In nicholasjclark/BBS.occurrences: BBS community composition and co-occurrence analysis

knitr::opts_chunk$set(echo = TRUE)
library(knitr)
opts_chunk$set(tidy.opts=list(width.cutoff=50),tidy=TRUE)

This appendix describes how to extract landcover variables at 40km resolution for each BBS survey site. These variables can be accessed in the BBS.occurrences library using data('Site.descriptors').

If need be, install the MODISTools2 package. Please note that older versions of MODISTools may have problems with broken links! We have found it is better to remove old versions first and install the most up-to-date version

remove.packages('MODISTools')
install.packages("https://modis.ornl.gov/files/modissoapservice/MODISTools2.tar.gz",
                 repos = NULL, type = "source")

Or, if you prefer manual installation, you can manually downlaod MODISTools2.tar.gz here (https://modis.ornl.gov/data/modis_webservice.html)

install.packages("yourfilepath/MODISTools2.tar.gz", repos = NULL, type = "source")

Load the previously compiled BBS data from Appendix_S1

load("./BBS.Data.cleaned/BBS.2003-2009.filtered.Rdata")

Gather lats and longs for collecting MODIS landover data from each unique site

coordinates <- data.frame(lat = latlon[,2], long = latlon[,1])

Find unique sites

library(dplyr)
uniq.coordinates = coordinates %>%
  dplyr::group_by(lat, long) %>%
  dplyr::distinct()

Source functions needed for MODIS downloading and processing from NJ Clark's GitHub account

source("https://raw.githubusercontent.com/nicholasjclark/
       LandcoverMODIS/master/R/Landcover_functions.R")

Download data for the specified years (but don't include years 2010-2012, these are not accessible). This stores the raw downloaded data in a new LandCover folder within your working directory. Note that the input coordinates file must have colnames lat and long. Also note that landcover downloads struggle if we access more than ~100 points at a time (depending on your system). Downloading multiple subsets over time is safer

download_landcover(years_gather = c(2003:2009), 
                   coordinates = uniq.coordinates[1:100, ],
                   size = c(40,40))

Repeat the above function until all of the necessary files have been downloaded. Once ALL of the necessary files are downloaded, then summarise them and calculate proportions of landcover. But please note, you should not summarise before all necessary raw files are downloaded (raw files will get deleted once the function completes in order to save memory!). This function writes a .csv summary file (containing proportional cover for a range of habitat categories) in the LandCover folder and deletes the raw files. Note, these landcover variables can be directly accessed in the BBS.occurrences library using data('Site.descriptors').

summarise_landcover() 

nicholasjclark/BBS.occurrences documentation built on July 19, 2020, 8:31 p.m.