GIS and Remote Sensing in Geosciences

Data formats

The following classes of GIS data may be distinguished:

vector data: the features of interest are represented by points, lines or polygons. The vector format is suitable for discrete features like roads, buildings, (coastlines) or land ownership patterns. Design software packages (e.g. AutoCAD) work almost exclusively with vector data. The difference to GIS vector data, however, is that the latter carries geographically meaningful coordinates.
raster data: the data consists of a regular grid of cells, comparable to a digital image. Actually, raster data are just an image with a spatial reference assigned. Raster data formats should be used for all themes carrying spatially more or less continuous information which cannot easily be converted into discrete units should, e.g. satellite imagery or digital elevation data.

In general, each software package has its specific formats for vector and raster data. These formats are sometimes, but not always compatible with other software packages and shall not be discussed in detail in this place. However, there are some formats readable by virtually all packages and therefore best suitable for data exchange.

Raster

TIFF

The TIFF format is very well suitable for exchanging satellite imagery. The coordinates are stored either directly in the file (so-called Geotiff) or in a so-called world file. The world file has to be stored in the same folder as the image and has to carry the same name, but with the extension .tfw. The same concept exists with jpg files (world file .jgw), but it is not recommended to use jpg for spatial data due to reduced quality by compression.

ASCII Raster

ASCII Rasters follow a relatively simple scheme, just representing a text file with a six-line header and the cell values. Simple ASCII rasters can be created manually with a text editor. Even though the format was developed by ESRI, it is compatible with many other software packages and well suitable for data exchange. However, many packages require an import to the software-specific raster format for further analysis.

Vector

Shapefile

Even though developed by ESRI, the Shapefile can be considered as the lingua franca of vector data processing. Shapefiles are readable by many software packages, though some require conversion to the software-specific format for further processing. One single shapefile may contain point, line or polygon features, but never more than one of these feature types. Shapefiles actually consist of at least three single files with the same name (differing by extension) which have to be stored in the same folder.

Data table

Some software packages can use simple columnar text files with x and y coordinates and one attribute per further column. Such files require a header line. This concept only works for point features and the data has to be imported to the program-specific data format for further processing.