WeatherDataAI Order data

Daily weather for anywhere on land, back to 1940.

Pick a location or a set of countries, choose your variables and years, and pay. Clean CSV files land in your inbox. No app to learn, no NetCDF to unpack.

250,000+land grid points
31,000+days of history
21daily variables
Click anywhere on land to drop a point.
—
Land grid Your selection Map drawn at 0.5°. Data is delivered at 0.25° (~28 km).

Data

One consistent global grid, every day since 1940.

Observations from government agencies, private networks and Earth observation platforms are assimilated into numerical weather prediction models, then refined with machine learning. The output is a single gridded record you can model against without stitching stations together.

Resolution0.25°About 28 km (17 mi) grid spacing
Period1940 → nowDaily values, current year to order date minus 7 days
CoverageAll landEvery country, 250,000+ grid points
FormatCSVMetric or imperial, opens in Excel, R or pandas

Variables

21 daily variables, with departures and 30-year normals.

Daily values run from 1940 to the present. Departures give each day's difference from normal at that grid point. Climatology is the 30-year daily normal: one 365-day series per variable, ordered once with no year range.

Variable IDNameMetricImperialDepartureClimatology
WM-104218_average_temperature_2023.csvMulti-location sample
Point_ID,Country,Continent,Region,latitude,longitude,2023-01-01,2023-01-02,2023-01-03
112840,Japan,Asia,East Asia,35.625,139.625,5.2,6.1,4.8
112841,Japan,Asia,East Asia,35.625,139.875,5.6,6.4,5.0
112842,Japan,Asia,East Asia,35.875,139.375,3.9,4.7,3.6
112843,Japan,Asia,East Asia,35.875,139.625,4.4,5.3,4.1
…

Country orders arrive as one file per variable per year, one row per grid point. Single-location orders arrive as one file with every variable and year.

Delivery

Configure, pay, download.

Pay once, no account to create. We prepare your files and email them to you.

Step 1

Configure

Choose a point or countries, the variables you need, a year range and units. The price updates as you go.

Step 2

Pay with Stripe

Card, Apple Pay or Google Pay on Stripe's hosted checkout. Your receipt arrives by email.

Step 3

Get your CSVs

We email your CSV files, or a download link for large orders, within one business day.

API

Need it in your code? Use the API.

Historical observations and 00z daily forecasts by variable, date range and coordinate. Each day of data counts as one call. Overage is $0.01 per day of data.

# pip install py_weatherdataai
from py_weatherdataai.client import WeatherDataAI

client = WeatherDataAI("YOUR_API_KEY")
df = client.get_observation("2T", "2024-02-15", "2024-02-20", 41.586, -93.625, "Metric")

Daily updates to your bucket

Subscribe to fresh observations for your variables and regions, pushed to AWS S3, Google Cloud Storage or Dropbox as soon as the databank updates.

support@weatherdata.ai

FAQ

Questions

Anything else, email support@weatherdata.ai with your order number.

How is the price calculated?

A single location costs $5 per series for any date range, with a $10 minimum. A series is one variable as daily values, departures or climatology. Country orders cost $0.50 per 1,000 grid points per series per year, and a climatology counts as one year. The minimum is $25. The total is always shown before you pay.

How recent is the data?

The current year updates daily. Models are rerun as late observations arrive, so ordering the current year gives you data through roughly 7 to 10 days before your order date.

Can I use the data commercially?

Yes. The data is structured for commercial analysis, modeling and decision-making. It is provided as is; see the disclaimer for terms.

My order is too large for a card.

Orders over $25,000 are invoiced. The checkout button switches to an invoice request automatically, or email support and we will arrange payment by wire.

Does it cover the ocean?

No. WeatherDataAI covers land only. A point near the coast snaps to the nearest land grid cell.