Import a wind vector file generated from Python (e.g. via xarray or netCDF4). The engine automatically accommodates standard grid matrices, xarray dictionaries, or vector arrays.
Click to select or drag and drop your Python-exported JSON file
Accepts .json and .geojson exported from Python NetCDF output
The bundled converter uses the first available vertical level and writes one wind field per hour. The map draws a tapered directional trail for each vector; the trail is visual only and does not use earlier hourly frames.
Minimal Inline Python Snippet (10 m only)
import xarray as xr, json, numpy as np
# Open your NetCDF file
ds = xr.open_dataset("wind.nc")
u, v = ds["u10"].values, ds["v10"].values
lats, lons = ds["latitude"].values, ds["longitude"].values
vectors = []
for i, lat in enumerate(lats[::5]):
for j, lon in enumerate(lons[::5]):
u_val, v_val = float(u[0, i*5, j*5]), float(v[0, i*5, j*5])
speed = float(np.sqrt(u_val**2 + v_val**2))
direction = float((np.arctan2(-u_val, -v_val) * 180 / np.pi + 360) % 360)
vectors.append({"lat": round(float(lat), 2), "lng": round(float(lon), 2),
"u": round(u_val, 2), "v": round(v_val, 2),
"speed": round(speed, 2), "direction": round(direction, 1)})
with open("public/data/wind/wind_custom.json", "w") as f:
json.dump({"timestamp": "2026-10-05T12:00:00Z", "vectors": vectors}, f)
The built-in parser automatically accommodates any of the following Python data structures: