DCS Spatial Services now provides authoritative datasets in the Geocentric Datum of Australia (GDA2020). In addition, some DCS services continue to support requests for data in alternate datums.
GIS users should update software and workflows to support GDA2020 and/or associated transformations. Contact your software vendor for GDA2020 support.
GIS software supporting the NTv2 grid ‘conformal plus distortion’ transformation is recommended.
ICSM’s GDA Transformation Products and Tools may assist with planning a transformation strategy.
WGS 84 and Web Mercator: Potential for misalignment
WGS 84 and its ‘Web Mercator projection’ have historically been used as a defacto standard in web mapping applications. Unfortunately, these applications often assume that GDA94 ≈ WGS 84 ≈ GDA2020.
Misalignments of approximately 1.8 metres may occur when projecting GDA94 and GDA2020 spatial data into WGS 84 / Web Mercator. This issue affects all web mapping and GIS software.
WGS 84 in the web mapping context is actually a low-accuracy static ‘ensemble’ of different WGS 84 versions defined over time, and usually ignores the movement of the tectonic plates (up to 7 cm / year in Australia). Depending on the source, a WGS 84 dataset may be equivalent to GDA94, GDA2020, ATRF@date or other regional datums. Historically, this information is rarely used to qualify WGS 84 metadata. WGS 84, without further clarification, is therefore suitable for low accuracy applications only (~3 m). Refer to this WGS 84 information sheet for more information.
As a result, metre-level misalignments can occur when combining WGS 84 datasets from different sources. Many WGS 84 datasets would be better described in terms of well-defined datums of high accuracy where known (for example, GDA94, or GDA2020). Enquire after additional metadata to better define WGS 84 data before collating datasets.
Where DCS Spatial Services provides data and web services in WGS 84, these will be described in metadata as WGS firstname.lastname@example.org (i.e. equivalent to GDA94) or WGS email@example.com (i.e. equivalent to GDA2020). Historic datasets may remain unlabelled and are usually assumed to be equivalent to GDA94.