Geospatial Technology & Georeferenced Data

"Geomatics (also known as geospatial technology or geomatics engineering) is the discipline of gathering, storing, processing, and delivering geographic information, or spatially referenced information."

Wikipedia, accessed 30-dec-2012, http://en.wikipedia.org/wiki/Geomatics

Geospatial Technology Is Not Rocket Science

A person hearing the term "geospatial technology' for the first time might think that it has to do with space exploration and black holes. In fact, geospatial technology is about the earth surface - what's going on there and where it's occurring:
"geo" = "earth" and "spatial" = "relating to size, shape and position".

Some of the confusion arises from the fact that work in geospatial technology often involves interaction with satellite systems. But we need to remember that space-based tools such as the Global Positioning System (GPS) and LANDSAT exist to locate and quantify what's occurring on Earth. The only reason that we sometimes have to deal with "up there" space is that it's a great vantage point for observing things down here.

Spatial Databases - Combining "What Is" And "Where Is"

Over time, humans have devised tools to document their observations and experiences. We are strongly motivated to recall things accurately for our own benefit and to pass them on to others. Our ancestors developed wonderful tools for recording "what is" information. Oral and written descriptions were often supplemented with "paper" maps, but traditional methods of integrating language and maps always seem to fall short. Does it work best to embed a bunch of descriptive text in a paper map or does the other way around work better. And no matter what the scale of a paper map, it's always too small for for some lines of inquiry and too detailed for others.

Beginning in the 1960's, the development of geographic information system (GIS) concepts brought the power of computers to bear on those age-old problems. GIS is based on classic DBMS (database management systems) principles, but data storage structures and data processing capabilty are expanded to include recognition of absolute location. The specialized data structures that underlie the various branches of geospatial technology have a multitude of application-specific names. The closest thing to an all-encompassing generic name is "spatial database" and we intend to use that term throughout these pages. All spatial databases share a couple of defining properties.

  • "What is" and "where is" information are fully integrated in the same digital file structure.
  • "Where is" information is georeferenced, ie stored on a full-scale model of the earth's surface.

Spatial Databases - When Location Matters

When the spatial database/GIS approach was in its infancy, advocates used the slogan "Location Matters" to target opportunities for the new technology. Today, we've concluded that location matters in almost everything we do and spatial databases are widely used for storing and publishing data. This allows much of the work in geospatial technology to be built on easily-obtained data published in ready-to-use formats. Not everyone is aware of the wide availability of spatial databases. The fact that various disciplines have their own terms for their spatially referenced data is somewhat to blame - think "LANDSAT scene", "digital line graph", "gis layer", digital elevation model", etc.

Spatial DB Design Options - Rasters, Feature Classes, TINs, Etc.

The process of recording location information in spatial databases is called "mapping". The author of a spatial database chooses from a very limited number of basic designs. The choice is somewhat governed by the author's personal preference, but for the most part it is based on the character and complexity of the data. Number of descriptive attributes to be stored, compatibility with analytical software and whether data is discrete or continuous are some of the factors that are considered. Basic designs for mapping spatial data include:

Rasters - A single variable (the variable can be continuous or discrete) is mapped by assigning a numeric value to each cell in a grid that is laid out across a rectangular chunk of the landscape more info.

Point Feature Classes - Vector geometry is used to map features on the earth surface that can be reasonably represented by a discrete point more info.

Line Feature Classes - Vector geometry is used to map features on the earth surface that can be reasonably represented by a series of connected vertices more info.

Polygon Feature Classes - Vector geometry is used to map features on the earth surface that can be reasonably represented by a series of connected vertices forming a closed figure more info.

TINs - A TIN or triangulated irregular network is a highly specialized data structure used for storing 3-dimensional surfaces. A TIN is produced by interpolating triangular "plates" from a set of xyz coordinates. The TINs we encounter in our work are generally terrain models.

LAS - LiDAR Data Exchange Format