Mingke Erin Li

Future Work on DGGS

maria

Geospatial analytics for the globe

The OGC Abstract Specification has stated three basic functional algorithms that should be supported by DGGS, including quantization operations, algebraic operations, and interoperability operations [1]. Specifically, quantization operations realize the data assigning and retrieving, algebraic operations realize the cell navigation and basic spatial analysis, and interoperability operations realize the conversion of cell addresses between different DGGS specifications or between a DGGS specification and the traditional graticule [1]. However, other than these basic functional operations, other extended operations are still being expected, because although currently a bunch of geospatial analytics have been developed and widely used, most of them are based on planar geometries, and designed for traditional data structures like vector and raster. First, some spatial analysis operations need to be redefined with DGGS. For example, due to the flexibility of cell shapes used in DGGS, the cell adjacency and neighbor ordering (e.g., triangular or hexagonal cells) may be different from the traditional concepts of raster cells. Thus, the algorithms in geostatistics, interpolation, sampling, and hydrology analysis should be redeveloped. Second, some operations with DGGS are superior to those with traditional GIS, such as generalization and determine topological relationships. Previously, generalization needs to be done by, for instance, resampling a raster or removing a line’s redundant vertices, while with DGGS those unnecessary details can be reduced by switching to a coarser resolution. Third, some analytical algorithms can be omitted with DGGS, such as converting different data structures (e.g., raster and vector), transforming among different projections, or map registration. Because of the fixed cell geometry and locations in DGGS, information associated with specific cells are aligned automatically. Hence, operations with DGGS can avoid those complex data transformation process. Data integration becomes much more convenient in DGGS. Efforts have been made to develop hexagonal-cell raster [2-5], and the cell-based image analysis can still be realized with DGGS, like image enhancement, image algebra, focal statistics, etc. In fact, some research have been carried out to develop analytical functions in the context of DGGS, including the managing terrain data [6], interpolating spherical grids [7], determining topological relationships [8], geocoding textual documents [9], developing gazetteer service [10], etc.

Comprehensive DGGS platforms

As summarized in Table 2, there have been some implementations and platforms developed to create DGGS cells (e.g., H3, OpenEAGGR, DGGRID, rHEALPix, and Geogrid [11-16]), integrate multi-source datasets (e.g., PYXIS WorldViewTM [17]), or serve as an application for specific purposes (e.g., HEALPix and SCENZ-Grid; [18, 19]). Each of these DGGS implementations has its own superiority, supported functionalities, practical scenarios, and potential user communities. For example, PYXIS WorldViewTM is one of few DGGS platforms that offers a user-friendly interface so handy to operate even for school kids [17]). It is also the only one able to integrate multi-source data compliant with the OGC Abstract Specification, and carry out basic spatial analysis and generate statistical summaries. However, the constructive DGGS elements including the basic polyhedron, cell shape, aperture, orientation, and projection methods are inaccessible for users to change. On the contrary, other DGGS implementations allow users to define the DGGS components and transform among different grid designs without using longitudes and latitudes (e.g., H3, OpenEAGGR, DGGRID, and rHEALPix [11-15]), while the major purpose is to create or index discrete global grids instead of conducting spatial analyses directly based on such applications. In addition, although being open-sourced, these implementations do not have a user-friendly interface and largely rely on command-line scripts, which narrowing the user communities to specialists in geospatial and programming. Therefore, although there have already been some existing DGGS implementations or platforms, a more comprehensive one is still in need, which would allow users to define their preferred DGGS elements, realize multi-source data integration, and able to run spatial analyses and even complicated spatial-modeling, with a user-friendly visualization interface. In other words, the comprehensive platform should bridge the gap between the massive datasets and the final decision-makers even without much geospatial expertise by offering various functionalities and recommended DGGS designs.

DGGS with higher dimensions

As a fully 3D representation of the entire Earth, volumetric DGGS address not only the Earth’s surface but also the third spatial dimension reaching to the interior and exterior of the surface. As 2D DGGS implementations discretizing the planet’s surface into uniform cells, 3D DGGS realize a volumetric discretization of the sphere with a hierarchical spatial partitioning scheme. Although some work has been done to develop such volumetric DGGS (e.g., [20-22]), there are still a lot of topics that need further discussion, including the representation of surface data, geospatial volumes storage, indexing, efficient visualization, potentials of application, etc. In the same manner, DGGS can be extended to include the time dimension to support the analysis of the Earth through time. Consequently, a series of terms or definitions need to be extended in the future. For example, to extend a 2D DGGS coordinate reference system which is analogous to an ISO19111 geodetic coordinate reference system to higher dimensions for including vertical or time-series measurement. Also, to extend the DGGS domain from the surface of the entire Earth to include vertical and time dimensions.

Citation

  1. OGC. Topic 21: discrete global grid system abstract specification. 2017 [cited 2019 Oct. 15]; Available from: http://www.opengis.net/doc/AS/dggs/1.0.
  2. Petersen, D.P. and D. Middleton, Sampling and reconstruction of wave-number-limited functions in n-dimensional Euclidean spaces. Information and Control, 1962. 5: p. 279-323.
  3. Mersereau, R.M., The processing of hexagonally sampled two-dimensional signals. Proceedings of the IEEE, 1979. 67(6): p. 930-949.
  4. Staunton, R.C., The design of hexagonal sampling structures for image digitization and their use with local operators. Image and vision computing, 1989.
  5. de Sousa, L.M. and J.P. Leitão, HexASCII: a file format for cartographical hexagonal rasters. Transactions in GIS, 2018. 22(1): p. 217-232.
  6. Dutton, G., Geodesic modelling of planetary relief. Cartographica: The International Journal for Geographic Information and Geovisualization, 1984. 21(2-3): p. 188-207.
  7. Carfora, M.F., Interpolation on spherical geodesic grids: a comparative study. Journal of Computational and Applied Mathematics, 2007. 210(1-2): p. 99-105.
  8. Li, J., H. Zhang, and W. Wang, Fast and robust point-in-spherical-polygon tests using multilevel spherical grids, in Next Generation Computer Animation Techniques, J. Chang, et al., Editors. 2017. p. 56-66.
  9. Melo, F.J.S. and B. Martins, Geocoding textual documents through a hierarchy of linear classifiers, in Portuguese Conference on Artificial Intelligence. 2015, Springer, Cham: Portugal.
  10. Adams, B., Wāhi, a discrete global grid gazetteer built using linked open data. International Journal of Digital Earth, 2017. 10(5): p. 490-503.
  11. Sahr, K. DGGRID. 2019 [cited 2019 Nov. 21]; Available from: https://www.discreteglobalgrids.org/software.
  12. Uber. H3: a hexagonal hierarchical geospatial indexing system. 2017 [cited 2019 Nov. 25]; Available from: https://github.com/uber/h3.
  13. OpenEAGGR (Open Equal Area Global GRid). 2017 [cited 2019 Nov. 26]; Available from: https://github.com/riskaware-ltd/open-eaggr.
  14. Bowater, D. and E. Stefanakis. rHEALPix. 2019 [cited 2019 Nov. 25]; Available from: http://atlas.gge.unb.ca/rHEALPix.
  15. Bowater, D. and E. Stefanakis, An open-source web service for creating quadrilateral grids based on the rHEALPix discrete global grid system. International Journal of Digital Earth, 2019: p. 1-17.
  16. Mocnik, F.-B., A novel identifier scheme for the ISEA aperture 3 hexagon discrete global grid system. Cartography and Geographic Information Science, 2018. 46(3): p. 277-291.
  17. Global Grid Systems. 2019 [cited 2019 Nov. 20]; Available from: https://www.globalgridsystems.com.
  18. JPL. HEALPix. 2018 [cited 2019 Nov. 20]; Available from: https://healpix.sourceforge.io/index.php.
  19. SCENZ-GRID. 2010 [cited 2019 Nov. 27]; Available from: https://www.seegrid.csiro.au/wiki/SCENZGrid/WebHome.
  20. Yu, J. and L. Wu, Spatial subdivision and coding of a global three-dimensional grid: Spheoid Degenerated-Octree Grid, in 2009 IEEE International Geoscience and Remote Sensing Symposium. 2009. p. II-361-II-364.
  21. Rohlf, J.H. and M.D. Hancher, System and methods for storing and retrieving geospatial data, U.P.a.T. Office, Editor. 2012: Washington, DC, US.
  22. Wan, G., et al., Sphere shell space 3D grid. ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2013. XL-4/W2: p. 77-82.