GeoDataAccess.jl

GeoDataAccess.jl is a Julia package for collecting geospatial data from multiple public APIs. It provides a unified interface across weather, hydrology, seismology, air quality, and more — with built-in caching and GeoInterface.jl integration for flexible spatial queries.

Data Sources

Source	Type	Coverage	Resolution	API Key
`OpenMeteoArchive`	Raster (ERA5)	Global, 1940–present	Hourly/Daily, 25 km	No
`OpenMeteoForecast`	Raster	Global, 16-day forecast	Hourly/Daily, 9 km	No
`NOAANCEI`	Station-based	Global stations, 1763–present	Daily	No
`NASAPower`	Raster	Global, 1981–present	Daily, 55 km	No
`TomorrowIO`	Raster	Global, 2000–present	Hourly/Daily, 4 km	Yes
`VisualCrossing`	Raster	Global, ~50 years	Daily/Hourly, 1 km	Yes
`USGSEarthquake`	Event-based	Global catalog	Event-based	No
`USGSWaterServices`	Station-based	US, 1.5M+ sites	Daily/15-min	No
`OpenAQ`	Station-based	Global, 11K+ stations	Hourly/Daily	Yes
`NASAFIRMS`	Event-based	Global, near real-time	375 m / 1 km	Yes
`EPAAQS`	Station-based	US stations	Hourly/Daily	Yes
`LandfireSource`	Raster	CONUS, AK, HI	30 m	No (email)
`NOAAGFS`	Raster (GRIB2)	Global, forecast	0.25° (~25 km)	No
`ERA5`	Raster (NetCDF/GRIB)	Global, 1940–present	0.25° (~25 km)	Yes

Installation

using Pkg
Pkg.add("GeoDataAccess")

Quickstart

The core workflow is: create a DataAccessPlan, inspect it, then fetch to download the data as JSON files.

using GeoDataAccess
using GeoDataAccess: DataAccessPlan, fetch, OpenMeteoArchive
using Dates

# Create a plan: hourly temperature and precipitation for NYC
plan = DataAccessPlan(OpenMeteoArchive(), (-74.0, 40.7),
    Date(2024, 7, 1), Date(2024, 7, 3);
    variables = [:temperature_2m, :precipitation],
    frequency = :hourly)

DataAccessPlan for openmeteoarchive
  Extent:    Point(40.7, -74.0)
  Time:      2024-07-01 to 2024-07-03 (3 days)
  Variables: temperature_2m, precipitation
  frequency: hourly
  timezone: GMT
  API calls: 1
  Est. size: 1.125 KiB

  Request 1: GET 1 point(s), hourly 3 days → openmeteoarchive/3669ebf219e4f089.json

# Execute the plan — returns file paths to cached JSON
files = fetch(plan)

1-element Vector{String}:
 "/home/runner/.julia/scratchspac" ⋯ 68 bytes ⋯ "oarchive/3669ebf219e4f089.json"

# Read the JSON data
using JSON
data = JSON.parsefile(files[1])
keys(data)

KeySet for a JSON.Object{String, Any} with 9 entries. Keys:
  "latitude"
  "longitude"
  "generationtime_ms"
  "utc_offset_seconds"
  "timezone"
  "timezone_abbreviation"
  "elevation"
  "hourly_units"
  "hourly"

Or skip the plan and call fetch directly on a source:

using GeoDataAccess: fetch, OpenMeteoArchive

files = fetch(OpenMeteoArchive(), (-74.0, 40.7),
    Date(2024, 7, 1), Date(2024, 7, 3);
    variables = [:temperature_2m, :precipitation],
    frequency = :hourly)

1-element Vector{String}:
 "/home/runner/.julia/scratchspac" ⋯ 68 bytes ⋯ "oarchive/3669ebf219e4f089.json"