High-Accuracy
Site-level NWP Forecast Data
Numerical weather forecast data for wind and solar power forecasting: 15-day horizon, 15-minute resolution, 19 weather elements. Provide your site coordinates and receive grid-interpolated single-point forecast series, delivered in full as CSV.
Higher Horizon & Resolution for Power Forecasting
Full delivery per model run, covering every core element required by wind farms and PV plants
360-hour Horizon
Each model run covers +0h to +360h, 96 points/day — 1441 time points delivered in full, no incremental patches; new runs fully replace old ones.
Site-level Interpolation
Simply provide site coordinates — the NWP grid is interpolated to a single-point series, with no need to fetch grid cells or compute interpolation yourself. Managed per site ID (farm_id).
19 Elements, Built for Wind & Solar
10m / 100m (hub-height) wind speed and direction, temperature, pressure, humidity, three-layer + total cloud cover, four solar radiation components — ready for power forecasting models.
Rigorous Quality Control
Strictly regular 15-minute intervals with no gaps or duplicates; sample verified 1441 points with zero missing values; radiation components satisfy internal consistency checks.
Data Delivery Format
One CSV file per site per model run: a fixed 19-column header, one data point every 15 minutes, sorted in ascending time order
| date_time | spd100 | t2m | rh2 | tcc | tr |
|---|---|---|---|---|---|
| 2026-08-27_20:00 | 2.83 | 289.08 | 95.9 | 1.0 | 0.0 |
| 2026-08-27_20:15 | 3.19 | 288.82 | 96.5 | 1.0 | 0.001 |
| 2026-08-27_20:30 | 3.29 | 288.71 | 96.8 | 1.0 | 0.004 |
| 2026-08-27_20:45 | 3.30 | 288.58 | 97.0 | 1.0 | 0.008 |
| 2026-08-27_21:00 | 3.16 | 288.36 | 97.2 | 1.0 | 0.012 |
See the Field Dictionary for all 19 columns; unit conversions (K → ℃, Pa → hPa, radiation accumulation differencing) are covered in Usage Notes.
# Unit conversion and irradiance differencing in three lines import pandas as pd df = pd.read_csv("weatherforecast_102.95839-36.35208_2026082712.csv") df["t2m_c"] = df["t2m"] - 273.15 # K → ℃ df["ghi"] = df["tr"].diff() / 900 # W/m² df[["date_time", "spd100", "t2m_c", "ghi"]].head()
Industry Solutions
From wind speed to irradiance — the data foundation for renewable power forecasting and weather alerting
Wind Power Forecasting
Hub-height wind speed directly determines power output, yet ordinary weather data only offers 10m wind. We provide 100m (near-hub-height) wind speed and direction, plus temperature and pressure for air-density correction.
Solar Power Forecasting
Irradiance is the direct driver of PV output. Differencing the four radiation components at 15-minute steps yields per-interval irradiance, and cloud cover helps separate sunny / cloudy generation scenarios.
Weather Service Alerting
Gale, temperature-drop and low-cloud alerts need both long horizon and high resolution. 15 days × 15-minute granularity covers both early warning lead time and nowcasting corrections.
Wind Resource & Site Assessment
Accumulating successive model runs builds a long site-level series, supporting statistical assessment of wind and solar resources and turbine-selection cross-checks.
Data Quality Control
Grid-point interpolation delivered per site — every metric traceable to the verified sample
15-minute steps with no gaps or duplicates; 1441 time points including both endpoints.
All 19 columns of the sample are complete; missing values are defined as empty strings — transparent rules.
Radiation components satisfy observational relations (rr ≈ 0.216 × tr, dr + sr ≈ 1.21 × tr).
Full delivery per model run; new runs fully replace old data — no mixing of issue times.
Four Steps to Integrate
From sample evaluation to production, with technical support throughout
Provide site coordinates and your use case, receive real sample data
Get your farm_id, confirm model-run schedule and element scope
CSV file push, or HTTP API pull (spec v1.0)
Apply unit conversions per the Field Dictionary, verify, and launch
Request Free Sample Data
A real 15-day site forecast CSV — 1441 rows × 19 columns — to evaluate data quality and business fit