KNMI provides precipitation forecast data. This package provides an interface to download and work a bit easier with the Nowcast precipitation forecast data from KNMI.
Nowcast precipitation forecast up to 2 hours ahead, per 5 minutes, over the Netherlands. The forecast contains 25 time steps: +0 minutes to +120 minutes. Forecasted is the precipitation sum per 5 minutes, on a grid of 1x1 km. The forecast is made with an operational KNMI implementation of pySTEPS (https://pysteps.github.io). The forecast is initiated with the KNMI 5-minute real-time precipitation accumulation product: RTCOR-5m.
More information on the dataset here
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
# To run the notebooks install requirements-dev.txt aswell
pip install -r requirements-dev.txtTo use the KNMI data platform you need an API key. You can request one here
See the notebooks folder for examples.
Download the latest available h5 file from the KNMI data platform.
from knmi_precipitation.download import OpenDataAPI
api_key = "your-api-key"
oda = OpenDataAPI(api_key)
latest_file = oda.download_latest_file('./data/test.h5')Load a .h5 file and get the forecast data in mm/hr and dBZ at a specific location.
from knmi_precipitation.h5_radar_data import H5RadarData
data = H5RadarData('./example_data/RAD_NL25_RAC_FM_202409041355.h5')
lng, lat = 5.319185, 51.687406
value_mm_hr = data.get_mm_hr_from_lon_lat('image1', lng, lat)Export the forecast data of image1 from a .h5 file to a GeoTIFF file warped to EPSG:4326.
from knmi_precipitation.h5_radar_data import H5RadarData
radar = H5RadarData('./example_data/RAD_NL25_RAC_FM_202409041355.h5')
data.export_geotiff('image1', './data/RAD_NL25_RAC_FM_202409041355_1.tif')