Ensemble precipitation forecasts made with Quantile Regression Forests and deterministic Harmonie-Arome inputs
**Data Quality Warning:** Data released prior to 2025-07-10 is unreliable due to a configuration error. Please disregard affected data.
A gridded 51-member ensemble of precipitation forecasts that are created using a tree-based machine learning method, quantile regression forests (QRF), and inputs from the deterministic Harmonie-Arome (HA) Cy43 forecasts. The target data set is rain-gauge-adjusted radar data that is upscaled by taking 3x3 km means and then a rolling maximum is taken in a 9 x 9 km box. Inputs to the machine learning model include HA precipitation, and indices of atmospheric instability. Spatial and temporal dependencies are restored using the minimum divergence Schaake Shuffle (SSh). Hourly forecasts are issued 8 times per day (00, 03, 06, 09, 12, 15, 18 and 21 UTC) for 60-hours into the future.
Simple
- Date (Publication)
- 2025-04-15
- Edition
-
1
- Citation identifier
- urn:xkdc:ds:nl.knmi::QRF-RT-SSh/v2025/
- Purpose
-
Calibrated and skilful gridded ensemble forecasts of precipitation in the Netherlands
- Status
- On going
- Point of contact
-
Organisation name Individual name Electronic mail address Role Koninklijk Nederlands Meteorologisch Instituut (KNMI)
Point of contact
- Maintenance and update frequency
- Continual
-
DG Discovery Properties Vocabulary 1.0
-
-
Weather forecast
-
Precipitation
-
Machine learning
-
Quantile regression forests
-
- Use limitation
-
pre-operational, so no guaranteed delivery yet
- Access constraints
- Other restrictions
- Other constraints
- Classification
- Unclassified
- Spatial representation type
- Grid
- Distance
- 3000 meters
- Language
- English
- Character set
- UTF8
- Topic category
-
- Climatology, meteorology, atmosphere
))
- Begin date
- 2025-04-15
- End date
- 9999-12-31
- Supplemental Information
-
The Minimum Divergence version of the Schaake Shuffle ([Scheuerer et al., 2017]( https://doi.org/10.1002/2016WR020133)) is used for statistical ensemble member generation. The Quantile Regression Forest algorithm as described by [Taillardat et al., 2016]( https://doi.org/10.1175/MWR-D-15-0260.1)) is used as post-processing method.
- Reference system identifier
- http://www.opengis.net/def/crs/EPSG/0/4326
- Distribution format
-
Name Version NetCDF
4
- Distributor contact
-
Organisation name Individual name Electronic mail address Role Koninklijk Nederlands Meteorologisch Instituut (KNMI)
Point of contact
Digital transfer options
- OnLine resource
-
Protocol Linkage Name landingpage
https://dataplatform.knmi.nl/catalog/datasets/index.html?x-dataset=QRF-RT-SSh&x-dataset-version=v2025 KNMI Data Platform
- Hierarchy level
- Dataset
- Statement
-
Bi-linear interpolation from lambert to regular lat-lon
Metadata
- File identifier
- 82ba10ab-f2e4-40c0-91b8-f21454618134 XML
- Metadata language
- English
- Character set
- UTF8
- Hierarchy level
- Dataset
- Date stamp
- 2025-12-09T13:03:22.56468Z
- Metadata standard name
-
ISO 19115
- Metadata standard version
-
ISO 19115:2003 NL Kernset 1.3 KNMI 2.1.0
- Metadata author
-
Organisation name Individual name Electronic mail address Role Koninklijk Nederlands Meteorologisch Instituut (KNMI)
Point of contact
- Other language
-
Language Character encoding Dutch; Flemish English UTF8
Overviews
Spatial extent
)))
Provided by
