Skip to contents

Access satellite data through ERDDAP

Source: R/erddap.R
accessERDDAP.Rd

Downloads a subset of an ERDDAP gridded dataset and returns it as an sf point object with one row per grid cell and time step — the same shape the other access functions return, so matchData() joins it unchanged.

Usage

accessERDDAP(
  vars,
  years = NULL,
  months = NULL,
  bounding_box,
  dates = NULL,
  dataset = "MUR",
  overwrite = FALSE
)

Arguments

vars

variables to read, from the dataset's own catalog. MUR offers SST, SST_ERROR and ICE; the VIIRS entries offer CHL.

years

years to read. Required unless dates is given.

months

months to read. Required unless dates is given.

bounding_box

named list with xmin, xmax, ymin, ymax, or an sf/sfc object. Longitudes negative west, as elsewhere.

dates

the exact dates to read, as YYYYMMDD strings, YYYY-MM-DD strings, or Date objects

dataset

which dataset to read: a name from erddap_datasets(), or a spec from erddap_dataset() describing any other

overwrite

re-read days already cached

Value

one row per grid cell per day, with YEAR, MONTH, DAY and a column per requested variable

Why this one needs no account

MUR and the VIIRS products are also at PO.DAAC, where they need an Earthdata login. ERDDAP serves them openly and subsets server-side, so a bounding box costs a small download rather than a global file. That is the whole reason this route was chosen: no credentials for this package to handle, and none for you to configure.

Satellite SST is not model SST

MUR measures the foundation temperature — below the daily warming layer — where a model's SST is its topmost level, and satellite chlorophyll is an optical retrieval where a model's is a state variable. Both land in columns called SST and CHL, which is what makes them drop into an existing pipeline, and is exactly why matchData() records <var>_source.

Two further cautions specific to satellites. MUR is gap-free by construction — it is an analysis, so cloud is interpolated over rather than left as NA, and SST_ERROR is where the uncertainty of that shows up; fetch it alongside if the interpolation matters. And VIIRSCHL is DINEOF gap-filled, so its holes are filled too; VIIRSCHL2018 is the raw retrieval and is gappy under cloud, which fill_satellite_gaps() is for.

See also

erddap_datasets() for what ships, erddap_dataset() for anything else on an ERDDAP server

Examples

if (FALSE) { # \dontrun{
bb <- list(xmin = -70, xmax = -66, ymin = 41, ymax = 44)

# MUR SST at 0.01 degrees - far finer than any model here
sst <- accessERDDAP(vars = c("SST", "SST_ERROR"),
                    dates = unique(observations$date), bounding_box = bb)

matched <- matchData(observations, sst)

# VIIRS chlorophyll
chl <- accessERDDAP(vars = "CHL", years = 2022, months = 6,
                    bounding_box = bb, dataset = "VIIRSCHL")
} # }