Most historic eruptions with magnitudes of Me 5 or more (Section 1.3.2) were the first on record for the responsible volcano. The deadliest eruption in history – that of Tambora in 1815, which directly or indirectly killed perhaps more than 100,000 people – is just one example. None of the volcanoes responsible for the four largest eruptions of the past century or so, Katmai 1912 (Alaska), El Chichón 1985 (Mexico), Mt Pinatubo 1991 (Philippines) or Chaitén 2008 (Chile), had previously erupted in recorded history, nor were they considered potentially hazardous. The message is clear – the largest eruptions in future are likely to come from previously little-known, even unheard-of, volcanoes.

To go about identifying these potentially dangerous volcanoes, it is necessary to deduce the timing, magnitude, intensity, style and gas yield of their past eruptions. And to understand their significance, we also need to determine and quantify the nature and extent of their impacts on the atmosphere, environment, climate and society. The latter endeavour is especially challenging since attributing cause and effect can be perilous. Association of climate change to a particular eruption requires careful compilation of evidence. For instance, short-term global climate change can have numerous explanations such as the El Niño Southern Oscillation, the North Atlantic Oscillation and variability in solar radiation.