Swift Inventory of Low-Level Screaming Parties in the UK, 2009-2014

Quality Control

The public has submitted many of these data. Although the level of misidentified records may be higher than in datasets collected exclusively by trained field-workers, this is mitigated by the large sample size. These data have been mapped and checked for sensitivities and typographical/ geographical errors. Data has been checked using the NBN Record Cleaner

Method steps

1. The Swift Inventory collates data about low-level screaming swifts from as many sources as possible. In 2009, members of the public were asked to complete a form to report screaming parties of swifts, flying at rooftop level. Records collected in 2009, and in previous years, were provided by swift conservation groups (Concern for Swifts - Scotland, London's Swifts, Northern Ireland Swifts, Swift Conservation and UK Swifts). Data were also drawn from previous RSPB surveys (swift survey in 2007, and Homes for Wildlife survey in 2008). The data source is included as an attribute in the dataset. The Inventory will be updated annually. In 2010, an interactive online form was used for reporting records. Additional records were contributed by Kirtlington Wildlife and Conservation Society, Cambridgeshire Bird Club, Dorset Wildlife Trust, Bristol Naturalists Society, Glasgow Museums BRC, SOC-Highland and Perth and Kinross Council, Ludlow 21. The Count attribute is reported as the approximate count of swifts. It is difficult to get an exact count of the number of low-flying swifts, because of their speed, and the likely proximity to buildings, which obscure the swifts from view. When a count range was reported, the mean value is included in the dataset. The count attribute may refer to either, the approximate number seen at a site on one occasion, or the approximate number seen over numerous occasions.

Coverage is UK wide. There may be some bias towards more densely populated areas. Many records were provided without adequate geographic information to enable them to be mapped. Where possible, a postcode was assigned to these records using Royal Mail's Postcode Finder. All records with valid postcodes were assigned a grid reference corresponding to the centroid of the postcode area. For streets spanning multiple postcode areas, the grid reference corresponds to the centroid of a central postcode on the street. Where it was not possible to assign a postcode to a location, Google, Google Maps or Ordinance Survey maps were used to assign latitude and longitude or a grid reference. Records with latitude and longitude were mapped at 1 km resolution. In some cases, grid references were assigned by locating villages on maps. The resolution of these records ranged from 1 - 10 km. More detail about the format of the grid reference is provided in the dataset.