Alaska Volcano Observatory

Fierstein and Hildreth (2001) provide information about the magitude of the 1912 eruption at Novarupta and Katmai: "The world's largest volcanic eruption of the 20th century broke out at Novarupta [see fig. 1 in original text] in June 1912, filling with hot ash what came to be called the Valley of Ten Thousand Smokes and spreading downwind more fallout than all other historical Alaskan eruptions combined. Although almost all the magma vented at Novarupta, most of it had been stored beneath Mount Katmai 10 km away, which collapsed during the eruption. Airborne ash from the 3-day event blanketed all of southern Alaska, and its gritty fallout was reported as far away as Dawson, Ketchikan, and Puget Sound [see fig. 21 in original text]. Volcanic dust and sulfurous aerosol were detected within days over Wisconsin and Virginia; within 2 weeks over California, Europe, and North Africa; and in latter-day ice cores recently drilled on the Greenland ice cap."
Schaaf (2004) contains excerpts and summaries of eyewitness accounts of the eruption; a copy of this publication is available online at https://www.nps.gov/articles/aps-v11-i1-c9.htm. Detailed, firsthand narratives of the events as experienced at Kodiak are given in Perry, 1914, and Erskine, 1962.
Hildreth and Fierstein also published a free download tome of information about this eruption: https://pubs.usgs.gov/pp/1791/
From Fierstein and others (1998): "On the afternoon of June 6, 1912, an ominous cloud rose into the sky above Mount Katmai on the Alaska Peninsula. The cloud quickly reached an altitude of 20 miles, and within 4 hours, ash from a huge volcanic eruption began to fall on the village of Kodiak, 100 miles to the southeast. By the end of the eruption on June 9th, the ash cloud, now thousands of miles across, shrouded southern Alaska and western Canada, and sulfurous ash was falling on Vancouver, British Columbia, and Seattle, Washington. The next day the cloud passed over Virginia, and by June 17th it reached Algeria in Africa.
"During the 3 days of the eruption, darkness and suffocating conditions caused by falling ash and sulfur dioxide gas immobilized the population of Kodiak. Sore eyes and respiratory distress were rampant, and water became undrinkable. Radio communications were totally disrupted, and with visibility near zero, ships couldn't dock. Roofs in Kodiak collapsed under the weight of more than a foot of ash, buildings were wrecked by ash avalanches that rushed down from nearby hillslopes, and other structures burned after being struck by lightning from the ash cloud.
"Similar conditions prevailed elsewhere in southern Alaska, and several villages were abandoned forever. Animal and plant life was decimated by ash and acid rain from the eruption. Bears and other large animals were blinded by ash and starved when large numbers of the plants and small animals they lived on were wiped out. Millions of dead birds that had been blinded and coated by volcanic ash littered the ground. Aquatic organisms, such as mussels, insect larvae, and kelp, as well as the fish that fed upon them, perished in ash-choked shallow water. Alaska's salmon-fishing industry was devastated, especially from 1915 to 1919, because of the starvation and failure of many adult fish to spawn in ash-choked streams.
"In 1916, a National Geographic Society expedition led by Robert Griggs visited Mount Katmai and found a 2-mile-wide crater where its summit had been before 1912. Nearby, the expedition discovered a newly formed lava dome they called 'Novarupta' and huge flows of volcanic ash filling what they named the 'Valley of Ten Thousand Smokes' for the numerous plumes of steam rising from the still hot ground. Griggs' descriptions of these spectacular features helped persuade President Woodrow Wilson to create Katmai National Monument (now National Park) in 1918.
"In the 1950's, volcanologists discovered that the 1912 eruption was actually from Novarupta, not Mount Katmai. Novarupta's eruption had removed so much molten rock (magma) from beneath Mount Katmai that it caused a cubic mile of Katmai's summit to collapse."
Hildreth (1983) gives the following detailed description of the events of the eruption: "Because there were no geophysical instruments in Alaska, no scientfic observers in the district, and no eyewitnesses with a view of the VTTS [Valley of Ten Thousand Smokes], most reports of seismic, acoustic, and eruptive phenomena were not necessarily very accurate, chronologically or descriptively. The most reliable data are the visual observations from aboard the steamer Dora and the record of tephra falls at Kodiak [see figs. 1, 4, in original text] (Martin, 1913; Griggs, 1922). Martin visited Kodiak, Uyak, Katmai village, and Cold Bay [see fig. 1 in original text] in August of 1912 and gathered what accounts he could from residents of the district, but he did not venture inland to the volcanoes nor was he able to interview anyone from the two bands of native hunter-fishermen said to have been within 30 and 40 km, respectively, north and south of Novarupta at the time of the initial outburst. Reconstructions of the eruptive sequence by Griggs (1922) and Fenner (1923, 1925, 1950) are heavily inferential, overinterpretive both of Martin's data and of latter-day interviews with people said to have been at Savonoski when the eruption began (see below), and suffer badly from the incorrect assumption that much of the tephra had come from Mt. Katmai. The following is a critical reappraisal of what appear to be the facts, as assembled largely by G.C. Martin (1913; and unpublished USGS field notes); all times cited are adjusted to Alaskan local time.
"Earthquakes were reported to have been felt at Katmai village (30 km SSE of Novarupta) as early as the evening of 31 May, and severe shocks were felt at Uyak, Kanatak, and Nushugak (200 km WNW) on 4 and 5 June. On the morning of 6 June, explosions were heard at Nushugak and at Seldovia (240 km ENE) but there were no reports of accompanying seismicity or ash clouds. The same morning the Dora left Uyak at 0845 and steamed north-eastward up Shelikof Strait, the Katmai-group volcanoes in full view with a 'strong westerly breeze and fine clear weather,' but no one aboard noticed a tephra column until 1300 (when the vessel was ~88 km southeast of Novarupta). That column was clearly Plinian and, as shown below by the stratigraphic data, it was predominantly rhyolitic; the cloud overtook the Dora by 1500 and began dropping ash at Kodiak (170 km ESE) [see figs. 1, 4, in original text] by 1700.
"Separate interviews with two individuals said to have been in or near Savonoski (30 km N) on the morning of 6 June were conducted in 1918 (Griggs, 1922, p. 17) and in 1923 (Fenner, 1925, p. 216); these have been interpreted to suggest that the ash flow in the VTTS began in the morning several hours prior to the first high tephra column. This is quite unlikely in view of: (1) the record of the Dora; (2) the fact that the basal tephra layer in the VTTS, the distinctively rhyolitic Layer A, does not occur atop the ash-flow deposit; and (3) the common association of pumiceous pyroclastic flows with the collapse of vertical eruption columns. Preliminary eruptive activity at Novarupta on the morning of 6 June is quite reasonable and may be the best explanation of the noise and dust reported that morning at Savonoski and the explosions heard at Nushugak and Seldovia; but a major eruption column and a rhyolitic ash flow several km^3 in volume are certainly excluded prior to 1300.
"Major shocks that punctuated virtually continuous seismic activity were reported locally at ~1300 and ~2300 on 6 June and ~2240 on 7 June. These times are not necessarily very accurate. The first instrumentally recorded teleseism (at Seattle) is reported to have arrived at 1241 on 6 June (Fenner, 1925). Harvard, Ottawa, and several Eurasian stations recorded many teleseisms originating in southwest Alaska between 1805 on 6 June and ~2100 on 11 June. One of these near midnight on 6/7 June has been estimated at magnitude 6.4 and another at 0606 on 10 June at magnitude 7.0 (Coffman and von Hake, 1973). Earthquakes following the main phases of the eruption were felt at Cold Bay (60 km SSW) on 50 of the 70 days through mid-August (Fenner, 1925).
"A great explosion, accompanied by an earthquake felt at Cold Bay ~1300 on 6 June (Martin, 1913), was audible throughout the region and close in time to the first recorded teleseism and to the first sighting of tephra by the Dora. Another blast at ~1500 on 6 June, for which no accompanying earthquake was noted, may have been the most severe outburst of all, being heard for hundreds of kilometers (Martin, 1913). Martin suggested, but did not cite any evidence for, explosions synchronous with the large earthquakes felt nearby at ~2300 on 6 June and 2240 on 7 June. There were, however, reports of loud noises between midnight and 0200 on 8 June, at Katanak (95 km SW) and at Cordova (590 km ENE) (Martin, 1913; Fenner, 1925). Explosions continued to be heard, some as far away as Juneau (1200 km) until 10 June, though how many of the noises may have been thunder is impossible to assess.
"Major tephra eruptions, first noted by the Dora at 1300 on 6 June, continued to fall heavily on downwind settlements nearby until 9 June. The principal sector of dispersal was southeastward [see fig. 1 in original text], and the most distant positive record of ashfall was in Puget Sound (2400 km SE); atmospheric effects were worldwide (Griggs, 1922; Volz, 1975). In the main downwind direction at Kodiak (170 km ESE) there were three discrete periods of ashfall [see fig. 4 in original text]: (1) 1700 6 June until 0910 7 June; (2) ~1200 7 June until 1430 8 June; and (3) during the night of 8/9 June. The 9th of June dawned clear, and no further ash-falls were recorded on Kodiak Island. A major vapor plume and, possibly, sporadic ashfalls close to the source lasted all summer (Martin, 1913; Griggs, 1922; Fenner, 1925). Timing of the emplacement of the Novarupta dome is poorly known; its extrusion followed the last major tephra fall and was complete at the time of discovery on 31 July 1916.
"Details of correlation between audible explosions, tephra falls, and seismic events are not well known. Martin (1913) is the principal source, but his article has been misread and overinterpreted and ensuing errors propagated in successive publications. The seismicity does not correlate very well with eruptive events, and indeed there need be no correspondence (e.g. Filson and others, 1973; Nairn and others, 1976; Yokoyama and others, 1981). Much of the seismic activity seems more likely to have been related to fitful subsidence of Katmai and Novarupta calderas.
"At the onset of eruption, no one is known to have had a view of the VTTS (much less the Novarupta area itself), and only the party of native fishermen then southwest of Katmai village en route to Cold Bay may have been in a position to see Mt. Katmai. The Savonoski interviews (above) smack of ex post facto embellishment, although Martin (1913, p. 147) accepted a third hand report in 1912 that the villagers there may have seen Mt. Katmai after its collapse, but before 9 June. It is not clear where they might have had such a view, because, even notwithstanding the ash clouds, Mt. Katmai is not visible from Savonoski or along their escape route down the lake to Naknek [see fig. 1 in original text]. Martin, Griggs, and Fenner accepted such reports to mean that Mt. Katmai had lost its top by the afternoon of 6 June. This may be true, but it is not clear how Mt. Katmai could have been seen at all during the eruptive interval at Novarupta or what meaning should be attached to such expressions as 'blown off' or 'blown up.' No definitive geologic evidence has been found to fix the timing of Mt. Katmai's collapse, but it may be a reasonable inference that in addition to the syneruptive earthquakes, much of the summer-long seismicity can be attributed to such collapse, as well as to tectonic adjustments over a somewhat wider area, or even to dome emplacement at Novarupta."
Hildreth (1983) also reports that "pumice in the intial fall unit (A) is 100% rhyolite, but fall units atop the ash flow are >98% datcite; black andesitic scoria is common only in the ash flows and in near-vent air-fall tephra." He also states: The Novarupta lava dome is "diameter 380 m, its hieight ~65 m, and the intermediate (mostly dacite) lava interbanded with the rhyolite consitutes no more than 5% of the exposure; most of the conspicuous banding reflects textural variation in the rhyolite."
Fierstein and Hildreth (2001) estimate that the Novarupta eruption of 1912 ejected "at least 17 cubic km of fall deposits and about 11 cubic km of ash-flow tuff (ignimbrite) * * * emplaced in about 60 hours, representing a magma volume of about 13 cubic km (Fierstein and Hildreth, 1992). Hildreth (1987) estimates the volume of the Novarupta lava dome to be 0.005 cubic km.