Climate Resources Authors: Daniel J. McGlinn Description: This collection of files documents the processing of the Global Biodiversity Information Facility (GBIF) geographic data and the 5-arc minute resolution WorldClim Bioclim data to produce a species freezing exposure datafile: 'MinimumFreezingExposure.csv'. The data citations are also included as a text file. The Rscripts (i.e., files with the .R extenstion) are to be run in the R programming language and they require the following R packages are installed: 'sp','raster','rgdal','foreach','snow','snowfall','doSNOW'. The primary Rscript is called 'GBIF_run_all.R' and it documents the sequential order in which the scripts should be run in. The directory that this script is called from must contain the unpacked GBIF raw (.csv) datafiles in a directory called 'gbif_raw_data' and the working directory must also contain the files 'names-tr.txt','speciesTraitData.csv', and 'headers_commadelim_rerun.txt' which are not provided in this repository. Lastly the relevant GIS datalayers (also not provided) that are called must be present in the directory '~/gis/' Records from the GBIF dataset were filtered using the following criteria: 1. The scientific names were in a reasonable format of a Latin binomial or trinomial (e.g., only letters). 2. The record had numeric latitude and longitude in decimal degrees where the latitude was between -90° and 90° and the longitude was between -180° and 180°, and neither coordinate was exactly equal to zero, which is often used in these databases as a placeholder for lack of information. 3. The record’s latitude was not equal to longitude, as this would most likely be indicative of a data entry error. 4. The record was not a duplicate record according to the GBIF “occurrence_id” field. 5. The record was not located within 50 km of the GBIF headquarters in Copenhagen, Denmark (55.68°N, 12.59°E), to minimize the chance that a record was given a coordinate that corresponded to where the data were housed but not where the plant was actually collected. 6. The record contained a valid entry in the GBIF field “country_interpreted”. This country information was independently checked against a global country shapefile. The spatial coordinate of the record had to match the record in the “country_interpreted” field or at the minimum match the continent with which that “country_interpreted” field was associated. These comparisons provided a benchmark of reasonable geospatial accuracy that the record had to possess and ensured the record was from a terrestrial location. The scripts will create the following directories if they do not exist: 'gbif_data_chunks','filtered_results','genus_sort', & 'genus_results'. The output of the script 'GBIF_export_freezing_exposure', is a datafile, 'MinimumFreezingExposure.csv', which contains two columns: 'Species' : the scientific binomial 'Freeze.tmin.lo': a text string which is either 'FreezingUnexposed' or 'FreeezingExposed'. 'FreezingExposed' denotes species that were exposed to freezing temperatures over some portion of their range. This variable was derived using the documented geographic distribution of the species in the GBIF dataset combined with the WorldClim Bioclim variable BIO6 (= Min Temperature of Coldest Month). Specifically, if the 2.5% lower quantile of BIO6 across a species occurrances was less than or equal to 0 degrees C then the species was characterized as 'FreezingExposed' otherwise the species was characterized as 'FreezingUnexposed'. Files: GBIF_run_all.R GBIF_make_data_chunks.R GBIF_make_splist.R GBIF_setup_geog_data.R GBIF_geog_filter.R GBIF_climate_query.R GBIF_climate_summary.R GBIF_export_all_records.R GBIF_package_summaries.R GBIF_output_freezing_exposure.R MinimumFreezingExposure.csv citations.txt References: Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones and A. Jarvis, 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25: 1965-1978. Global Biodiversity Information Facility (GBIF). http://www.gbif.org/