We provide in this database median maps at 1 km resolution of the hourly accumulated precipitation over Belgium from 1940 to 2016. We used the analog technique to compute for every day in the past (1940 – 2016) the 25 best analogs selected from the high resolution RMI RADCLIM radar database that is available from 2017 to 2022. The analogs were determined based on ERA5 data from ECMWF. The median was computed based on this 25 analogs. Please read the LICENSE file for more information on the data licenses. Citation If you use this dataset for a publication, please cite the dataset article: Debrie, E., Demaeyer, J., and Vannitsem, S.: Hourly precipitation series over Belgium based on the Analogue Technique, Earth Syst. Sci. Data Discuss. [preprint], doi:, in review, 2025. The complete description of the method and performance is provided therein. Installation Download the dataset in a given folder, and in a terminal, and still in this folder, enter the following command: unzip RADCLIM-Analogs-median.zip This will unpack the dataset. You need at least 40Gb of free space on your disk to perform this operation. Usage The database is using Zarr format and is composed of two Zarr archives. One archive contains the database of the 25 best analogue days for each target day between 1940 and 2016. The second one contains all hourly median precipitation fields over Belgium from 1940 upon 2016. A notebook showing how to load and explore the dataset with the Python language is available here: https://github.com/ElkeDebrie/RADCLIM-Analogs. Note on the coordinate systems used in the dataset The original RADCLIM product was set on a grid with the Belgian Lambert 2008 used as projection. The product has a spatial resolution of 1 km with each estimate representing the averagedprecipitation on a square of size 1 km. The product covers an area from 0.3W to 9.7E in longitude and from 47.4N to 53.7N in latitude. The dataset coordinates x and y are the coordinates on the plane defined by the projection. For convenience, each data points latitude and longitude was also computed and provided in the dataset as latitude and longitude dataset coordinates. RADCLIM documentation The original RMI RADCLIM product documentation is available at https://opendata.meteo.be/cases/202107/radclim_userguide.pdf.