Mapping the March 2026 Escalation

The call, up front

Key Judgments

Executive summary

11,809 buildings, narrowly concentrated

Between the war start of 2 March 2026 and the satellite cohort close of 29 April 2026, Israeli strikes and ground operations damaged or destroyed an estimated 11,809 buildings across three regions of Lebanon, of which 1,266 are at the severe tier. The damage is not spread evenly. It concentrates on a 100-village belt south of the Litani river, with four villages in the Bent Jbeil district hit at rates between 19 and 40 percent of their building stock. Schools, mosques, and hospitals took disproportionate damage relative to their share of the building inventory.

The 16 April US-brokered ceasefire reduced strike tempo but did not produce a stable post-escalation environment. The Ministry of Public Health recorded 2,491 killed and 7,719 wounded by 25 April. The Integrated Food Security Phase Classification placed 1.24 million people, roughly a quarter of Lebanon's population, in acute food insecurity through August.

How the census is built

The verification frame

This assessment measures physical damage during the 58-day window from the pre-war satellite image of 5 March 2026 to the cohort-close image of 29 April 2026. The window opens three days after the 2 March escalation and closes 13 days into a fragile US-brokered ceasefire that took effect at midnight Beirut time on 17 April and was extended for three weeks on 23 April.

Damage is detected from two independent satellite sources at 10-metre resolution. A radar signal (Sentinel-1) detects structural change such as collapsed roofs and broken walls. An optical signal (Sentinel-2, five spectral bands) detects surface change such as rubble and burn scars. A claim is reported as verified only when both signals agree at the building level and the village it sits in carries at least one independently recorded strike during the window. The verification threshold cuts roughly 70 percent of the raw radar signal as false positive. The figures below are what survives the cut. A false-positive baseline derived from 22 zero-strike control cadastres establishes that any village above 4.40 percent of its building stock is decisively above noise, against a median undamaged baseline of 0.76 percent.

The report covers three analytically distinct regions: southern Lebanon (250 cadastres across 7 districts, the saturation-targeted frontline), the Bekaa (41 cadastres across 4 districts, the dispersed interior), and Dahieh and the Beirut suburbs (the urban overlay on 2024 rubble). Combined, the three frames cover roughly 910,000 building footprints.

The aggregate picture

The shape of the destruction

Two facts shape everything that follows. First, the destruction is geographically narrow. Twenty-five named villages south of the Litani carry roughly 70 percent of southern Lebanon's verified damage on 8 percent of its building stock. Twenty-five villages, not whole districts. Second, the damage rate at the regional level sits low against the building inventory. Across the seven southern governorates, 1.85 percent of buildings show triangulated damage; the cohort total across all three regions is 1.30 percent. These regional rates anchor the headline figure, but they understate the village-level concentration the rest of this assessment reads.

The 11,809 buildings carry triangulated damage at the moderate-or-severe tier. Of those, 1,266 cleared the severe threshold, meaning structural collapse, total roof loss, or comparable destruction. The census derives from satellite signals cross-checked against an open-source strike-incident database, not field survey. The verification cut is conservative; the raw radar signal alone would give a damage rate three to four times larger across the same area.

Read by district, the concentration is sharper still. Marjaayoun runs at 6.25 percent of its building stock and Bent Jbeil at 5.26 percent, both an order of magnitude above the next tier.

The top of the curve

The frontline saturation cluster

Within southern Lebanon, ten villages carry the bulk of the absolute damage count. These ten villages alone hold roughly 4,900 damaged buildings, 47 percent of southern Lebanon's verified count, on about 4 percent of its stock. Six of the ten sit in Bent Jbeil and Marjaayoun districts.

The frontline arc, district by district

Bent Jbeil district

Bent Jbeil sits on the immediate Israeli border and is the cohort's most-targeted district by every measure. It carries 3,445 verified buildings damaged across 38 cadastres, 609 of them at the severe tier, on a 65,537-building stock. That severe count is half of the southern severe total on 11.6 percent of southern building stock. The district holds the cohort's three most-extreme single-village damage rates: Rachaf at 43 percent, Beit Lif at 39 percent, and Aitaroun at 32 percent. Aitaroun alone records 963 damaged buildings, the highest absolute count in the cohort.

The frontline arc, district by district

The Marjaayoun salient

Marjaayoun carries the highest district rate in the cohort at 6.25 percent, with 3,890 damaged buildings. The salient runs from Meiss el-Jabal and Houla through Khiyam, Taybet Marjaayoun, and Aadaysseh to Qantara, the village destroyed with its bridge on 26 March, where 66 of 199 damaged buildings cleared the severe threshold. Meiss el-Jabal records 819 damaged buildings at 21 percent of stock, the second-highest absolute count in southern Lebanon.

What produced the damage

Strike geography

The damage census verifies only buildings whose cadastre carries at least one recorded strike. The underlying strike record holds every geolocated Israeli strike on Lebanon across the cohort and the two weeks that follow. Airstrikes dominate, and they cluster on the southern border strip, with secondary concentrations in the Bekaa, where the pattern reads as supply-line interdiction, and over Dahieh, where the 8 April Beirut mass-strike day produced a sharp late-cohort spike.

The dispersed interior

The Bekaa

The Bekaa took roughly one-tenth the damage rate of southern Lebanon. 407 buildings across 41 cadastres show triangulated damage, 18 at the severe tier, against a 249,130-building stock for a regional rate of 0.16 percent. The pattern is dispersed rather than concentrated. West Bekaa, where the Sohmor and Yohmor frontline runs, holds the highest local rate at 0.63 percent, still below the southern frontline. Every Bekaa cadastre received at least one strike, and the strike-to-damage ratio reads as supply-line interdiction rather than saturation bombardment.

The urban overlay

Dahieh and the Beirut suburbs

Dahieh, Beirut's southern suburbs, took 936 newly damaged buildings during the March window on top of the rubble already on the ground from the 2024 war, 31 of them at the severe tier, for a rate of 0.98 percent against the 2026 overlay. Bir el Abed (203 buildings, 2 percent) and Haret Hreik (218 buildings, 1.7 percent) anchor the sample. The 8 April Beirut mass-strike day, which killed 357 nationally across 100 airstrikes and hit central Beirut without warning, produced a substantial fraction of the late-cohort damage growth. Because the 2026 census reads change against a pre-war image, it captures only the new damage, not the standing destruction of the prior war.

Beyond the building census

Civilian facilities

Schools, mosques, and hospitals took damage at rates several times the building-level rate of the same areas. Schools in southern Lebanon registered triangulated damage at 33 percent of sampled facilities and mosques at 21 percent. In the Dahieh sample, schools ran at 16 percent and water and sewage facilities at 63 percent. The named facilities below cleared the verified-severe threshold.

The strike sequence on the Tibnine Governmental Hospital on 15 to 16 April put the last large hospital south of the Litani out of operation, and the four-paramedic killing in Mayfadoun on 15 April extended the pattern into the post-escalation window.

What the map encloses

The Yellow Line

The damage cluster describes the territory the IDF formalised through the post-ceasefire Yellow Line designation. The verified-damage belt runs continuously along the southern border, from Beit Lif in the west through Aitaroun, Aita ash-Shaab, and Yaroun to the Marjaayoun salient at Houla, Markaba, Meiss el-Jabal, Khiyam, Taybet Marjaayoun, Aadaysseh, and Qantara. The 510 km Yellow Line zone declared in week 7 and the 55-village no-go list issued in week 8 enclose almost exactly the cadastres carrying the highest damage rates. The targeting cleared the territory the buffer encloses, and the buffer codified the cleared territory.

What the damage sits inside

The human bracket

The damage maps onto a displacement and casualty crisis the 16 April ceasefire did not stabilise. Formal-shelter occupancy peaked at 141,000 on 16 April. 30,000 returned within 48 hours of the ceasefire, but 10,000 of those returnees came back to shelters by 24 April, citing uninhabitable housing and ongoing insecurity. The Ministry of Public Health recorded 2,491 killed and 7,719 wounded by 25 April since the war start.

Five of six Litani crossings were destroyed between 13 and 26 March, with the sixth at Khardali hit only on its access roads. An estimated 150,000 people south of the river lost reliable north-south movement, and no bridge was reconstructed during the 13-day post-ceasefire window. By cohort close, 55 border villages were under formal IDF no-go designation and 1.24 million people, roughly a quarter of Lebanon's population, were placed at acute food insecurity through August by the IPC working group.