project_velocity Projecting shifts in habitat for 686 North American marine species under RCP 2.6 and 8.5.
Publication: Morley JW, Selden RL, Latour RJ, Frolicher TL, Seagraves RJ, Pinsky ML (2018) Projecting shifts in thermal habitat for 686 species on the North American continental shelf. PLoS ONE 13(5): e0196127. https://doi.org/10.1371/journal.pone.0196127
Contact: James Morley University of North Carolina, Chapel Hill jwmorley@email.unc.edu
Malin Pinsky Rutgers University malin.pinsky@rutgers.edu
Contains Rdata files used in the R-scripts. The raw temperature projection data in the sub-directory ‘Temp proj files’ and the formatted temperature projection files in the subdirectory ‘prediction_files_Nov2017’ are too large to store on GitHub, however, and are stored on the lab server in the shared data_largefiles: folder in a directory called projections_PlosOne2018.
Additional large files can be found stored on the lab's Box account.
Contains the figures from the projections project.
An alternative method to combine all the trawl survey data with the ‘trawlData’ package. This is the script that was ultimately used for the manuscript.
Fill in missing SST values for a portion of the trawl hauls, using satellite temperature observations from the Hadley NOAA/OI
Standardizes species names across surveys based on some files that have species name corrections. Also corrects some species names. Drops species/life stages that are not of interest (e.g., larval fish). Picks the species that are ultimately included in the analysis based on frequency of occurence.
Develops the projection grid based on the resolution of the SODA climate data. This projection grid was used to conduct the future climate projection and to integrate with the bathymetry environmental data in ‘Bathymetry.R’.
Depth data is brought in from GEBCO database and a depth data grid is developed and rugosity is calculated based on depth. Then sediment data is brought in from a data set that was put together elsewhere (contact for that is Becca Selden). Finally, the projection grid was developed based on the SODA climate resolution, in other words, we determined where these three data sources all overlap and are shallower than 400 m. Overall, this script is not cleaned up well.
Merge together a number of files containing the environmental data for hauls, which creates a master file for the trawl-hauls.
Fits the niche model to each species and calculates some sdm model evaluation statistics.
Drop species with poorly fit habitat models and then reformat the region value for conducting projections. Then reads in and formats all the climate projection data into a data structure that can be used to predict with the habitat models. Also produces some figures for the climate projection and some code to qaqc the climate data.
Conducts projections based on the delta biomass GAM model. The script integrates the habitat models with the climate projection models and then saves the aggregated (20 year time bins) output. This script is activated by a shell script, which was used on a CPU cluster.
Summarizes the projection output and makes the figures used in the Plos One (2018) manuscript.