiDAR point data to produce a seamless gridded elevation data set. A Digital Elevation Model (DEM) is a surface model generated from the LiDAR returns that correspond to the ground with all buildings, trees and other above ground features removed. The cell values represent the elevation of the ground relative to sea level. The DEM was generated by interpolating the LiDAR ground points to create a 1 foot resolution seamless surface. Cell values correspond to the ground elevation value (feet) above sea level. A proprietary approach to surface model generation was developed that reduced spurious elevation values in areas where there were no LiDAR returns, primarily beneath buildings and over water. This was combined with a detailed manual QA/QC process, with emphasis on accurate representation of docks and bare-earth within 2000ft of the water bodies surrounding each of the five boroughs.

NYC 1foot Digital Elevation Model: A bare-earth, hydro-flattened, digital-elevation surface model derived from 2010 Light Detection and Ranging (LiDAR) data. Surface models are raster representations derived by interpolating the LiDAR point data to produce a seamless gridded elevation data set. A Digital Elevation Model (DEM) is a surface model generated from the LiDAR returns that correspond to the ground with all buildings, trees and other above ground features removed. The cell values represent the elevation of the ground relative to sea level. The DEM was generated by interpolating the LiDAR ground points to create a 1 foot resolution seamless surface. Cell values correspond to the ground elevation value (feet) above sea level. A proprietary approach to surface model generation was developed that reduced spurious elevation values in areas where there were no LiDAR returns, primarily beneath buildings and over water. This was combined with a detailed manual QA/QC process, with emphasis on accurate representation of docks and bare-earth within 2000ft of the water bodies surrounding each of the five boroughs. Please see the following link for additional documentation- https://github.com/CityOfNewYork/nyc-geo-metadata/blob/master/Metadata/Metadata_DigitalElevationModel.md

Topographic Position Index (TPI) is a topographic position classification identifying upper, middle and lower parts of the landscape. This dataset includes a mask that identifies where topographic position cannot be reliably derived in low relief areas.

The TPI product was derived from Smoothed Digital Elevation Model (DEM-S; ANZCW0703014016), which was derived from the 1 arc-second resolution SRTM data acquired by NASA in February 2000. A masked version of the TPI product was derived using the slope relief classification product.

The TPI data are available at 1 arc-second and 3 arc-second resolution.

The 3 arc-second resolution dataset was generated from the 1 arc-second TPI product and masked by the 3” water and ocean mask datasets.

Lineage: Source data 1.\t1 arc-second SRTM-derived Smoothed Digital Elevation Model (DEM-S; ANZCW0703014016). 2.\t1 arc-second slope relief product 3.\t3 arc-second resolution SRTM water body and ocean mask datasets.

Topographic position index calculation TPI is a measure of topographic position, classified into three classes corresponding to upper slopes, mid-slopes and lower slopes. The method follows that of the "Drainage Channels Class" section of Warner, Cress and Sayre (2008) which is based on the TPI method of Jenness (2006) and Weiss (2001).

The TPI classification uses relative elevation as a fraction of local relief; where the relative elevation is high compared to the local relief the class is upper slope, and where the relative elevation is low compared to local relief the class is lower slope. Intermediate values are classified as mid-slopes. This use of residuals compared to a smoothed elevation model to produce relative elevations is similar to the method described by McRae (1992).

Relative elevation is the difference between local (cell) elevation and the mean elevation over a 300 m radius circle (approximately: the calculation actually uses 10 grid cells at 1 arc-second resolution). Local relief is calculated as the standard deviation of elevation over the same circular region. The classification is:

TPI = \t1 if relative_elevation < -0.5 * local relief (lower slopes) \t3 if relative_elevation > 0.5 * local relief (upper slopes) 2 otherwise (mid slopes)

In relatively flat areas the finite accuracy of a DEM limits its ability to discriminate topographic position. The mask included with the TPI layer identifies areas that are too flat to reliably identify upper, middle and lower landscape positions. It is based on the 'Slope-Relief' classification and the TPI mask has values of 1 where there is sufficient relief for TPI to be meaningful and 0 where TPI should not be used.

The TPI calculation was performed on 1° x 1° tiles, with overlaps to ensure correct values at tile edges.

The 3” arc-resolution version was generated from the 1” TPI class and mask products. This was done by aggregating the 1” data over a 3 x 3 grid cell window and taking the mean of the nine values that contributed to each 3” output grid cell. The result was then converted to integer format, avoiding truncation errors and ensuring that (for example) values between 1.5 and 2 were assigned to class 2, and values between 2.5 and 3 were assigned to class 3. The 3” TPI and TPI mask data were then masked using the SRTM 3” ocean and water body datasets.

The data of 30 m elevation difference distribution for the Himalayas, within the scope of the highest elevation value refers to the Himalayas and the lowest elevation value difference.Relative elevation difference is to describe a regional terrain characteristics of macroscopic indicators, to quantitative description of morphology, geomorphic type divided into important indicators.Can use the focus of the field in GIS software to analyze statistics to calculate the relative relief, including statistical type selection RANGE - calculation neighborhood pixels, in the RANGE (the difference between the maximum and minimum), again through the GIS software further processing.Grid size using the same as the 30 m DEM grid size.The data of 30 meters from the Himalayas DEM data is obtained by ArcGIS software processing.

In this joint demonstration project for the Tampa Bay region, NOAA's National Ocean Service (NOS) and the U.S. Geological Survey (USGS) have merged NOAA bathymetric and USGS topographic data sets into a hybrid digital elevation model (DEM) with all data initially referenced to the ellipsoid, but transformable to any of 28 orthometric, 3-D, or tidal datums.A seamless bathymetric/topographic digital elevation model (DEM) was developed by merging the "best available" bathymetric data from NOAA and topographic data for USGS. Each of the datasets was initially processed independently to apply the "best available" criteria to select the data to be merged. Prior to merging, the selected data were transformed to a common reference coordinate system, both horizontally and vertically.The selected topography points within the shoreline buffer zone and the bathymetry points were gridded to produce a raster surface model with a 1-arc-second (30-meter) grid spacing to match the resolution of NED. The points were input to an implementation of the ANUDEM thin plate spline interpolation algorithm, which is optimized for generation of topographic surfaces. The bathymetry points could have been gridded independently of the topographic data, but the shoreline zone land elevations were included in the interpolation to ensure a better match of the bathymetric and topographic surfaces for the subsequent mosaicing step. To avoid introduction of any interpolation edge effects into the merged elevation model, the output grid from the interpolation was clipped to include only land elevations within 300 meters of the shoreline.The final processing step involved the mosaicing of the bathymetry grid and the NED elevation grid. The values in the 300-meter overlap area were blended by weighted averaging, where the weights for each grid are determined on a cell-by-cell basis according to the cell's proximity to the edges of the overlap area. The resulting final merged product is a seamless bathymetric/topographic model covering the Tampa Bay region at a grid spacing of 1-arc-second (30-meter). The vertical coordinates represent elevation in decimal meters relative to the GRS80 ellipsoid, and the horizontal coordinates are decimal degrees of latitude and longitude referenced to the NAD83 datum.This dataset is intended for geospatial applications that require seamless land elevation and water depth information in coastal environments.

This layer indicates the relative elevations of transportation nodes. It outlines relationships between street segments on different planes via node-and-segment connectivity (e.g. 0 for surface, -1 for an underpass, and 1 for an overpass). Relative Elevation Nodes (REN) are a work-around for inherent limitations when navigating through three-dimensional street networks represented on the two-dimensional plane (i.e. computer screen). REN help to establish routing options and are particularly useful where complex overpass/underpass structures exist because they indicate street segments as being on different road-levels; where on the computer screen all streets appear to be connected.

Additional tables and supporting documentation are available in the Data Dictionary and User Manual.

This dataset contains four alternative digital elevation models (DEMs) at 1 m resolution and model performance statistical metrics for the Global Change Research Wetland (GCReW) site on the Rhode River, a tributary of the Chesapeake Bay in Maryland, USA, for the year 2016. Three DEMs were created by using different strategies for correcting positive biases in Light Detection and Ranging (LiDAR)-based DEMs that are common in tidal wetlands. These included (1) applying a single average offset based on a literature review, (2) using the LiDAR Elevation Correction with NDVI (LEAN)-method, and (3) applying plant community-specific offsets using a local vegetation cover map. Existing LiDAR data at 1 m resolution collected in 2011 was the basis for these DEMs. The fourth DEM was created by using Empirical Bayesian Kriging to extrapolate between measured ground points. The elevation is provided in meters relative to the North American Vertical Datum of 1988 (NAVD 88). To calibrate the four approaches, the elevation of the entire marsh complex was surveyed at 20 m x 20 m resolution to document the distribution of elevation relative to tidal datums from a single year. Two Trimble R8 real-time kinematic (RTK) GPS receivers were used to survey 525 points over the complex from July 26, 2016, to August 15, 2016. Relative plant cover was also documented. Tidal datums were calculated from the nearby Annapolis, MD tidal gauge located 13 km from GCReW.

Elevation Relative to Sea Level measured via Uncategorized in m. Part of dataset Seabed2030 Ice Surface Elevation, 2022

Culminating more than four years of processing data, NASA and the National Geospatial-Intelligence Agency (NGA) have completed Earth's most extensive global topographic map. The mission is a collaboration among NASA, NGA, and the German and Italian space agencies. For 11 days in February 2000, the space shuttle Endeavour conducted the Shuttle Radar Topography Mission (SRTM) using C-Band and X-Band interferometric synthetic aperture radars to acquire topographic data over 80% of the Earth's land mass, creating the first-ever near-global data set of land elevations. This data was used to produce topographic maps (digital elevation maps) 30 times as precise as the best global maps used today. The SRTM system gathered data at the rate of 40,000 per minute over land. They reveal for the first time large, detailed swaths of Earth's topography previously obscured by persistent cloudiness. The data will benefit scientists, engineers, government agencies and the public with an ever-growing array of uses. The SRTM radar system mapped Earth from 56 degrees south to 60 degrees north of the equator. The resolution of the publicly available data is three arc-seconds (1/1,200th of a degree of latitude and longitude, about 295 feet, at Earth's equator). The final data release covers Australia and New Zealand in unprecedented uniform detail. It also covers more than 1,000 islands comprising much of Polynesia and Melanesia in the South Pacific, as well as islands in the South Indian and Atlantic oceans. SRTM data are being used for applications ranging from land use planning to "virtual" Earth exploration. Currently, the mission's homepage "http://www.jpl.nasa.gov/srtm" provides direct access to recently obtained earth images. The Shuttle Radar Topography Mission C-band data for North America and South America are available to the public. A list of complete public data set is available at "http://www2.jpl.nasa.gov/srtm/dataprod.htm" The data specifications are within the following parameters: 30-meter X 30-meter spatial sampling with 16 meter absolute vertical height accuracy, 10-meter relative vertical height accuracy, and 20-meter absolute horizontal circular accuracy. From the JPL Mission Products Summary, "http://www.jpl.nasa.gov/srtm/dataprelimdescriptions.html". The primary products of the SRTM mission are the digital elevation maps of most of the Earth's surface. Visualized images of these maps are available for viewing online. Below you will find descriptions of the types of images that are being generated:

• Radar Image
• Radar Image with Color as Height
• Radar Image with Color Wrapped Fringes
  -Shaded Relief
• Perspective View with B/W Radar Image Overlaid
• Perspective View with Radar Image Overlaid, Color as Height
• Perspective View of Shaded Relief
• Perspective View with Landsat or other Image Overlaid
• Contour Map - B/W with Contour Lines
• Stereo Pair
• Anaglypgh

The SRTM radar contained two types of antenna panels, C-band and X-band. The near-global topographic maps of Earth called Digital Elevation Models (DEMs) are made from the C-band radar data. These data were processed at the Jet Propulsion Laboratory and are being distributed through the United States Geological Survey's EROS Data Center. Data from the X-band radar are used to create slightly higher resolution DEMs but without the global coverage of the C-band radar. The SRTM X-band radar data are being processed and distributed by the German Aerospace Center, DLR.

Note: Geoscience Australia no longer supports users' external hard drives. The data can either be downloaded from the ELVIS Portal or from the Related links. The 1 second Shuttle Radar Topography Mission (SRTM) Digital Elevation Models Version 1.0 package comprises three surface models: the Digital Elevation Model (DEM), the Smoothed Digital Elevation Model (DEM-S) and the Hydrologically Enforced Digital Elevation Model (DEM-H). The DEMs were derived from the SRTM data acquired by NASA in February 2000 and were publicly released under Creative Commons licensing from November 2011 in ESRI Grid format.

DEM represents ground surface topography, with vegetation features removed using an automatic process supported by several vegetation maps. This provides substantial improvements in the quality and consistency of the data relative to the original SRTM data, but is not free from artefacts. Man-made structures such as urban areas and power line towers have not been treated. The removal of vegetation effects has produced satisfactory results over most of the continent and areas with defects identified in supplementary layers distributed with the data, and described in the User Guide.

DEM-S represents ground surface topography, excluding vegetation features, and has been smoothed to reduce noise and improve the representation of surface shape. An adaptive smoothing process applied more smoothing in flatter areas than hilly areas, and more smoothing in noisier areas than in less noisy areas. This DEM-S supports calculation of local terrain shape attributes such as slope, aspect and curvature that could not be reliably derived from the unsmoothed 1 second DEM because of noise.

DEM-H is a hydrologically enforced version of the smoothed DEM-S. The DEM-H captures flow paths based on SRTM elevations and mapped stream lines, and supports delineation of catchments and related hydrological attributes. The dataset was derived from the 1 second smoothed Digital Elevation Model (DEM-S) by enforcing hydrological connectivity with the ANUDEM software, using selected AusHydro V1.6 (February 2010) 1:250,000 scale watercourse lines and lines derived from DEM-S to define the watercourses. The drainage enforcement has produced a consistent representation of hydrological connectivity with some elevation artefacts resulting from the drainage enforcement.

Further information can be found in the supplementary layers supplied with the data and in the User Guide.

This dataset provides maps of the elevation of coastal wetlands relative to tidal ranges for the conterminous United States (CONUS) at 30 m resolution for 2010. It also includes maps of tidal amplitude, relative sea-level rise for the period 1983-2001, and maps for coastal lands and low marsh areas based on the probability of being below the mean higher high tide water line for spring tides (MHHWS). Uncertainty layers for elevation maps are also provided.

This data provides an illustration of the height above the nearest stream approach to flood inundation mapping based on the TauDEM vertical distance to stream function. This example uses a 10 m resolution National Elevation dataset for Onion Creek in Texas. Height above the nearest stream may be thought of as a “relative elevation function” which measures for every DEM cell in the landscape the difference in elevation between that cell and the cell to which it flows on the stream channel. This is like a “water depth” or “stage height” function defined using terrain analysis continuously across the landscape. This relative elevation function, combined with a depth in each stream reach provide a simplified terrain based approach to flood inundation mapping premised on the following:

1. Each reach has a water depth hw, from a hydraulic model such as SPRNT or RAPID.
2. Each reach has an ID
3. Each grid cell has the ID of the reach it connects to (catchment grid) and the height above the stream hs
4. Flood extent is “rapidly” mapped as If(hw(id) > hs(id)) Inundation depth = hw(id) - hs(id) Else Inundation depth = 0

The data here can also be used to compute reach averaged hydraulic properties as follows 1. For each reach the stream network file gives reach length L. 2. For a series of water depths using the height above nearest stream intersected with catchment raster the innundation water volume V, surface area As and bed area Ab are obtained. 3. Reach average properties are then computed as Cross section Area A = V/L Wetted perimeter P = Ab/L Top width = As/L Hydraulic Radius = A/P

This approach is a simplification over finer scale hydraulics, and the inaccuracy due to introduction of this simplification still needs evaluation. This approach is also dependent on how well the DEM represents the channel and flooded area. This is expected to improve as we get better LIDAR DEMs and develop better ways to hydrologically condition DEMs that do not involve pit filling.

Surface elevation table (SET) measurements from 26 SETs at 9 marsh sites in the Plum Island Sound Long-Term Ecological Research Site in the Great Marsh, Massachusetts. SET measurements are useful for determining the relative elevation change of marsh sediments. Precise measurements of sediment elevation in marshes is useful for determining rates of elevation change in response to changes in sea level.

The U.S. Geological Survey (USGS) St. Petersburg Coastal and Marine Science Center conducted research to quantify bathymetric changes along the Florida Reef Tract (FRT) from Miami to Boca Chica Key, Florida. Changes in seafloor elevation were calculated from the 1930s to 2016 using digitized hydrographic sheet sounding data and light detection and ranging (lidar)-derived digital elevation models (DEMs) acquired by the National Oceanic and Atmospheric Administration (NOAA) in 2016 and 2017. Most of the elevation data from the 2016/2017 time period was collected during 2016, and, as an abbreviated naming convention, this time period was referred to as 2016. An elevation change analysis between the 1930s and 2016 data was performed to quantify and map historical impacts to seafloor elevation and to determine elevation-change statistics for 15 habitat types found within the study area along the FRT. Annual elevation-change rates were calculated for each elevation-change data point. Seafloor elevation-change along the FRT was projected 25, 50, 75 and 100 years from 2016 using these historical annual rates of elevation change. Water depth was projected 25, 50, 75 and 100 years from 2016 using historical rates of annual elevation change plus 2016 local sea level rise (SLR) data from NOAA. Data were collected under Florida Keys National Marine Sanctuary permit FKNMS-2016-068.

The Digital Elevation Model represents ground surface topography between points of known elevation. The elevation data was calculated using the altimeters and Global Positioning System (GPS) sensor used for the benefit of airborne magnetic and radiometric data on the same survey. The elevation is the height relative to the Australian Height Datum GDA94 (AUSGEOID09). The processed elevation data is checked for quality by GA geophysicists to ensure that the final data released by GA are fit-for-purpose. These line dataset from the Gairdner Airborne Magnetic Radiometric and DEM survey, SA, 2018 survey were acquired in 2018 by the SA Government, and consisted of 104788 line-kilometres of data at 200m line spacing and 60m terrain clearance.

Presentation for AWRA Geospatial Technologies Conference held Virtually August 4-13, 2020. This presentation on August 12. https://www.eventscribe.com/2020/AWRAGIS/

Flood inundation is difficult to map, model, and forecast because of the data needed and the computational demand. Recently an approach based on the relative elevation, or Height Above Nearest Drainage (HAND), which is derived from a digital elevation model (DEM), has been suggested for both flood mapping and obtaining reach hydraulic properties and synthetic rating curves. These products are only as good as the underlying DEM from which they are derived and better DEMs offers the potential for improving model representations of streams and parameters derived from DEMs used in hydrologic and flood inundation modeling and mapping. This presentation will review the approach for using relative elevation in flood modeling, describing how HAND is calculated, how it is used to map flood inundation for stream reach catchments and how it is used to determine stream reach properties, identifying shortcomings and giving ideas for improvements. As we obtain more detailed information on bathymetry, topography and hydrography it is important to establish a consistent data model for the river bed that is used in HAND related work that aligns and reconciles elevation and hydrography. This presentation will discuss approaches for using hydrography to remove DEM obstacles, the segmentation of streams used in deriving HAND reach average hydraulic properties and ideas for quantifying reach average roughness based on HAND and flood inundation mapped from remote sensing of previous floods with measured discharges.

Orographic shading or relief map, based on diffuse anisotropic lighting, created from the Digital Surface Model (MDS) of the 2017 LiDAR data. It has been made in grayscale and with a resolution of 50 centimeters. This approach uses terrain visualization techniques that are alternative to traditional ones and that are based on diffuse lighting, rather than direct lighting. These methods enhance the relative elevation of each point against the classic analytical shading, showing the darkest concave surfaces and the brightest convex ones. Ground illumination combining direct and diffuse light sources improves object recognition in high-resolution MDE.

The topographic site effect plays a vital role in controlling the characteristics of earthquake ground motions. Due to its complexity, the factors affecting topographic amplification have not been fully identified. In this study, 100 ground motion simulations generated by double-couple point sources in the homogeneous linear elastic half-space are performed based on the 3D (three-dimensional) Spectral Element Method, taking the Menyuan area of Qinghai Province, China as a local testbed site. A relationship between incident direction and the strength of topographic amplification has been observed. The horizontal ground motion is affected by the back-azimuth, which is typically chosen to be the direction from seismic station to seismic source measured clockwise from north. Specifically, the east-west PGA (Peak Ground-motion Acceleration) is significantly amplified when back-azimuth is about 90° or 270°, and the north-south PGA is significantly amplified when back-azimuth is around 0° or 180°. The vertical ground motion is affected by the dipping angle, which is the angle from vertical at which an incoming seismic wave arrives. The vertical PGA is strongly amplified when the seismic wave is almost horizontally incident (e.g., dipping angle = 78°). A correlation study between geomorphometric parameters and frequency-dependent topographic amplification indicates that relative elevation and smoothed curvature contain similar information, both of which are closely related to the topographic amplification of horizontal components, but not the vertical component. Our study reveals the influence of source and propagation path on topographic amplification and provides a reference for considering the topographic site effect in real engineering sites.