END OF AN ERA: THE FINAL EXPLOSIVE ERUPTIONS OF KEANAKĀKO‘I TEPHRA AT KĪLAUEA
A widespread assumption is that the golden pumice of Sharp et al. (1987), the product of one or more high lava fountains, represents a transition between the two periods. It is preserved mainly west of the caldera (an unusual dispersal direction given prevailing N and NE trade winds), where it rests on lithic fall and surge deposits of the lethal 1790 eruption. In 1823, Ellis noted the golden pumice near Kīlauea summit. No tephra overlies it, so modern workers assumed that the golden pumice is the only post-1790 explosive deposit of the 19th century.
This assumption is wrong. We recently found that at least six other explosive eruptions took place after 1790, starting with a fall of pumice and achneliths (together called the eastern pumice) and followed successively by falls of fine lithic ash, Pele’s hair, lithic lapilli, inversely size-graded mixed vitric and lithic ash and fine lapilli, and lithic blocks and lapilli. This tephra was deposited mainly south of the caldera, a common dispersal direction that is very different from that of the golden pumice. Until recently, no outcrop was known to include both the golden pumice and the other six deposits.
In September 2016, we discovered an exposure just SSW of the caldera in which the golden pumice clearly underlies the six other explosive deposits. Several centimeters of alluvium separate the golden pumice from the eastern pumice and the younger fall deposits. From previous work, the golden and eastern pumice are chemically distinct, and microbeam analyses of pumice samples from below and above the alluvium match those of the golden and eastern pumice, respectively.
Thus the golden pumice clearly belongs to the Keanakāko‘i explosive period and does not record a transition to the current effusive period. Instead, the Keanakāko‘i ended with lithic-rich phreatomagmatic and phreatic eruptions younger than the golden pumice, not with a high lava fountain. We speculate that the change from explosive to effusive activity took place rapidly (a few years or less?) as the magma supply rate increased, keeping the feeding conduit full and impeding involvement with groundwater.