No matter of GIS or DGGS, the goal is digitalization, meaning to model the spatial information, store it in the computer memory, probably do some analysis or statistics, and visualize it on the screen. The nature of modeling is to simplify a complex phenomenon to some degree, and thus to set the stage for analysis. An example is the Earth’s surface: we all know that Earth is not a regular solid, the surface is rough. Even we use geoid (best-fit, global mean sea level) to present the surface, it is still complex because of bumps and hollows. Therefore, geology scientists simplified it by an ellipsoid or even spheroid. So, when talking about the advantages of DGGS, we are essentially talking about how good DGGS is in modeling and analyzing spatial information, at a degree of simplification while without losing significant fidelity.
Hence, from these perspectives, DGGS can:
- Locate a spatial position via a cell/cells, at a certain resolution level, by a unique, finite cell address. When storing this information in the computer memory, the location (cell address) is accurate, without any round-up. This accurately reflects the topography of the Earth and acknowledges the essential discreteness of location representation.
- If we use lat/lon to present a spatial location, when stored in the computer memory, decimal digits cannot be infinite. In other words, we cannot locate a dimensionless point precisely.
- Simple spatial operations such as an equality test can result in semantic errors with floating-point values.
- All computer representations of location, both raster and vector, are essentially discrete. Given n bits, we can distinguish 2[^n] unique values, and all other points must be mapped to these.
- The core of all geospatial applications are data models representing locations. Thus, even minor efficiency improvements in location representation can result in substantial performance increases.
- Refine the cells with an almost equal area at each of the resolution levels. Spatial resolution is always explicit and approximately constant. Each digit in the index corresponds to a single precision in the representation.
- Graticule or Web Mercator Tiles are commonly used grid systems, but their cell size and shape are not uniform. Thus, it is not practical for modeling or statistics. Because of the inherent flattening, the apparent spatial resolution varies across the surface of the Earth.
- Align the spatial attributes (landcover/temperature/elevation/etc.) with the exact location defined by cells, which conducts congruent geography.
- Traditional GIS slices the spatial information into layers, not effective for location-based queries.
- Do analysis more accurately because DGGS essentially consist of spherical cells considering Earth’s curvature. A shift from geoprocessing in the Euclidean space to the geographic domain can greatly reduce the projection deformations.
- Vector and raster models are essentially flat data models in the Euclidean space.
- Provide hierarchical resolutions.
- With traditional GIS, people usually observe phenomena at a single resolution, or define the multi-resolution vaguely. Also, because of the likely usage of projections, the area may be distorted so that the resolution is not necessarily consistent.
- Avoid hiatus at longitudes or latitudes, so that people can treat the Earth as an intact surface
without having to adopt special map projections for polar regions.
- With traditional GIS methods, researchers have to apply special map projections for polar regions or high-latitude areas, which makes the polar region isolated from the rest of the Earth.
-
Serve as a uniform data model to transform geospatial information from various data sources and to be independent of the original data formats.
-
Benefit parallel computing given its discrete nature.
- Convey information without a visual deformation of the content when being displayed, because all parts of the Earth’s surface are treated consistently and fairly.
- Projections are frequently used in the traditional GIS, where the distortions are unavoidable. End-users are left with the responsibility to understand and account for those distortions in the interpretation of results.