On 4 August 2020, a series of explosions occurred in the Port of Beirut, Lebanon, widely reported to have been caused by detonation of a large quantity of ammonium nitrate (approximately 2750 tonnes) following a fire in the warehouse where it was being stored. The final and largest explosion caused considerable damage to the surrounding area and at the time of writing resulted in at least 181 deaths and over 6000 injuries.
Shortly after the explosion, social media users began sharing videos showing the initial fire, detonation, and propagating blast wave. In many of these videos, the moment of detonation and blast wave time of arrival (\(t_\mathrm {a}\)) at the observer’s position and/or recognisable landmarks are clearly discernible from the footage and audio. Thus, these videos make it possible to approximately determine the time of arrival of the shock front at different distances from the source of the explosion.
In this article, we examine 16 videos posted online [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]Footnote 1 at various locations across the city of Beirut (Fig. 1). We correlate the calculated distance–time relationship with well-known semi-empirical laws [17] in order to estimate the approximate yield of the 2020 Beirut explosion by minimising the mean absolute error between the data and semi-empirical predictions.
There is a pressing need for rapid, accurate assessment of the size of the explosive yield in such events, both to inform first responders of the likely injuries and structural damage at various distances from the explosion and to provide a factual context for political and media discussion.