Digital Earth reference model
From Wikipedia, the free encyclopedia
A digital Earth reference model defines a fixed global reference frame for the Earth using four principles of a digital system, namely:
-
- Discrete partitioning using regular or irregular cell mesh, tiling or Grid;
- Data acquisition using signal processing theory (sampling and quantizing) for assigning binary values from continuous analog or other digital sources to the discrete cell partitions;
- An ordering or naming of cells that can provide both unique spatial indexing and geographic location address;
- A set of mathematical operations built on the indexing for algebraic, geometric, Boolean and image processing transforms, etc.
The distinction between "digital" versus "analog" Earth reference model is made in the manner the entire Earth surface is covered. Tessellation refer to a finite number of objects/cells that cover the surface as discrete partitions while Lattice refer to ordered sets of points that cover the surface in continuous vector space. The mathematical frame for a digital Earth reference model is a tessellation while the mathematical frame for an analog Earth reference is a lattice.
The value of a digital Earth reference model to encode information about the Earth is akin to the value obtained from other digital technologies, namely synchronization of the physical domain with the information domain, such as in digital audio and digital photography. Efficiencies are found in data storage, processing, integration, discovery, transmission, visualization, aggregation, and analytical, fusion and modeling transforms. Data reference to a Digital Earth Reference Model (DERM) becomes ubiquitous facilitating distributed spatial queries such as “What is here?” and “What has changed?”. Image and signal processing theory can be utilized to operate on data referenced to a DERM.
The DERM structure is data independent allowing for the general quantization of all georeferenced data sources onto the common grid. Application, algorithms and operations can then be developed on the grid independent of data sources.
Approaches using an analog reference require rigorous manual conflation to satisfy the creation of digital products such as digital maps or other cartographic, navigation or geospatial information (see also GIS). However, digital models are weaker at geometric transformations where translation, scaling and rotation must conform to the discrete cell locations wherein on an analog model with a continuum of locations geometric transformation are straight forward with no requirements for reprocessing or resampling.
Contents |
[edit] Historic Development
The use of the Military grid reference system (MGRS), a regular grid extension of the UTM lattice reference, and a hierarchy of UTM coordinates to name each grid provided an accurate scale dependent position for observations, artillery targeting and navigation is one of the earliest forms of a tesseral based Earth reference. The World Meteorological Organization adoption of a rectilinear grid based on cell defined by geographic degrees (International Exchange Grid see also GRIB) used for synthesis and dissemination of atmospheric data is another. Many alternative methods of gridding and indexing the Earth surface have been proposed.
[edit] New Forms
These early forms of gridded data models have also become known as discrete global grids. Criteria for an optimal discrete global grid have been developed by Michael Goodchild – known as the Goodchild criteria. In the early 1990’s the US-EPA identified the need for a statistically appropriate global grid for the investigation and exchange of biodiversity information (EMAP). Concurrently, the proliferation of space based Earth observation systems (EOS) introduced a new requirements for global grid models. The Committee for Earth Observing Satellites identified and investigated the need for new models to enable imagery and sensor data. ISO Technical Committee 211 began work on projects to define standards for imagery and gridded data in 1997.
A cell shape in such representations can be critical to the validity, adaptability and usefulness of the grid. As rectilinear structures are intuitive but lack optimization characteristics as a tessellation especially when tiled to a sphere, other schemes including voronoi regions, peano curves, triangles and hexagonal tilings have been advanced as superior alternatives.
Many ordering and naming models have been implemented as geospatial database indexing for efficient data retrieval (R-Trees, QTM, HHC). Few of these models have encompassed a complete digital Earth reference model where both a formation of digits that represent a hierarchy where the index contains a parent child relationship and a formation of digits that monotonically converges by a set modulus to all vector Reals.
[edit] Location as the Frame for all Information
The term DERM was coined by Tim Foresman in context with a vision for an all encompassing geospatial platform as an abstract for information flow in support of Al Gore’s vision for a Digital Earth. Here the emphisis is on the requirements for a "Digital Earth" reference model with the emphasis on the need for a data structure for Digital Earth application, as opposed to an "Earth reference model" that is Digital. Several key challenges were identified for this DERM, namely data interoperability and integration. The International Society on Digital Earth has a standing committee considering DERM implementations and standards which includes both the Earth reference frame and the ancillary requirements for metadata and attribute semantics.
By synchronizing location within a hierarchical linear indexing, a DERM can act to connect all locations at any level of granularity just as IP addressing connects all nodes on a network.
For Network Design Engineers: “User’s location will become information that is as common as date is today … the challenge is to integrate the concept of physical location into the design of the Internet which relies on logical addressing”, Julio Navas, Tomasz Imielinski, GeoCast Geographic Addressing and Routing Mobile Computing and Networking (1997). Applications such as ubiquitous computing, augmented reality, command & control, and location-based services depend on a self synchronized physical model of the space they reside in.
