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 zones are used as reporting units of the air quality in Belgium. These zones are defined based on Article 3 of the 2004/107/EC Directive and Article 4 of the 2008/50/EC Directive. A zones shall mean a part of the territory of a Member State, as delimited by that Member State for the purposes of air quality assessment and management (source: 2008/50/EC).2008/50/EC Article 4: Establishment of zones and agglomerations: Member States shall establish zones and agglomerations throughout their territory. Air quality assessment and air quality management shall be carried out in all zones and agglomerations.

Daily forecasts for each Belgian commune, from current day up to the next 14 days. Available parameters are : min temperature, max temperature, weather type, wind speed, wind direction, maximum hust speed, quantity of precipitation. This product is not publicly available.

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.

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.

This dataset contains the spatial metadata of the complete collection of aerial photos of the National Geographic Institute (NGI), from 1946 to today. The aerial photos themselves are not part of the dataset. The collection comprises more than 100,000 aerial photos. The images are predominantly panchromatic (black and white). From the late 1970s onwards, colour images were also taken, and from the end of 2004 onwards all images were taken in colour. The analogue aerial photos in the collection were scanned in high resolution. More recent aerial photos were captured directly in digital format. This specific dataset includes three layers: the centroids (centre points) of the individual aerial photos, the footprints (the outlines of the area on the ground covered by each photo), and the flight paths of the aircraft used during image acquisition. For each aerial photo, all known metadata is included in the attribute table, such as the acquisition date, the flight and the strip, the dimensions, the scale, the ground resolution, the radiometry (black and white or colour), the coordinates of the centre point, and the camera, lens and film used. This data makes it possible to find out which aerial photos are available for a given location and period, and what the characteristics of each photo are. The data can be consulted via the associated web service (WFS). The aerial photos can be ordered in high resolution via https://shop.ngi.be/nl/luchtfotos/.

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.

Summer and winter mean relative humidity given in %. The relative humidity fields are available for the present target year 1975, which corresponds to the middle of the 30-year period 1961-1990.

An hillshade is a homogeneous and regular points grid, indicating the grey tone deriving from their orientation relative to the chosen fictitious light source. The Hillshade DSM 1m is a representation of the hillshading of the DSM 1m.

UNDER EMBARGO - This dataset is part of BE/2023 sampling campagn in SW Greenland fjords (Igaliku and Tunulliarfik). Pelagic community was analysed using Imaging Flow Cytometry (iFCM) with an ImageStream®X Mk II. Cells were grouped into functional size classes—pico-, nano- and microplankton—according to measured cell length. Cells lacking chlorophyll autofluorescence were classified as heterotrophic or chemotrophic organisms, including heterotrophic picoplankton/bacteria (HP; ≤2 µm) and heterotrophic nanoplankton (HN; 2–20 µm). No larger heterotrophs (>20 µm) were visually detected. Autofluorescent cells were considered phototrophic, although this fraction may also include mixotrophic taxa, and comprised picophytoplankton (AP; ≤2 µm), nanophytoplankton (AN; 2–20 µm), and microphytoplankton (AMicro; 20–100 µm). To estimate the biovolume of each plankton class, the two-dimensional cell surface area measured by the IDEAS® imaging software was multiplied by the mean cell width, assuming that cell width approximates the third spatial dimension. Carbon biomass was subsequently derived from biovolume using established carbon–volume relationships. For the HP fraction, carbon content was estimated using the bacterial conversion proposed by Romanova and Sazhin (2010), where volume is expressed in µm³. Although the HP fraction may also include heterotrophic picoeukaryotes, and its biomass may therefore be partly underestimated, this conversion was applied because the fraction was assumed to be numerically dominated by bacteria. For the other protist groups, carbon biomass was derived following Menden-Deuer and Lessard (2000). Carbon values were converted from pg C cell⁻¹ to carbon biomass (µg C L⁻¹) based on cell abundance.