WEBVTT

00:00:00.000 --> 00:00:10.000
Beirut, 4 August 2020: what does red prove?. This map examines damage after the explosion. Does every red area mean a destroyed building?

00:00:10.000 --> 00:00:22.000
First, a real view of the city. The space-station photograph from 11 August provides context. It does not show the explosion itself.

00:00:22.000 --> 00:00:36.000
Separate the photograph from the analysis layer. The colours on the right are not natural surface colours. They are an analysis of radar records before and after.

00:00:36.000 --> 00:00:48.000
Radar: a returning signal instead of light. The satellite sends a signal and measures its return. Surface changes can change that signal; the map uses this.

00:00:48.000 --> 00:01:02.000
Look at the concentration around the port. Red and orange marks cluster around the port. This helps narrow the area to examine first.

00:01:02.000 --> 00:01:14.000
The map unit is a pixel, not a building. NASA describes each coloured cell in this product as 30 × 30 metres. One cell does not mean one building.

00:01:14.000 --> 00:01:28.000
Zooming does not create new detail. Magnification only shows the existing pixels. It cannot produce building boundaries or a building count.

00:01:28.000 --> 00:01:42.000
Independent check: UNOSAT’s optical assessment. UNOSAT reports visible structural damage in separate satellite photographs. That is not a pixel-by-pixel validation of this map.

00:01:42.000 --> 00:01:56.000
Read the report’s limits too. UNOSAT explicitly says minor damage may be invisible from space and field validation is still needed.

00:01:56.000 --> 00:02:08.000
An uncoloured area does not mean “safe”. No colour on this map is not a certificate of no damage. We cannot decide the condition of individual homes from it.

00:02:08.000 --> 00:02:20.000
Finding: areas to inspect, not verdicts on buildings. Colours help investigate likely damage locations. This map cannot establish the cause of the explosion or who was responsible.

00:02:20.000 --> 00:02:36.000
Credit the source; bound the conclusion. ARIA/EOS produced the map. CSINT compared scale, dates and report limits; we did not reprocess the radar data.