CDAWeb

Julia interface to NASA's CDAWeb RESTful services for accessing space physics data.

Quick Start

using Pkg; Pkg.add("CDAWeb")
using CDAWeb, Dates

ds = cda"OMNI_HRO_1MIN"
display(values(ds))         # data variables with descriptions, from the master CDF (data files may lack some); keys(ds) the names
t0, t1 = "2023-01-01", "2023-01-02"
var = cda"OMNI_HRO_1MIN/SYM_H"  # == `ds["SYM_H"]`
getmeta(var, "CATDESC")   # ISTP attributes: CATDESC, FIELDNAM, UNITS, FILLVAL, DEPEND_1

getdata(ds, t0, t1)  # the whole dataset
x = getdata(var, t0, t1)  # == `var(t0, t1)`: data in [t0, t1), a lazy CDFVariable
A = Array(x)              # FILLVAL and values outside VALIDMIN/VALIDMAX are NaN; time is the last dimension
t = times(x)
getdim(x, 1)              # values along dimension 1: its DEPEND_1 (e.g. a spectrogram's energies) or `axes(x, 1)`
getmeta(x, "VALIDMIN")    # attributes of the fetched files, which may differ from the master CDF's
BGSEc, Magnitude = getdata.(cda"AC_H0_MFI/BGSEc,Magnitude", t0, t1)

Files are cached in ~/.cdaweb/ (or $CDAWEB_DIR); clear_cache!() clears it.

Discovery

display(get_datasets("PSP", "MAG", "RTN"))   # find datasets: id, time range, label
get_variables(; dataset = "SOLO SWA PAS")   # every data variable of every matching dataset, with descriptions
get_variables("density"; dataset = "MMS1 FPI brst")

• Each space-separated word of the terms must match a whole word of the dataset's id, label, mission, spacecraft or instrument, case-insensitively: plural -s/-es allowed, a trailing * makes a prefix ("magnet*"). Separators inside a word match any separator ("SYM-H" finds SYM_H); a Regex matches as given.
• get_variables terms must all match one data variable's name or description (CATDESC), and dataset terms its dataset as above.
• Datasets whose mission or spacecraft match more terms come first ("Wind" lists WI_* before ACE's "Solar Wind Experiment"), single-mission before merged (OMNI), then by id.
• When nothing matches, an info line gives each term's matches alone.
• display shows a row per result, up to 80 with labels cut at 160 characters, variables under a line per dataset, then the matches per id prefix (spacecraft code) to narrow by; show(stdout, MIME"text/plain"(), x) shows all.
• getmeta(ds) is the dataset's CDAS entry: Id, Label, TimeInterval.Start/End, PiName, Notes, Doi, SpaseResourceId.
• CDAS metadata (the dataset list) is cached in ~/.cdaweb/metadata/ for a day.

References

• CDAS RESTful Web Services
• CDA_Webservice - speasy (Python)

Elsewhere

• speasy pursues a similar goal with support for multiple data sources including AMDA and CSA. This package, however, focuses on finer control over cached data with better performance and offers straightforward, direct access to those files allowing offline access and reproducibility (see speasy#237 and speasy#122).
• xhelio-cdaweb: NASA CDAWeb data access for heliophysics — MCP server + Python library