The {nert} package streamlines access to various Cloud-Optimised GeoTIFF (COG) datasets provided by the Australian Terrestrial Ecosystem Research Network (TERN), allowing you to easily incorporate these environmental datasets into your R analytics. Currently supported datasets in the {nert} package include:
{nert} is available through the R-Universe with pre-built binaries (this is the easy way).
To get started:
options(
repos = c(
AAGI = "https://aagi-aus.r-universe.dev",
CRAN = "https://cloud.r-project.org"
)
)
install.packages("nert")
Note that for Linux users, you will need to install system libraries to
support geospatial packages in R, e.g., {sf} and {terra} as well as
some packages for downloading data via
curl, please see Note for Linux
Installers. If you run into errors, e.g.,
Bad GitHub Credentials, please read this: Managing Git(Hub)
Credentials
and set up your GitHub credentials in R and try again.
You can install the development version of {nert} from GitHub with:
o <- options() # store original options
options(pkg.build_vignettes = TRUE)
if (!require("pak")) {
install.packages(
"pak",
repos = sprintf(
"https://r-lib.github.io/p/pak/stable/%s/%s/%s",
.Platform$pkgType,
R.Version()$os,
R.Version()$arch
)
)
}
pak::pak("AAGI-AUS/nert")
options(o) # reset options
COGs are not like working with other datasets in R that you might fetch
from the web. They only fetch the information about the object for you
in your R session and only at the time that you call it, e.g.,
plot() or writeRaster().
A normal GeoTIFF is like downloading an entire 500-page book because you want one paragraph.
A COG is like having an index that lets you jump directly to page 347 and read only that page.
COGs allow you to very quickly create a SpatRaster object locally that
represents the entire dataset and then work with a sub-portion from
there. Be aware that if you are not careful, you could end up
downloading a large GeoTIFF file fully. It is best to create the local
object and then crop(), mask() or extract() the area of interest
to save time, bandwidth and free memory. Then once you’ve done that, try
plotting or using the object in models or calculations. Now, on to
working with {nert} to read COGs from TERN!
The {nert} package provides a number of convenient functions such as
read_smips() which allow you to fetch COGs from the TERN Data Portal
for use in your R session. The below code fetches the SMIPS soil
moisture raster for the 1st of January 2024, and then uses the {terra}
package’s extract() function to get a soil moisture estimate for the
Adelaide CBD. Note that since these are cloud-optimised COGs, they
employ Just-In-Time data streaming: that is, only the bytes necessary
for the spatial extent we actually need are downloaded, resulting in
significant time and disk-space savings.
library(nert)
library(terra)
#> terra 1.9.46
r <- read_smips(date = "2024-01-01")
#> Warning: nert has set the GDAL option `GDAL_HTTP_USERPWD` to your TERN API key for the
#> rest of this R session.
#> ℹ GDAL sends the key with every `/vsicurl` request in this session, including
#> requests to servers other than TERN.
#> ℹ Restart R to clear it.
extract(r, xy = TRUE, data.frame(lon = 138.6007, lat = -34.9285))
#> ID smips_totalbucket_mm_20240101 x y
#> 1 1 46.07692 138.6037 -34.93254
Plot the Australia-wide rasters with plot().
plot(r)
The {nert} package also provides a convenient function
collect_tern_data() which allows you to easily grab multiple TERN
datasets across multiple times and multiple spatial locations. For
example, the below code grabs the SMIPS “totalbucket” and SLGA “CLY”
clay content estimates at 0-5cm and 5-15cm depths, across the first week
of 2024 at Merriden, WA and Tamworth, NSW:
dat <- collect_tern_data(
xy = data.frame(lon = c(118.28, 150.84), lat = c(-31.48, -31.07)),
date_range = c("2024-01-01", "2024-01-07"),
datasets = c("SMIPS", "CLY"),
smips_collection = "totalbucket",
depth = c("000_005", "005_015"),
stat = "EV",
verbose = FALSE
)
head(dat)
#> date lon lat SMIPS_totalbucket CLY_EV_000_005 CLY_EV_005_015
#> <Date> <num> <num> <num> <num> <num>
#> 1: 2024-01-01 118.28 -31.48 0.2271653 11 15
#> 2: 2024-01-01 150.84 -31.07 93.4498901 28 31
#> 3: 2024-01-02 118.28 -31.48 0.2135102 11 15
#> 4: 2024-01-02 150.84 -31.07 90.6175690 28 31
#> 5: 2024-01-03 118.28 -31.48 0.2344290 11 15
#> 6: 2024-01-03 150.84 -31.07 88.5260391 28 31
{nert} is undergoing active development and is not yet on CRAN. If you installed {nert} using the R-Universe (the preferred method), you can keep {nert} up-to-date locally like so:
update.packages()
and answering yes or y when asked if you would like to upgrade
{nert}.
If you are using Linux, you will likely need to install several system-level libraries, {pak} will do it’s best to install most of them but some may not be installable this way. For Nectar with a fresh Ubuntu image, you can use the following command to install system libraries to install {pak} and then install {fifo}. In your Linux terminal (not your R console, the “terminal” tab in RStudio should do here in most cases) type:
sudo apt update && sudo apt install libcurl4-openssl-dev libgdal-dev gdal-bin libgeos-dev libproj-dev libsqlite3-dev libudunits2-dev libxml2-dev
Set up the container.
devcontainer up --workspace-folder .
Run tests and check stuff.
devcontainer exec --workspace-folder . R -e "devtools::check()"
Render this file.
devcontainer exec --workspace-folder . R -e "devtools::build_readme()"
To cite nert:
citation("nert")
#> To cite package 'nert' in publications use:
#>
#> Sparks AH, Pipattungsakul W, Edson R, Rogers S, Moldovan M (2026).
#> nert: Curated Access to TERN Environmental Raster Data. R package
#> version 1.1.0. https://aagi-aus.github.io/nert/
#>
#> A BibTeX entry for LaTeX users is
#>
#> @Manual{,
#> title = {{nert}: Curated Access to TERN Environmental Raster Data},
#> author = {Adam H. Sparks and Wasin Pipattungsakul and Russell Edson and Sam Rogers and Max Moldovan},
#> year = {2026},
#> note = {R package version 1.1.0},
#> url = {https://aagi-aus.github.io/nert/},
#> }