Lecture 02 CE797 F24.pdf

Faculty of Engineering Lecture (2) CE 797: Special Topic Civil Eng. Department Raster Data Model Digital Elevation Models and Dr. Suhail A. Almadani...

Faculty of Engineering Lecture (2) CE 797: Special Topic Civil Eng. Department Raster Data Model Digital Elevation Models and Dr. Suhail A. Almadani Chapter (1) ( 2 of 2) GIS Applications in Civil Eng. 1.3 DIGITAL ELEVATION MODELS A digital elevation model (DEM) consists of an array of uniformly spaced elevation data. DEMs are a primary data source for terrain mapping and analysis. A traditional method for producing DEMs is to use a stereoplotter and stereo pairs of aerial photographs (i.e., pairs of aerial photographs of the same area taken from slightly different positions to produce the 3-D effect). The stereoplotter creates a 3-D model, allowing the operator to compile elevation data. This old method is accurate but requires experienced operators and is time-consuming. Another old method is to interpolate a DEM from the contour lines of a topographic map. Several new techniques for DEM generation have been developed in recent years. The following sections cover three such techniques using optical sensors, interferometric synthetic aperture radar (InSAR), and light detection and ranging (LiDAR). Other techniques, which are not included here, are unmanned aerial systems-based photogrammetry and terrestrial laser scanning. 1.3.1 Optical Sensors To make DEMs, two or more optical satellite images of the same area taken from different directions are needed. These stereo images should be taken within a short time interval so that their spectral signatures do not differ significantly. Two optical sensors that readily meet the requirement are Terra ASTER and SPOT 5. ASTER provides a nadir view and a backward view within a minute, and SPOT 5 provides a forward view and a backward view along its orbit. ASTER DEM has a spatial resolution of 30 m. SPOT 5 DEM has a spatial resolution of 20 m. DEMs can also be generated from very high-resolution satellite images such as World-View images if stereo pairs are available. 1.3.2 InSAR InSAR uses two or more SAR images to generate elevations of the reflective surface, which may be vegetation, man-made features, or bare ground. SRTM (Shuttle Radar Topography Mission) DEMs, for example, are derived from SAR data collected by two radar antennas placed on the Space Shuttle in 2000. SRTM DEMs cover over 80 percent of the landmass of the Earth between 60° N and 56° S (Farr et al. 2007). For the United States and territorial islands, they have elevation data spaced 1 arc-second (about 30 meters in the midlatitudes) apart between 0° and 50° latitude and spaced 1 arc- second apart in latitude and 2 arc-seconds apart in longitude between 50° and 60° latitude. For other countries, SRTM DEMs are available at a 90-meter resolution. Higher resolution DEMs than SRTM can now be made from SAR images collected by Sentinel-1, TerraSAR-X, and RADARSAT-2. For example, Airbus Defense and Space distributes DEMs made from TerraSAR-X stereo images at the spatial resolutions of 10 meters, 4 meters, and 1 meter. Name Orbit Type Orbit Height (km) Repeat Cycle (days) Launched in Out of service since Organisation TerraSAR-X Sun Synchronous 514 11 2007 DLR - Germany Textbook: Introduction to Geographic Information Systems, Kang-tsung Chang, McGraw-Hill (2019) Page 1 of 3 Faculty of Engineering Lecture (2) CE 797: Special Topic Civil Eng. Department Raster Data Model Digital Elevation Models and Dr. Suhail A. Almadani Chapter (1) ( 2 of 2) GIS Applications in Civil Eng. 1.3.3 LiDAR The use of LiDAR data for DEM generation has increased significantly since the mid-1990s. The basic components of a LiDAR system include a laser scanner mounted in an aircraft, GPS, and an Inertial Measurement Unit (IMU). The laser scanner has a pulse generator, which emits rapid laser pulses (0.8 — 1.6 µm wavelength) over an area of interest, and a receiver, which gets scattered and reflected pulses from targets. Using the time lapse of the pulse, the distance (range) between the scanner and the target can be calculated. At the same time, the location and orientation of the aircraft are measured by the GPS and IMU, respectively. The target location in a three-dimensional space can therefore be determined by using the information obtained by the LiDAR system. A major application of LiDAR technology is the creation of high resolution DEMs, with a spatial resolution of 0.5 to 2 meters (Figure 1.4). These DEMs are already georeferenced based on the WGS84 ellipsoid. Because LiDAR can detect multiple return signals for a single transmitted pulse, it can produce DEMs of different height levels such as ground elevation (from LiDAR last returns) and canopy elevation (from LiDAR first returns). Thus, LiDAR can be used to estimate forest heights. LiDAR data can be converted to a DEM. 1.4 OTHER TYPES OF RASTER DATA 1.4.1 Digital Orthophotos A digital orthophoto is a digitized image of an aerial photograph or other remotely sensed data in which the displacement caused by camera tilt and terrain relief has been removed (Figure 1.5). The USGS began producing digital orthophoto quads DOQs in 1991 from 1:40,000 scale aerial photographs of the National Aerial Photography Program NAPP. These USGS DOQs are georeferenced and can be registered with topographic and other maps. The standard USGS DOQ format is either a 3.75-minute quarter quadrangle or a 7.5-minute quadrangle in black and white, color infrared, or natural color, with a 1-meter ground resolution. A black-and-white DOQ has 256 gray levels, like a single-band satellite image, whereas a color orthophoto is a multiband image, each band representing red, green, or blue light. DOQs are useful for checking the accuracy of such map layers as roads and parcel boundaries. 1.4.2 Land Cover Data Land cover data are typically classified and compiled from satellite imagery and are thus often presented as raster data. The USGS, for example, offers a series of three land cover databases: NLCD 2001, NLCD 2006, and NLCD 2011. All three databases use a 16-class scheme classified from Landsat images with a spatial resolution of 30 meters (http://www.mrlc.gov/index.php). Textbook: Introduction to Geographic Information Systems, Kang-tsung Chang, McGraw-Hill (2019) Page 2 of 3 Faculty of Engineering Lecture (2) CE 797: Special Topic Civil Eng. Department Raster Data Model Digital Elevation Models and Dr. Suhail A. Almadani Chapter (1) ( 2 of 2) GIS Applications in Civil Eng. 1.4.3 Bi-Level Scanned Files A bi-level scanned file is a scanned image containing values of 1 or 0 (Figure 1.6). In GIS, bi-level scanned files are usually made for the purpose of digitizing. They are scanned from paper or Mylar maps that contain boundaries such as soils and parcels. A GIS package usually has tools for converting bi-level scanned files into vector-based features. Maps to be digitized are typically scanned at 300 or 400 dots per inch (dpi). 1.4.4 Digital Raster Graphics A digital raster graphic (DRG) is a scanned image of a USGS topographic map (Figure 1.7). The USGS scans the 7.5-minute topographic map at 250 to 500 dpi to produce the DRG with a ground resolution of 2.4 meters. USGS DRGs are georeferenced to the UTM coordinate system, based on either NAD27 or NAD83. 1.4.5 Graphic Files Maps, photographs, and images can be stored as digital graphic files. Many popular graphic files are in raster format, such as TIFF (tagged image file format), GIF (graphics interchange format), and JPEG (Joint Photographic Experts Group). 1.5 DATA CONVERSION AND INTEGRATION Although integrated raster and vector data models have been proposed in the literature, raster and vector data remain separate in practicm,9999i5n41e. How to use these two types of data together for projects is therefore of interest to GIS users. This section discusses conversion and integration of raster and vector data. 1.5.1 Rasterization Rasterization converts vector data to raster data (Figure 1.8). Rasterization involves three basic steps. The first step sets up a raster with a specified cell size to cover the area extent of vector data and initially assigns all cell values as zeros. The second step changes the values of those cells that correspond to points, lines, or polygon boundaries. The cell value is set to 1 for a point, the line’s value for a line, and the polygon’s value for a polygon boundary. The third step fills the interior of the polygon outline with the polygon value. Errors from rasterization are usually related to the design of the computer algorithm, the size of the raster cell, and the boundary complexity. 1.5.2 Vectorization Vectorization converts raster data to vector data (Figure 1.12). Vectorization involves three basic elements: line thinning, line extraction, and topological reconstruction. Lines in the vector data model have length but no width. Raster lines in a scanned file, may occupy several pixels in width. ▪ Raster lines must be thinned, ideally to a 1-cell width, for vectorization. ▪ Line extraction is the process of determining where individual lines begin and end. ▪ Topological reconstruction connects extracted lines and shows where digitizing errors exist. Results of raster-to-vector conversion often exhibit steplike features along diagonal lines. A subsequent line smoothing operation can help reduce those artifacts from raster data. Textbook: Introduction to Geographic Information Systems, Kang-tsung Chang, McGraw-Hill (2019) Page 3 of 3