Total lightning observations, i.e., cloud-to-ground (CG) strokes and intra/inter-cloud (IC) pulses, in Belgium. BELLS data is given in Universal Time! (local time winter = UT +1; local time summer = UT + 2). Each record contains following fields: 1. Network type, 0 = LF, 1 = VHF-TOA, 2 = VHF-ITF, 3 = VLF 2. Year, 1970 to 2032. 3. Month, with January as 1 and December as 12. 4. Day of the month, 1 to 31. 5. Hour, 0 to 23. 6. Minute, 0 to 59. 7. Second, 0 to 60. 8. Nanosecond, 0 to 999999999. 9. Latitude of the calculated location in decimal degrees, to 4 decimal places, -90.0 to 90.0. 10. Longitude of the calculated location in decimal degrees, to 4 decimal places, -180.0 to 180.0. 11. Altitude in meters, -4950 to +60535. This field will be 0 if the altitude is not provided in the lightning data. 12. The altitude uncertainty in meters. This field will be 0 if the altitude uncertainty is not provided in the lightning data. 13. Estimated peak current in kiloamps, -9999 to 9999. 14. VHF Range normalized power, -9999.0 to 9999.0 15. Multiplicity for flash data (1 to 99) or 0 for strokes. 16. Cloud pulse count for the lightning record. This field will be -1 if the lightning did not contain the cloud pulse count attribute. 17. Number of sensors participating in the solution, 2 to 99. 18. Degrees of freedom when optimizing location, 0 to 99. 19. The error ellipse angle as a clockwise bearing from 0 degrees north, 0 to 180.0 degrees 20. The error ellipse semi-major axis length in kilometers, 0 to 50.0km. 21. The error ellipse semi-minor axis length in kilometers, 0 to 50.0km. 22. Chi-squared value from location optimization, 0 to 999.99 23. Rise time of the waveform in microseconds, 0 to 99.9 24. Peak-to-zero time of the waveform in microseconds, 0 to 999.9 25. Maximum rate-of-rise of the waveform in kA/usec (will be a negative rate if discharge is negative), -999.9 to 999.9 26. Cloud indicator, 1 if Cloud-to-cloud discharge, 0 for Cloud-to-ground 27. Angle indicator, 1 if sensor angle data used to compute position, 0 otherwise 28. Signal indicator, 1 if sensor signal data used to compute position, 0 otherwise 29. Timing indicator, 1 if sensor timing data used to compute position, 0 otherwise 30. Flash ID, a unique integer value identifying the flash grouping a stroke belongs to or 0 if not set. This product is not publically available yet.

The DTM 5m, digital terrain model at 5m-resolution, is a homogeneous and regular point grid indicating the height of the ground level in order to model its surface, without taking into account objects on the surface such as buildings and vegetation. The DTM 5m is the result of the assembly of different data coming from various sources and with different coordinate systems and resolutions. The DTM 1m is achieved by interpolating in Lambert 2008 source data in Lambert 72 and at a 1m-resolution from the Flemish (2013-2015) and Brussels (2021) regions, and by adding Lambert 2008 data at 1m-resolution from the Walloon Region (2021-2022). The DTM 5m is obtained by resampling of the DTM 1m. In addition, the DTM 5m is improved and updated internally using direct photogrammetric surveys based on the most recent aerial photographs.

Hail products are derived from the observed vertical profiles of radar reflectivity and the NWP vertical profiles of temperature. Three types of products are generated. poh : probability of hail of any size (larger than 0.5 cm diameter)expressed in %. posh : probability of severe hail(larger than 2cm)expressed in %. mesh : maximum expected size of hailexpressed in mm of hailstone diameter. All products are generated every 5 minutes. This product is not publically available yet.

Since August 2019, users of the RMI smartphone app are able to send an observation of the meteorological conditions at a certain place and a certain time. The observations provide information about the weather conditions and potentially severe weather to the other users and to RMI. The collection of citizen weather reports is a valuable complement to the information obtained with the classical instruments like stations, radar and satellite. The data can be exploited for nowcasting, warnings and model verification, and eventually in assimilation. A general introduction of the data and their characteristics can be found in Reyniers et al. (2023). A basic quality control is implemented on the received observations via a plausibility check. This plausibility check determines whether an observation is plausible, suspicious or false, by comparing it to the INCA-BE nowcasting system using a simple thresholding scheme. INCA-BE is RMI's operational nowcasting system described in Reyniers et al. (2021). There is no strict spatial extent since there is no restriction at the input side: users can send observations from all over the globe. The bulk of the observations are received from within Belgium. Note that the plausibility check is not available for reports from outside Belgium.

The average household size of private households by administrative unit (region, province, district and municipality) for Wallonia.

Characteristics of the cadastral parcels - Land use corresponds to the dataset describing the cadastral nature of the parcels such as recorded by the FPS Finance for tax purposes. This dataset is composed of seven classes. The first class shows, at the national level, for each type, the number of parcels, their cadastral income and their total surface area, as well as the median cadastral income of the plots. The second class includes this information at the level of the three regions. The following classes do the same at the level of provinces, arrondissements, municipalities, cadastral divisions and statistical sectors. The dataset can be freely downloaded as a zipped CSV.

The data consists in tables of estimated extreme precipitation quantities for precipitation durations ranging from 10 minutes to 30 days and return periods ranging from 2 to 200 years for each Belgian municipality.

Landuse-Particular Zone is the INSPIRE 2D vector data set of the "Particular zones" layer originated from the Top10vector product of the National Geographic Institute in Belgium . It includes the "ExistingLandUseDataSet" features and the "ExistingLandUseObject" features in accordance with the INSPIRE specifications.

Beach Litter count data collected across 10 beaches of the Belgian coast in 2023 and 2024 by volunteers of the citizen science association Proper Strand Lopers. One of the key indicators of abundance, composition and trends of litter in the marine environment is the amount on beaches. OSPAR monitors litter on 100m stretches at over 70 beaches in the North-East Atlantic following common monitoring guidelines. The monitoring records litter in 112 predefined litter items in 11 types: Plastic / polystyrene, Metal, Paper and cardboard, Wood, Sanitary waste, Cloth, Rubber, Glass, Pottery/ceramics, Medical waste and Faeces.

UNDER EMBARGO - This dataset is part of BE/2023 sampling campagn in SW Greenland fjords (Igaliku and Tunulliarfik). Samples for DNA extraction were collected along fjord transects at several depths of the water column. The focus of the study was to determine the taxonomical composotion of bacterial community in two Arctic fjords.