Pyroclastic Density Current

Primary reference(s)

Branney, M.J. and P. Kokelaar, 2002. Pyroclastic density currents and the sedimentationnof ignimbrites. Geological Society Memoir 27. Geological Society of London.

Cole, P.D., A. Neri and P.J. Baxter, 2015. Hazards from pyroclastic density currents. In: Sigurdsson, H., B. Houghton, S. McNutt et al. (eds.), The Encyclopedia of Volcanoes, 2nd edition. Academic Press, pp. 943-956.

Additional scientific description

The following terms may be considered sub-types of pyroclastic density currents (PDCs): pyroclastic flow, block-and-ash flow, pumice flow, lateral blast, pyroclastic surge. The pyroclastic flow and surge are two end members (dense and dilute end, respectively). The term ‘ignimbrite’ is commonly used as a general term describing pumice- and ash-rich PDC deposits of very varied volumes (Druitt, 1998; Branney and Kokelaar, 2002), but has also been used to refer, predominantly, to the large-volume end of this spectrum (e.g., Wilson and Hildreth, 2003).

PDCs are produced from volcanic eruptions across many orders of magnitude, from small-volume events (<0.001 to 1 km3) to caldera-forming eruptions with volumes around 101–103 km3 of erupted material (Druitt, 1998; Dufek et al., 2015). PDCs are hot, unstoppable, gas-particle mixtures that move extremely quickly across the ground surface at velocities of tens to hundreds of kilometres per hour and have temperatures of typically between 200 and 600°C (Cole, 2015; Dufek et al., 2015). Most PDCs propagate to distances of between a few to tens of kilometres from the source (Ogburn, 2012). For exceptionally large-magnitude events, PDCs may travel over 100 km and cover areas of up to 103–104 km2 (Takarada and Hoshizumi, 2020). Many of the aforementioned variables can be used as hazard metrics for PDCs: flow speed, flow density, temperature, dynamic pressure, flow and deposit thickness, maximum runout, invasion area, etc.

Two different flow parts commonly form PDCs: a dense, basal undercurrent dominated by particle-particle interactions; and a dilute, upper part whose motion is mainly dominated by turbulence (Branney and Kokelaar, 2002; Sulpizio et al., 2014; Cole, 2015). The dense basal part strongly interacts with (and is controlled by) the topographic surface as it erodes and deposits material along its path (Doronzo, 2012). The dilute upper part tends to be less controlled by topography and may decouple from the main dense undercurrent, overcoming topographic obstacles and following diverse propagation paths (e.g., Fisher, 1995; Ogburn et al., 2014). Extensive numerical modelling of PDCs has been conducted over recent decades, to better understand PDCs and quantify their hazard (Sulpizio et al., 2014; Dufek et al., 2015). Most past efforts have focused on simulating either the dense basal (e.g., Patra et al., 2005) or the dilute upper part of PDCs (e.g., Bursik and Woods, 1996), but several multiphase flow models have also been presented (e.g., Suzuki et al., 2005).

Between 1500 and 2017 AD, PDCs were the most deadly of all volcanic hazards: there were 102 fatal incidents and 59,958 fatalities caused directly by PDCs. 50% of PDC fatalities were recorded up to 10 km from a volcano and 90% up to 20 km (Brown et al., 2017). The 1883 eruption from Krakatau volcano (Indonesia) resulted in PDC fatalities up to 80 km from the volcano, aided by the passage of PDCs over the sea (Carey et al., 1996).

Metrics and numeric limits

Not available.

Examples of drivers, outcomes and risk management

PDCs can kill all living things and destroy structures by abrasion, impact, burial and heat.

Risk to building structures has not been systematically assessed but dense PDCs can bury buildings and destroy their openings (windows, doors) and, in dilute PDCs, dynamic pressures of a few kilopascal can cause moderate to heavy damage to buildings (Valentine, 1998; Zuccaro et al., 2008). 

Deaths commonly result from thermal injury (including laryngeal and pulmonary oedema), asphyxiation and impact or blast trauma (Baxter, 1990). Survivors of PDC inundation can suffer from severe burn injuries requiring specialist treatment (Loughlin et al., 2002).

Indirect casualties can include accidents, for example related to evacuation or unsafe driving conditions, heart attacks and cascading hazards such as fires, famine and disease. Indirect deaths can dwarf the numbers of direct deaths (Brown et al., 2017).

High resolution (spatial and temporal) monitoring of lava-dome extrusion rates, and topography, can enable dome collapse PDCs to be anticipated, resulting in timely evacuation (Pallister et al., 2013). Probabilistic volcanic hazard assessments of PDCs (e.g., Sandri et al., 2018) are increasing in number and methods are improving.