Lebanon imagery and land cover data

uav imagery by Omran NAJJAR, acquired 2025-12-24 to 2025-12-24, 2.9 cm per pixel, published on OpenAerialMap.

**UNOSAT code: FR20191016LBN** This map illustrates satellite-detected fire hotspots based on the analysis of Visible Infrared Imaging Radiometer Suite (VIIRS) accessed via NASA FIRMS, between October 15-16, 2019. 30 hotspots were detected in Lebanon and 121 hotspots were detected in three analysed Governorates (i.e. Homs, Lattakia and Tartous) in Syria. This is a preliminary analysis and has not yet been validated in the field. Please send ground feedback to UNITAR - UNOSAT.

This land cover dataset at sub-national scale (Level 3) shows a detailed classification with information on the the crops representing at least 10% of the area. It is published on a dekadal basis. The data is provided in near real time from January 2009 to 2021. **Data publication:** 2020-01-01 **Supplemental Information:** No data value: 255 Unit: class Map code : L3-LCC-D **Citation:** FAO 2018. WaPOR Database Methodology: Level 3. Remote Sensing for Water Productivity Technical Report: Methodology Series. Rome, FAO. 72 pages. Licence: CC-BY-4.0 **Contact points:** Metadata Contact: [WaPOR](mailto:wapor@fao.org) Resource Contact: [WaPOR](mailto:wapor@fao.org) **Data lineage:** A supervised classification methodology is applied to assign a specific class to each pixel of the image. Training data consist of seasonal and long term metrics derived from dekadal NDVI time series, phenology and spectral reflectance data combined with reference data denoting the exact location of each of the classes specified in the table below. Data component developed through collaboration with the FRAME Consortium. More information can be found at: http://www.fao.org/in-action/remote-sensing-for-water-productivity/en/ **Resource constraints:** Creative Commons Attribution 4.0 International (CC-BY-4.0) **Online resources:** Download the data from Google Cloud Storage

The layer indicates the quality (in percentage %) of the original classifier and whether a pixel was re-classified during post-processing. The data is provided in near real time from January 2009 to 2021. **Data publication:** 2020-01-01 **Supplemental Information:** No data value: 255 Unit: % Map code: L3-QUAL-LCC-S **Citation:** FAO 2018. WaPOR Database Methodology: Level 3. Remote Sensing for Water Productivity Technical Report: Methodology Series. Rome, FAO. 72 pages. Licence: CC-BY-4.0 **Contact points:** Metadata Contact: [WaPOR](mailto:wapor@fao.org) Resource Contact: [WaPOR](mailto:wapor@fao.org) **Data lineage:** A combination of factors influences the accuracy of the classification across a land cover classification map. All land cover maps contain a fraction of falsely classified pixels. At level 3, the land cover classification quality layer combines the result of the machine learning classifier quality output, with flags that indicate when pixels were reclassified during post-processing: - a flag value of 240 indicates that a pixel was reclassified using a fixed mask (used in isolated cases). - a flag value of 241 indicates that a pixel was reclassified during post-processing. Pixel values represent the accuracy of the classification expressed in percentage (%). A quality value close to 100 represents more certainty regarding the classification, whilst pixel values close to 0 indicate pixels for which the classification is less accurate. **Resource constraints:** Creative Commons Attribution 4.0 International (CC-BY-4.0) **Online resources:** Download the data from Google Cloud Storage