dwc-qa
dwc-qa copied to clipboard
tdwg
What is a georeference?
This question comes from various sources. Related questions are, "What is the minimum requirement for a georeference?" and "What is the gold standard for a georeference?"
Related to this, a question/comment from Rich Rabeler (University of Michigan Herbarium):
"One other point related to the verbatim coordinates comes to mind. One feature of Symbiota is an autoconversion of DMS lat/long to decimal lat/long. One of the reasons that you might be seeing quite a few what one could call "bare" georeferenced data is the assumption that having Symbiota do the conversion completes the process since there are no other required fields that one needs to enter. Partly it's training and partly resources. I suspect the Covid-19 effect on collection budgets may find us working with fewer resources and possibly having to take the "good enough" approach more than some would like."
From @debpaul:
"Here I see lots of "opportunities" to automate filling some of the metadata fields to "reveal" that the process you describe is happening -- rather than just you "knowing" that's what happened. It doesn't have to take more time on staff part, but it does take a) planning, and b) knowing what your options are for some programmatic help --- so that you c) know what you can / ought to ask for from your software providers / developers."
What is a georeference?
Drawing from the definition of georeference in the glossary of Georeferencing Best Practices (Chapman and Wieczorek 2020)...
georeference
the process (verb) or product (noun) of interpreting a locality description into a spatially mappable representation using a georeferencing method. Compare with geocode. The usage here is distinct from the concept of georeferencing satellite and other imagery (known as georectification).
Since it relies on other concepts with definitions in the glossary, I'll include some of those here too.
locality
the verbal representation of a location, also sometimes called "locality description".
location
a physical space that can be positioned and oriented relative to a reference point, and potentially described in a natural language locality description. In georeferencing, a location can have distinct representations based on distinct rules of interpretation, each of which is embodied in a georeferencing method.
rules of interpretation
a documented set of steps to take in order to produce a standardized representation of source information.
georeferencing method
the type of spatial representation produced as the output of a georeferencing protocol. In this document we discuss three particular methods of representation in detail, the shape method, the bounding box method, and the point-radius method.
georeferencing protocol
the documented specific steps to apply to a locality, based on the locality type, to produce a particular type of spatial representation.
locality type
a category applied to a locality clause that determines the specific georeferencing method that should be applied.
locality clause
a part of a locality description that can be categorized into one of the locality types, to which a specific georeferencing method can be applied.
geocode
the process (verb) or product (noun) of determining the coordinates for a street address. It is also sometimes used as a synonym for georeference.
What is the minimum requirement for a georeference?
Again, from Georeferencing Best Practices..
3.3 Georeferencing Methods The distinction between georeferencing methods is in the basic approach taken to capture spatially enabled location data. Within each method there should be protocols for how to produce georeferences based on the input locality description and supporting information. The goal of any georeferencing method and its specific, documented protocols should be to create a spatial representation of the entire location, including all uncertainties involved, with sufficient accompanying information and documentation to make the georeference reproducible.
Five methods are covered: point, point-radius, bounding box, shape, probabilistic. Of these, the point method (coordinates only) is considered impossibly deficient except to provide a view onto a map. All of the others have in common that they create a spatial representation that contains all the reasonable interpretations of the entire location. That is, they include the interpreted extent of the location.
From the Best Practices glossary:
extent
the entire space within the boundary a location actually represents. The extent can be a volume, an area, or a distance.
boundary
the spatial divide between what is inside a location and what is outside of it.
What is the gold standard for a georeference?
A full georeference is one that allows a potential user to determine if it is fit for a particular purpose.
Again from the Georeferencing Best Practices...
The best locality description is one that is specific, concise, complete, and unambiguous; the more detailed the description, the better the chance of getting a georeference that is reproducible and that minimizes the uncertainty.
From the Best Practices glossary:
uncertainty
a measure of the incompleteness of one’s knowledge or information about an unknown quantity whose true value could be established if complete knowledge and a perfect measuring device were available (Cullen & Frey 1999). Georeferencing methods codify how to incorporate uncertainties from a variety of sources (including accuracy and precision) in the interpretation of a location. Compare accuracy, error, bias, precision, and false precision.
To be reproducible and complete, the georeference must be documented with all of the georeference terms that apply (in addition to the Darwin Core Location terms on which the georeference interpretation is based). These include: {decimalLatitude, decimalLongitude, coordinateUncertaintyInMeters, geodeticDatum} or {footprintWKT, footprintSRS} plus georeferencedBy, georefrencedDate, georeferenceProtocol, georeferenceSources, georeferenceVerificationStatus, and georeferenceRemarks.