Data access

Data are available at the AWS portal: https://registry.opendata.aws/smn-ar-wrf/.

Downloading can be carried out in different ways:

URL adress
Files can be download by simply going to the following link: https://smn-ar-wrf.s3-us-west-2.amazonaws.com/index.html.
Data are stored using the Amazon Simple Storage Service (S3). For more information about this tool visit https://docs.aws.amazon.com/es_es/AmazonS3/latest/userguide/Welcome.html.

AWS CLI
This procedure makes use of routines proper to the tool AWS Command Line Interface (CLI). For more information regarding their use visit the following link.
Here, we present two examples, to download an individual file and to download files from an entire directory:

#!/usr/bin/env bash

# to download the file with the second forecast hour from the 00 UTC cycle for March 21, 2022 to the output_path: 
aws s3 cp --no-sign-request s3://smn-ar-wrf/DATA/WRF/DET/2022/03/21/00/WRFDETAR_01H_20220321_00_002.nc output_path

# to download all forecast times for the 00 UTC cycle for March 21, 2022 to the output_path:
aws s3 cp --no-sign-request s3://smn-ar-wrf/DATA/WRF/DET/2022/03/21/00/ --recursive output_path

Python
This can be done using the library s3sf.
For example, to download an individual file we may write:

import s3fs
# to download the file with the second forecast hour from the 00 UTC cycle for March 21, 2022
s3_file = 's3://smn-ar-wrf/DATA/WRF/DET/2022/03/21/00/WRFDETAR_01H_20220321_00_002.nc'
fs = s3fs.S3FileSystem(anon=True)
data = fs.get(s3_file)

R
This can be done using the library aws.s3.
For example, to download all the files for 1 day:

library("aws.s3")
 
# Define the function wrf.download to download the files
wrf.download <- function(wrf.name = wrf.name){
      save_object(
      object = paste0(wrf.name),
      bucket = "s3://smn-ar-wrf/",
      region = "us-west-2",
      file = substring(wrf.name, 28),
      overwrite = TRUE)}

# Define the date of the data
anual = 2022
mes = 9
dia = 3
ciclo = 0
time = "01H"   # Frequency of the forecast of interest 01H or 24H (string format) 

# The input date is converted to string
anual <- sprintf("%04d", anual)
mes <- sprintf("%02d", mes)
dia <- sprintf("%02d", dia)
ciclo <- sprintf("%02d", ciclo)
 
# The names of the files are defined
wrf.names <- get_bucket_df(
    bucket = "s3://smn-ar-wrf/",
    prefix = paste0("DATA/WRF/DET/", anual, "/", mes, "/", dia, "/", ciclo),
    max = Inf,
    region = "us-west-2")
 
wrf.names.rows <- which(grepl(time, wrf.names$Key, fixed = TRUE) == TRUE)
wrf.names <- wrf.names[wrf.names.rows, ]
 
# We run the función wrf.download 
lapply(wrf.names$Key, FUN = wrf.download)