Query and download data from OB DAAC

obdaac-download.sh

#!/usr/bin/env bash

# Prerequisite:
#
# (1) Create a `~/.netrc` file (if it doesn't exist); e.g., `touch ~/.netrc`
#
# (2) Open the `~/.netrc` file in a plain text editor and add a single line like the following:
#
#   machine urs.earthdata.nasa.gov login your_edl_email@email.com password YourEDLPassword123
#
# Alternatively, if you want a configuration just for this script, you can put
# the netrc information in a local file (e.g., `./my_netrc`). Then, when
# calling `curl`, remove the `-n` argument and replace it with `--netrc-file ./my_netrc`.

COOKIEFILE=".urs_cookies"
LISTFILE="filelist"

STARTDATE="2023-05-01 00:00:00"
CURDATE=$(date +"%Y-%m-%d %H:%M:%S")

# Use system default netrc file
CURL="curl -nL -b $COOKIEFILE -c $COOKIEFILE"

# Use netrc file at a specific path
# CURL="curl --netrc-file ./my_netrc -L -b $COOKIEFILE -c $COOKIEFILE"

# Remove old credentials. They'll be reset in the next curl call.
# Authentication should happen through the `.netrc` file.
rm -f "$COOKIEFILE"

# Download list of files matching query
#   `-n` means to use authentication from the `.netrc` file.
#   `-L` means to follow redirects
#   `-w "\n"` at the end adds a trailing newline, to make it easier to loop over the file.

$CURL \
  -d "results_as_file=1" \
  -d "sensor_id=14" \
  -d "dtid=1019" \
  -d "sdate=$STARTDATE" \
  -d "edate=$CURDATE" \
  -d "subType=1" \
  -d "addurl=1" \
  -d "prod_id=chlor_a" \
  -d "resolution_id=4km" \
  -d "period=R32" "https://oceandata.sci.gsfc.nasa.gov/api/file_search" \
  -w "\n" > "$LISTFILE"

echo "Found $(cat $LISTFILE | wc -l) matching files."

while read -r url; do
  fname=$(basename "$url")
  if [ -f "$fname" ]; then
    echo "File $fname already exists. Skipping..."
    continue
  fi
  echo "Downloading URL: $url"
  $CURL -O "$url"
done < "$LISTFILE"

Note that line 39 (-d "dtid=1019") can (and probably should) be dropped from the query without affecting the results.