juliensimon/solar-wind

Real-Time Solar Wind

Part of the Space Weather Datasets collection on Hugging Face.

Real-time solar wind measurements from NOAA SWPC, combining plasma and magnetic field data from the DSCOVR and ACE spacecraft at the Sun-Earth L1 Lagrange point. Currently 24,700 minute-resolution readings spanning 2026-03-17 to 2026-04-03.

Dataset description

The solar wind is a continuous stream of charged particles flowing from the Sun. Its speed, density, and magnetic field orientation (especially Bz) are the primary drivers of geomagnetic storms. When Bz turns strongly southward (negative), it couples with Earth's magnetosphere and can trigger storms that affect satellites, power grids, and GPS.

This dataset is the missing link in the Sun-to-Earth causal chain: solar flare → CME → solar wind → Dst/Kp storm → orbital drag.

The measurements come from the DSCOVR (Deep Space Climate Observatory) and ACE (Advanced Composition Explorer) spacecraft orbiting the Sun-Earth L1 Lagrange point, approximately 1.5 million km upstream of Earth. At this vantage point, the instruments sample the solar wind roughly 15-60 minutes before it reaches the magnetopause, providing a critical lead time for geomagnetic storm prediction. DSCOVR's Faraday Cup measures the bulk plasma properties (proton density, speed, and temperature), while its fluxgate magnetometer measures the interplanetary magnetic field (IMF) vector in Geocentric Solar Magnetospheric (GSM) coordinates.

The Bz component of the IMF in GSM coordinates is the single most important parameter for geomagnetic coupling. When Bz is strongly southward (negative), the IMF opposes Earth's northward magnetic field at the dayside magnetopause, enabling magnetic reconnection that transfers solar wind energy into the magnetosphere. Sustained Bz below -10 nT typically produces moderate geomagnetic storms (Kp 6-7, Dst below -100 nT), while extreme events with Bz below -30 nT can trigger severe storms affecting power grids and satellite operations. The combination of solar wind speed, density, and southward Bz determines the rate of energy input, often parameterized by the epsilon coupling function or the Newell universal coupling function.

Typical quiet-time solar wind conditions show speeds of 300-450 km/s and densities of 3-10 protons/cm^3. Coronal hole high-speed streams elevate speeds to 600-800 km/s, while interplanetary CMEs can drive transient speeds above 1,000 km/s with enhanced magnetic fields. These signatures are directly usable for real-time Dst forecasting models such as the Burton equation and its successors.

Schema

Column	Type	Description
time_tag	datetime	Measurement time (UTC, ~1-minute cadence)
density	float	Proton density (particles/cm³)
speed	float	Bulk solar wind speed (km/s)
temperature	float	Proton temperature (K)
bt	float	Total magnetic field magnitude (nT)
bx_gsm	float	Magnetic field Bx in GSM coordinates (nT)
by_gsm	float	Magnetic field By in GSM coordinates (nT)
bz_gsm	float	Magnetic field Bz in GSM coordinates (nT) — key storm driver

Quick stats

• 24,700 readings (2026-03-17 to 2026-04-03)
• Average speed: 478 km/s, max: 858 km/s
• Minimum Bz: -27.9 nT (12,641 southward readings)

Usage

from datasets import load_dataset

ds = load_dataset("juliensimon/solar-wind", split="train")
df = ds.to_pandas()

# Solar wind speed time series
df.plot(x="time_tag", y="speed", title="Solar Wind Speed")

# Bz southward events (storm drivers)
southward = df[df["bz_gsm"] < -5]
print(f"{len(southward)} readings with Bz < -5 nT")

# Correlate with Dst index
# Join with juliensimon/dst-index on nearest hourly timestamp

Update frequency

Updated daily at 15:00 UTC via GitHub Actions. Each run fetches the latest 7-day rolling window from SWPC and appends new readings to the growing dataset.

Data source

NOAA Space Weather Prediction Center. Data from DSCOVR (primary) and ACE (backup) spacecraft at the Sun-Earth L1 point, 1.5 million km sunward of Earth (60 minutes ahead of arriving solar wind).

Related datasets

• dst-index — Hourly Dst geomagnetic storm index (driven by solar wind)
• donki-space-weather-events — CMEs, storms, shocks
• solar-flare-events — Individual flare detections
• space-weather-indices — Daily Kp, Ap, F10.7

Pipeline

Source code: juliensimon/space-datasets

Support

If you find this dataset useful, please give it a ❤️ on the dataset page and share feedback in the Community tab! Also consider giving a ⭐️ to the space-datasets repo.

Citation

@dataset{solar_wind,
  author = {Simon, Julien},
  title = {Real-Time Solar Wind (DSCOVR/ACE)},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/juliensimon/solar-wind},
  note = {Based on NOAA SWPC real-time solar wind data from DSCOVR and ACE}
}