The United Nations Development Program (UNDP), in coordination with the National Council for Scientific Research in Lebanon (CNRS-L), announced the results of a rapid assessment of building-level damage in southern Lebanon, which includes areas south of the Litani River, including the districts of Bint Jbeil, Marjayoun, Nabatieh, Tyre, and Sidon, in continuation of the previous assessment for Beirut and Mount Lebanon. The results showed that the total direct damage to buildings in southern Lebanon is estimated at approximately 1.38 billion US dollars, with an estimated rubble volume of About 3.1 million cubic metres. The complete destruction of 11,095 buildings was recorded, affecting 17,891 housing units, while 2,242 buildings were partially damaged (5,219 housing units), and 9,311 buildings were exposed to minor damage (18,282 housing units). The most affected properties in terms of the number of completely destroyed buildings in the Bint Jbeil District are Ainatha (1,658 buildings) and Bint Jbeil. (1,076 buildings), and in the district of Marjayoun, Mays al-Jabal (969 buildings) and Taybeh (824 buildings), followed by Borj al-Shamali (370 buildings) and Naqoura (216 buildings) in the district of Tyre. The evaluation used a methodology based on geographical artificial intelligence (GeoAI), supported by visual office verification operations without implementing field verification operations. High-resolution satellite images taken on 29 April 2026 were compared with reference images dating back to 23 October 2025, allowing the visual damage associated with the conflict to be monitored, including roof collapses, structural deformations and the accumulation of rubble at the building footprint level. The size of the rubble and the cost of damage were also calculated using building footprints, number of floors and standard replacement costs, in line with the UNDP methodology and previous assessments. Although the assessment provides a comprehensive and reliable picture of the extent of damage, it does not include basements or underground facilities, nor damage to vital infrastructure such as roads, bridges, electricity, water and communications networks. The numbers also reflect the situation as of April 29, 2026 and do not take into account ongoing recovery or reconstruction work. It is noteworthy that the assessment was conducted using high-resolution satellite images taken on April 29, 2026, and compared with reference images dating back to October 23, 2025, based on a geographic artificial intelligence (GeoAI) model specially trained for this purpose. The model results were subsequently subjected to a systematic visual verification process conducted by the UNDP team at the building-by-building level. This process confirmed the accuracy of the adopted methodology, as the overall accuracy rate reached about 85 percent, which provides a high level of confidence in the final results.
