Available with 3D Analyst license.
In ArcGlobe, you can view vector data as rasters. Rasterizing feature layers in ArcGlobe allows you to maintain any cartographic symbology that you may have saved in ArcMap. It's also an effective way to drape features on the globe surface. By default, 2D points, lines, and polygons are added to ArcGlobe as rasterized features; 3D points and lines are added as vectors.
Alternatively, in ArcGlobe, you can convert vector data into raster data for display. This is also supported in ArcScene. Converting data into a consistent raster format is also the recommended approach for many analysis tasks.
Displaying features as rasters is like draping a grid containing square cells over the study area. The size of the square cells and the type of vector data being represented will greatly impact the visual effectiveness of the layer. When using ArcGlobe to represent rasterized vectors on the fly, you have the option of having the features maintain a fixed map unit size to represent precise areas (like having a reference scale set in ArcMap) or maintain a pixel size so you can always see the feature regardless of your distance from it.
Converting vectors to a raster dataset is less dynamic. A code is assigned to each cell according to the feature (or features) that is in each cell on the raster. The code or value of a cell is a numeric value that corresponds to an attribute type. Each cell then represents a specified portion of the world and can be any size you define. There is the possibility of a loss of accuracy when multiple features coincide within the same cell. While the primary consideration for converting vectors to rasters is usually analysis requirements, you should also be aware of how the raster dataset represents the features in the view. For example, having a small cell size for sparse point features results in them seeming to disappear as you zoom out.
Below is further information on how various geometry types deal with being represented as rasters.
Point data represents any object with known coordinates that at a certain resolution appear only as a point. Although a well, a telephone pole, or the location of an endangered plant are all features that can be represented as points at some resolutions, at other resolutions they do, in fact, have area. For example, a telephone pole viewed from an airplane 2 kilometers high is represented by only a point, but the same pole viewed from an airplane 25 meters high is represented by a circle. The resolution where a point object becomes an area is important to consider when deciding on the cell size.
When displaying points as a raster in ArcGlobe, the geographic size of the point symbol and the raster cell size combine to create the rasterized image added to the view. If the symbols and cell sizes are small, the points disappear as you zoom away from the layer. If the symbols are large and the cell sizes are small, you see the points from longer distances and they become fine grained as you zoom in. If the symbols are large and the cell sizes are large, you see the points from long distances and they remain coarse in appearance as you zoom in.
When converting points to a raster for use in ArcGlobe, symbols become irrelevant and only the cell extent that encompasses a point receives the value of the attribute of the point data that is being converted. There is also the possibility of some generalization of the original data. For example, if two or more points fall within the extent of a cell, one of the points is randomly selected to be used as the cell value. Thus, it is possible to have fewer cells with values than there are points being converted.
By definition, a point has no area, but it is converted to a cell representing area. In a raster dataset, point features are represented by the smallest unit of a raster: a cell. It is important to remember that a cell has area as a property. The smaller the cell size, the smaller the area and, thus, the closer the representation of the point feature. Points given area in a raster dataset have an accuracy of plus or minus one-half the cell size.
Linear data represents all those features that, at a certain resolution, appear only as a line (such as a road, stream, or power line). A line, by definition, does not have area. In a raster dataset, a line can be represented only by a series of connected cells. As with a point, the accuracy of the representation varies according to the scale of the data and the resolution of the raster dataset.
As with point data, linear features become the width of the cell. If linear features representing roads are converted to a 1-kilometer cell size, the width of the roads becomes 1 kilometer. Thus, you should choose a cell size that is appropriate to the linear feature you are representing.
Polygonal, or area, data is best represented by a series of connected cells that most accurately portrays its shape. Examples of polygonal features include buildings, ponds, soils, forests, swamps, and fields.
When converting polygons to grids, each cell in the resulting output raster dataset from the conversion process is assigned the value of the feature that most fills the cell or that is encountered in the scan process within the cell.
Trying to represent the smooth boundaries of a polygon with a series of square cells can present some problems, the most infamous of which is called the jaggies, an effect that resembles stair steps. For larger raster datasets with millions of cells, jaggies become insignificant when a finer cell resolution is used.
To reiterate, the accuracy of the above raster representation is dependent on the scale of the data and the size of the cell. The finer the cell resolution and the greater the number of cells that represent small areas, the more accurate the representation and the larger the file size of the resulting raster dataset.