In BigelowLab/nam218grads: Provides grads-served opendap access to NAM218 forecast data

This package simplifies access to the NCEP NOMADS grads server for the NAM forecast models.

Requirements

• R v4+
• rlang
• dplyr
• sf
• stars
• ncdf4

Installation

remotes::install_github("BigelowLab/nam218grads")

knitr::opts_chunk$set(echo = TRUE)

Usage

Retrieve a URL for a resource

Retrieve a URL for a resource by date, like yesterday's forecast at 0000H. Note that the NCEP/NOMADS grads server only maintains the current forecast plus the prior 6.

library(stars)
library(nam218grads)
uri <- grads_uri(date = Sys.Date()-1, product = "nam_00z")
uri

Get a handle to a resource

It's just a ncdf4 object, but we wrapped common tasks in friendly functions.

x <- open_grads(uri)
# File http://nomads.ncep.noaa.gov:80/dods/nam/nam20210303/nam_00z (NC_FORMAT_CLASSIC):
# 
#      141 variables (excluding dimension variables):
#         float absvprs[lon,lat,lev,time]   
#             _FillValue: 9.99900026055401e+20
#             missing_value: 9.99900026055401e+20
#             long_name: ** (1000 850 700 500 250) absolute vorticity [1/s] 
#         float no4lftx180_0mb[lon,lat,time]   
#             _FillValue: 9.99900026055401e+20
#             missing_value: 9.99900026055401e+20
#             long_name: ** 180-0 mb above ground best (4 layer) lifted index [k] 
#      ...
#    4 global attributes:
#        title: NAM every 3 hours fcst staring from 00Z03mar2021, download Mar 03 02:38 UTC
#        Conventions: COARDS GrADS
#        dataType: Grid
#        history: Wed Mar 03 22:27:39 GMT 2021 : imported by GrADS Data Server 2.0

You can explore the contents of the resource.

times <- get_time(x)
times

levels <- get_lev(x)
levels

vars <- get_varnames(x)
vars

dims <- get_dims(x)
dims

Retrieve data

As an array

Retrieving data will return contiguous arrays of data. We provide a function get_var_array that expects the slab coordinates in integer run-length encoding for each dimension as [start_index, count]. This function returns an array of cell values - for range of pressure levels at different times.

a <- nam218grads::get_var_array(x, "tmpprs", 
                                list(lon = c(110, 100),
                                     lat = c(150, 75),
                                     lev = c(3, 5),
                                     time = c(1, 12)))
str(a)

As a stars object

The get_var function is a convenient wrapper around the get_var_array function. Users specify the slab limits with 'real world' units, and a stars object is returned (although returning an array is an option).

s <- nam218grads::get_var(x, "tmpprs",
                          bb = c(-77.0, -51.5, 37.9, 56.7),
                          time = times[1:12],
                          lev = levels[3:5])
s

You can also retrieve multiple variables. Note in the example below that ugrd10m and vgrd10m (wind component speeds at 10m above surface) do not have a lev dimension so we can ignore the lev argument.

winds <- nam218grads::get_var(x, c("ugrd10m", "vgrd10m"),
                          bb = c(-77.0, -51.5, 37.9, 56.7),
                          time = times[5:7])
winds

Close the handle

The resource is a ncdf4 object - be sure to close it when done.

close_grads(x)

Plot the levels at different times

Let's use the temperature at various pressure levels in variable s for some plotting. First time slice...

if (require(maps)){
  time_string <- format(times[1], "%Y-%m-%d %H:%M:%S")
  overlay_fun <- function() {
    map(col = 'red', add = TRUE)
    mtext(time_string, side = 1, line = -1)
  }
  plot(s, axes = TRUE, hook = overlay_fun)
}

And the last time slice...

if (require(maps)){
  time_string <- format(times[12], "%Y-%m-%d %H:%M:%S")
  plot(s[,,,,12], axes = TRUE, hook = overlay_fun)
}

BigelowLab/nam218grads documentation built on Oct. 4, 2022, 2:01 a.m.