Geospatial Data

The following section will cover the use of geospatial data for this project. Most of the information here was acquired from www.vterrain.org.

Creating three dimensional representations of the real world places involves converting feature data (elevation, roads, ground cover, etc.) into 3D models. Geospatial data refers to any such data that describes Earth's features over a certain region.

The main challenge in this project is to combine the different data sets together, such as mapping street information onto terrain contours. Urban environments will probably be the hardest to model because of the density and overlap of so many features.

The chief player in the development of GIS (global...) is the United States Geographical Survey, a government organization that specializes in recording almost every type of geographical data of interest. The majority of the data collecting is usually done by individual states. Simliar organizations exist in other countries as well.

Here's a list of various data sources that can be used in producing virtual environments. The data sets are generally downloadable free of charge (tax dollars at work), except for data that is deemed to be critical to national security. More specifically, this refers to elevation and image data which come in several detail levels, with the most detailed versions not in public domain.

Elevation

DEM (Digital Elevation Model)

DEM is a series of files that contain a regularly spaced grid of elevation data for a region. The coverage for DEM is mostly for the United States. A typical coverage areas are --- for 1:24000 data sets and --- for 1:250000 sets. Recently, the USGS has been phasing out the DEM file format and for the newer, more generic, but much more complex SDTS file format, which is supposed to be suitable for storing all geospatial data, especially DEM data.

DLG (Digital Line Graph)

DLGs also contain elevation data but store it using polygons representing equal height contours. The advantage is that this method takes up much less space than a regularly spaced elevation grid for relatively flat terrain. Just like DEMs, the USGS has been integrating DLG files into SDTS data sets.

GTOPO30 and SRTM (Shuttle Radar Topography Mission)

These data sets have virtually global coverage. GTOPO30 elevation data is on a regularly spaced grid of 1km per point. SRTM is based on GTOPO30 data except that it's much more promising for terrain generation since it has much more accuracy (30m to 90m).

Image data

Generally, providing image data coverage for the entire globe is more difficult than doing so for elevation because a. It is dependent on the time of day and atmospheric conditions (clouds). b. The ammount of storage needed for detailed image data is astronomical even by today's standards. A typical USGS Digital orthoquadrangle, which provides 1 meter resolution for a 7 square km area takes up 150 megabytes.

Multi-Feature Data

TigerLine

TigerLine data sets primary contain information on every street in the United States as polygonal data. It also contains information on features such as rivers, political boundaries and landmarks. The file format is ASCII text.

Coordinate Systems/Map Projections

The most straightforward way to describe any arbitrary position on earth is to use spherical coordinates, namely a longitude, latitude pair because the earth is mostly spherical.

Even though longitude, latitude is the most universal way, there is the need to display a small detailed area onto a map, which means converting 3D coordinates into 2D, called a projection. There are several dozen map projections and this is beyond the scope of this paper.

In addition, to make an accurate measurement, the Earth cannot be assumed to be a sphere. Instead, an ellipsoid model of the earth is used. There are many ellipsoid models (called datums) of the Earth, but they are all characterized mathematically by the length of the major and minor axis. Often, two coordinates that are the same in a coordinate system might not correspond to the same physical location on Earth because the coordinates could be based on different datums.

Here's a list of common coordinate systems

Geographical: (longitude, latitude)

This is the simplest and most universal system and has been used for hundreds of years. This is especially favored by sailors because the coverage is global.

Universal Transverse Mercator

This system uses a mercator projection based on projecting a 6 degree verticle slice of the earth onto a cylinder. Conceptually, this coordinate system lets you imagine the surface of the Earth as a flat sheet within certain bounds. The coordinate pairs called easting, and northing therefore correspond to points in a cartesian plane.