Abstract

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.

Supplemental information

The Minimum Divergence version of the Schaake Shuffle (Scheuerer et al., 2017) is used for statistical ensemble member generation. The Quantile Regression Forest algorithm as described by Taillardat et al., 2016) is used as post-processing method.


Metadata

Dataset name QRF-RT-SSh
Dataset version v2025
Status onGoing
Last metadata update March 19, 2025, 09:13 (UTC)
Update frequency continual
License https://creativecommons.org/licenses/by/4.0/
North bound latitude 53.5655
East bound longitude 7.2634
South bound latitude 50.7575
West bound longitude 3.3914
Dataset edition 1
Dataset manager Harun Kivril
Maintainer KNMI Data Services
Publication timestamp 2025-04-15T00:00:00Z
Reference system identifier EPSG4326
Dataset start time 2025-04-15
Dataset end time Unlimited
Identifier urn:xkdc:ds:nl.knmi::QRF-RT-SSh/v2025/
Lineage statement Bi-linear interpolation from lambert to regular lat-lon
Purpose Calibrated and skilful gridded ensemble forecasts of summer precipitation in the Netherlands
Use limitation pre-operational, so no guaranteed delivery yet
Show more Show less

Resources