Integration Platform for Canadian Terrain Data by Adopting Discrete Global Grid Systems

Terrain data sources in Canada primarily include the Canadian Digital Elevation Model (CDEM) and the High Resolution Digital Elevation Model (HRDEM) distributed by Natural Resources Canada [1, 2]. Given their different coverage, resolutions, accuracies, and horizontal and vertical datums, end-users in need of both datasets may come across complicated, duplicated pre-processing and end up with inconsistent terrain base data for their following application or academic analysis. Thus, a standardized framework for national terrain data across various scales is required for convenience at the user’s end.

Meanwhile, Discrete Global Grid Systems (DGGS) provide good opportunities to serve as a uniform data model to integrate geospatial information independent of the sources. A DGGS is a spatial reference system that uses a hierarchical tessellation of cells to partition and address the globe [3]. It is promising in the terrain data integration context because of the following key features:

1. Uniform cells. A DGGS is a single consistent framework with constant spatial resolution at each level.
2. Spherical cells. It is hard to define a single projected system that minimizes distortions over a large landscape such as Canada. Also, spherical cells account for Earth’s curvature and avoid potential computational errors. This makes DGGS superior to projected rasters.
3. Discrete cells. Cells are independent of each other so that they can facilitate parallel computing.
4. Hierarchical structure. Because of the hierarchical tessellation, DGGS naturally provide multiple resolution options.

This research explored the adoption of DGGS as an integration platform for Canadian terrain datasets. The DGGS configuration used in this research is the Icosahedral Snyder Equal Area Aperture 3 Hexagonal Grid (ISEA3H; Figure 1). The ISEA3H scheme has been suggested by Sahr et al. and has been adopted in several DGGS implementations [4]. Algorithms were developed to integrate the CDEM and HRDEM by direct quantization at various granularities, and modeled elevations were aggregated by mean, maximum, and minimum statistics across the resolution levels.

Figure 1. Icosahedral Snyder Equal Area Aperture 3 Hexagonal Grid (ISEA3H) at two successive resolution levels.

The modeling process in the ISEA3H DGGS was tested on an experimental region around Algonquin Provincial Park, Ontario, spanning 77.36 to 78.86° W and 44.51 to 46.01° N. The computation was conducted on the Advanced Research Computing cluster at the University of Calgary, with a combination of job-level parallelism and shared-memory parallelism. Quantization results at the finest resolution level are shown in Figure 2. Aggregation results at the coarsest resolution level are illustrated in Figure 3.

Figure 2. Quantization results in the experimental region at the finest level (level 29) in the ISEA3H DGGS.

Figure 3. Aggregation results in the experimental region at the coarsest level (level 16) in the ISEA3H DGGS.

The outcomes of the research can set the stage for a national elevation service across various scales for Canada. Future work includes the development of in-database analytical functions in the ISEA3H DGGS, such as the topographical and hydrological functions. An operational integration platform for national elevation data can then be developed to support decision-making in the real world. A DGGS-powered geospatial datacube is also prospective to offer more analysis-ready data to end-users in future.

Funding sources: Canadian Natural Sciences and Engineering Research Council Discovery Grant program.

Collaborators: Natural Resources Canada

References

[1] Natural Resources Canada. Canadian Digital Elevation Model, 1945-2011, Canada Centre for Mapping and Earth Observation, Strategic Policy and Results Sector, Natural Resources Canada, Government of Canada. [Online]. Available: https://ftp.maps.canada.ca/pub/nrcan_rncan/elevation/cdem_mnec/

[2] Natural Resources Canada. High Resolution Digital Elevation Model (HRDEM)-CanElevation Series, Canada Centre for Mapping and Earth Observation, Strategic Policy and Results Sector, Natural Resources Canada, Government of Canada [Online]. Available: https://ftp.maps.canada.ca/pub/elevation/dem_mne/highresolution_hauteresolution/dtm_mnt/

[3] Open Geospatial Consortium. "Topic 21: discrete global grid system abstract specification." http://www.opengis.net/doc/AS/dggs/1.0 (accessed 15 November, 2019).

[4] K. Sahr, D. White, and A. J. Kimerling, "Geodesic discrete global grid systems," CaGIS, vol. 30, no. 2, pp. 121-134, 2003.