SoftWright has four digital elevation models for the continental United States, all derived from files supplied by the U.S. Government:
· 30-second data The lowest resolution data base has 30 arc-second (latitude and longitude) spacing between data points.
· 3-second data This data base is the highest resolution available for the entire country and has 3 arc-second spacing between data points. The primary source of for the 3-second DEMs (Digital Elevation Models) is the 1:250,000 USGS topographic maps. Larger scale maps, such as 1:100,000 and 1:24,000 (7.5 minute quadrangles) were used in the digitization process to improve the accuracy of the 3-second data. The government provides no information concerning which maps were used for which areas of the country.
· 30-meter data This high-resolution data base uses data from 7½ minute USGS quad maps. The data points are spaced at intervals of 30 meters on a UTM (Universal Transverse Mercator) grid rather that latitude and longitude lines. These data files are also referred to as 7½ minute DEMs. While this data base prepared by the USGS is not complete, many areas have been digitized as 30-meter data. The USGS is continuing to digitize 30-meter data. For information about the availability of specific 30-meter quad files please contact SoftWright.
· 1-second data This high-resolution data base is the USGS National Elevation Dataset (NED) data based on a 1-second latitude-longitude grid.
Data is also available for Alaska and Hawaii in the 1-second, 3-second and 30-second data bases, and for Hawaii in the 30-meter data base.
SoftWright provides several sources of international topographic elevation data for use with the Terrain Analysis Package (TAP)™ software for radio propagation studies. The data bases are generally in one of the following different resolutions:
Topo data in 3-arcsecond resolution is available for many parts of the world. For details about the SRTM (Shuttle Radar Topography Mission) data collection process click here. In addition, data bases are available for selected areas in various resolutions:
- Caracas, Venezuela 30-meter resolution
- Slovenia, (part) 3-second resolution
- Brasil (part) 3-second resolution
Also, 30-arcsecond resolution averaged elevation data is available for much of the earth's land area covering over 90 countries. This data base is derived from the National Geophysical Data Center (NGDC) Global Land One-km Base Elevation (GLOBE) data and from the EROS Data Center (EDC) 30-arcsecond DEM project. Both of these sources consist of elevation values that are averages of approximately 100 elevations in 30-arcsecond square "cells". SoftWright has processed these data bases of average elevations and interpolated the values for use with the TAP software. Since the elevation values on the 30-second grid actually represent averages over an area, the data base provides good general information about the terrain.
SoftWright can have prepared digitized data from available paper topographic maps to produce topographic elevation data bases in the desired resolution for other areas. Satellite prepared data can also be acquired for certain areas of the world. However, the cost in time and money, for these options, will be substantially greater than using prepared data files.
The process of generating a digitized data base from these maps is a function of the elevation contours printed on the maps. Where the terrain is relatively flat, contours are typically scaled at small intervals, such as 10-20 foot contours. However where there are great variations in the elevations, topographic maps use a larger interval, such as 50 - 100 feet between contours. For this reason, the intervals used for digitizing the data base vary throughout the country. Once the contours are digitized, an evenly spaced grid of elevations is mathematically modeled at the prescribed interval. The correlation of the data base to the original data values is then checked for accuracy over the entire area. This mathematically generated elevation grid is stored as the digital elevation model (DEM).
For certain applications (such as terrain averaging) the low resolution 30-second data base may be adequate. For applications that require maximum topographic detail (such as microwave paths or other radio propagation studies) the 30-meter data or the highest available resolution should be used. The TAP program may be used to extract elevation values from the data files along user-specified radials for propagation or profile studies. It can be configured to use 30-meter data where available, and then to "fill in" with 3-second or 30-second data where the 30-meter data is not available.
The 30-meter data provides an increase in grid density of about nine times over the 3-second data. For example, the number of grid points within the area of a 7½" quadrangle is increased from 22,000 in a 3" data file to approximately 202,000 grid points with the 30 meter grid interval. This increase in the number of sampling points, combined with the fact that all 30-meter data is taken from 7½" quad maps for improved accuracy make the 30-meter data decidedly superior. The following profile plots illustrate the difference in detail between the 3-second data and the 30-meter data files:
The actual distance between elevation data points for each data base varies depending on the latitude. Figure 2 illustrates the variation:
In all digital elevation models, peculiarities occurring between locations in the data grid are not reflected in the data base. For this reason, one should not rely exclusively on any data base. It is absolutely necessary to consult topographic maps to ensure greater accuracy.
However, it is usually not necessary to examine a complete propagation path from topo maps. The raw DEM data plotted as a profile will usually indicate which areas require further investigation from topo maps.
When a profile plot is prepared from the data base, and a line of sight and corresponding Fresnel zone is plotted and examined, it usually becomes quite apparent what portions (if any) of the path require further examination. Typically only one or two points need to be investigated and the map elevations entered into the project elevation data file. Using the digitized data base usually reduces the time needed to prepare a reliable path model by 90 to 95% when compared to manual studies.
For example, the following figure shows a path plotted using elevations at 0.1 mile intervals (80 points). Including the line of sight and Fresnel zone for the operating frequency shows that only a few limited areas need further investigation from maps.
For example, in the profile shown in Figure 3, most of the path is obviously clear. However, the elevations from approximately 0.7 to 1.2 miles should be examined on a topographic map, and the actual elevations entered into the project elevation file. Then the profile can be plotted again to check the clearance.
The vertical resolution of the DEMs is one meter. However, the vertical accuracy is generally not within one meter. There is not 100% correlation with topographic maps because of the limitations of the digitization process. The vertical accuracy stated in the government documentation for the 3-second USGS DEMs is ± 30 meters. The stated accuracy for the 30-meter USGS DEMs is 7 or 15 meters, depending on the file. Many areas of the country have been checked and the correlation with map elevations is often within 10 to 20 feet. However, in some cases, the correlation decreases (especially in mountainous areas with drastic elevation variations within short horizontal distances). In such situations, differences of as much as 150 feet have been observed.
For the 30-second data base, there is no stated specification for vertical accuracy. However, SoftWright strives to obtain the highest resolution data available throughout the world and maintain the best possible accuracy throughout the data conversion process.
Since there is no systematic way of knowing the accuracy of a particular area in the data base, it is imperative that at least a few sample extracted elevation points be compared to map elevations, and that peaks on critical paths be examined to obtain improved accuracy from maps. While no digital elevation model is perfect, the 30-meter, 3-second and 30-second data is fully adequate for radio propagation modeling and many other uses when judiciously reviewed as recommended by SoftWright.
    |