Data Format Interoperability--Key Formats

Most GIS and geomorphometry programs have their own internal format, which is designed for optimal storage and processing for the particular program.  Most programs can convert a number of other formats into their internal format, and can export to a number of other formats.

Three formats have the most versatility, and widest acceptance, for the three main classes of data.  Rasters have traditionally been the data used by geomorphometry, but the increasing availability of lidar data sets may drive algorithm development toward the raw point cloud instead of the derived grids. 

GDAL provides translatiton to and from a large number of formats, relieving GIS programs from handling all variants of formats like Geotiff, and providing a standard against which to measure data set compatibility (if GDAL cannot read it, there is a problem with the implementation of the data format).

	Raster	Geotiff especially with the GDAL drivers.  Problems include: Tiling may cause issues Inconsistent handling of missing data GDAL can converta Geotiff that uses unusual options in the Geotiff specification, which allows an almost limitless number of coding options, or if the projection is not natively supported by the GIS program.
	Vector	Shapefile ESRI spec did not call for datum projection PRJ file not always present, and not always interpreted correctly Shapefile specs from ESRI GDAL can reproject data if the projection is not natively supported by the GIS program.  Geographic (lat/long) provides the most versatile coordinates.
	Point cloud	LAS with LAZ compression   Web format description: https://www.asprs.org/divisions-committees/lidar-division/laser-las-file-format-exchange-activities
	Raster/Vector Mix	KML format Lacks a way to manipulate or search for data attributes.  Combines raster and vector data in a single file. Can be compresses as KMZ file Good distribution format to non-GIS users Open standard at Open Geospatial Consortium (OGC).

With the ability to exchange data in common formats, users can use several programs to capitalize on the unique capabilities of each.

Some programs will require a separate import step, and write a new version of the program in their native format, while others will convert on the fly and keep the native version of the data set in memory.

Formats to watch:

• OGC Geopackage: supports raster (especially tiled data) and vector data; design had mobile devices as key motivating factor.  Apparently not yet a lot of data. GDAL is supposed to be able to import.

Formats to avoid:

• World files
• XYZ format
• ESRI geodatabases
• ESRI zLAS

Last revised 5/30/2021