This page contains a record of the results of georeferencing for the collections participating in the ORNIS project. There were two separate processes used to produce the georeferences repatriated in the ORNIS project:
ORNIS Georeferencing Work Center Results
- Canada and the United States were georeferenced under NSF funding using the steps outlined on the Georeferencing Workflow page following the MaNIS/HerpNet/ORNIS Georeferencing Guidelines (Wieczorek 2001) and GBIF Guide to Best Practices in Georeferencing (Chapman and Wieczorek 2006)
- Mexico was georeferenced by the Mexican Bird Atlas Project (Navarro et al. 2003). This source does not include uncertainty measures, but is a rich source of coordinates in Mexico for collections participating in ORNIS, for which NSF funding was insufficient to georeference these localities using published best practices.
- 1,348,203 occurrence records were downloaded from ORNIS
- 266,568 distinct localities were extracted
- 71,235 localities already had geographic coordinates
None of the records already having coordinates were redetermined using best practices under ORNIS due to funding constraints at the outset. Before being sent to work centers for georeferencing, locations for all records were compared to locations georeferenced in MaNIS
(although HerpNet georeferences were not available at the time that some comparisons were made). By pre-processing in this way,
- 20,691 (10.6%) georeferences were reused for ORNIS
- 174,642 localities were georeferenced by ORNIS georeferencing work centers(listed on the Georeferencing page)
- 19,808 (10.2%) localities could not be georeferenced (too vague, inconsistent, not found)
Details of each work center's activities are given on the Checklist
Not all occurrence records in ORNIS were downloaded for georeferencing by the work centers. Records that were not accessible via the ORNIS portal at the time a region was claimed were not downloaded. Also, records from the large monitoring datasets of the Avian Knowledge Network were not downloaded, as these were already georeferenced.