Various compilers break the time period of 1938-1948 into several different eruptions, or one continuous eruption. Jacob and Hauksson (1983) assert that during this time period Pavlof was in nearly continuous minor explosive eruption, with "significant eruptions in 1936, July 1937, and 1942. Lava flow may have been extruded in May 1948. Otherwise, minor, nearly continuous ash emissions. Photo taken in 1942 shows 1911 fissure no longer present (the wound had been healed), and a single cinder cone on upper NE flank (moat between this cone and summit, but flanks of cinder cone merged with Pavlof proper elsewhere.)"
From Kennedy and Waldron (1955): "Major explosive eruptions occurred in 1936 and early in 1942; during the intervening period the emission of ash-laden steam was almost continuous. Reports by Army and Navy fliers indicate almost continuous discharge of smoke from the crater since 1942. Activity was probably less during the summer of 1946 than at any time during the preceding 4 years."
Coats (1950) reports: 1936 minor explosive eruption, 1937 July minor explosive eruption; 1945 smoke; 1942 major explosive eruption; 1947 smoke; 1948 May lava flow?, minor explosive eruption
From Hantke (1955): "July 1937 multiple big ashfalls over "Half-Island, Alaska", 1947 smoke, May 1948 explosive activity and probably lava flows."
From Hantke (1951): June 6, 1945, weak explosions with steam and ash emissions every minute from small summit crater
Robinson (1948) reported that in 1945 Pavlof was "mildly active, puffs of black ash and steam at ~1min intervals."
Meredith (1998) writes that Pavlof "showered the country with ash" in 1937, and that on September 20, 1946 "Mt. Pavlov blew out spurts of smoke and ash." She reports that she asked "one of the men if he worried about living so close to an active volcano. 'No,' he said, 'we just call him Mount Puff-Puff.'"

Coats (1964) writes about Pavlof: "Reported as quite active for about a year. Probably Strombolian phase." (July 1962-June 1963).
Jacob and Hauksson (1983) report that this eruption began around 1960 and continued until sometime in 1963, and consisted of mild ash eruptions and lava flows.
Dr. Kenneth Morin took photographs of ash and steam erupting from Pavlof on May 28, 1960. These images are viewable on the World Wide Web at: www.avo.alaska.edu/volcanoes/v… .

Stone and Kienle (1975) report: "An eruption of the Pavlof Volcano was observed at 10:00 p.m. local time 12 November (0800 GMT 13 November) from a location some 33 km from the summit. The observers reported seeing 'shooting flames,' coupled with a possible lava flow down the northwest flank.
"A seismometer located 7 km southeast of the summit recorded increased micro-earthquake activity coincident with the time of the reported lava eruption.
"Seismic records showed onset of harmonic tremor on 13 November, 07:00 AST, coincident with the reported eruption of lava."
McNutt (1985) estimates the total dense rock equivalent volume of eruptive material for this eruption as 4.8 - 7.9 x 10^6 cubic meters. McNutt (1999) estimates the total dense rock equivalent volume of eruptive material for this eruption as 6.4 x10^6 cubic meters.

Pavlof volcano was in eruption nearly continuously from September 13, 1975 until March, 1977. Jacob and Hauksson (1983) summarize the 1975-1976 activity as follows: "13 September - 06 October, 1975: Period of generally weak ash emissions, often likened to the chugging of a locomotive. Both harmonic tremor and explosion earthquakes recorded. Strong tremor recorded 13-15 and 23-24 September with increased activity. Possible lava flow observed in October. 09 September, 1976 - 04 December, 1976: Period of weak ash emissions including several episodes of stronger activity. Ash fall (light) at Sand Point on 09 September. Harmonic tremor recorded on 09 September, 18-22 October, and 10-22 November. Possible lava flow observed in December, may be spatter-fed or a lahar. Many explosions recorded, including some strong enough to be felt in Pavlof Bay (15-20 km)."
Shackleford (1977) summarizes the 1975 activity as follows: "Ash eruptions began from Pavlof on 13 September, 1975. There was no preceding rise in microearthquake frequency. Strongest activity took place in the months of September-October, ash columns achieving a height of 2400 m. A lava flow was reported moving down the N flank on 31 October, but similar observations in 1973 and 1974 were subsequently shown to be in error, so this report may be in error also. It may well be a lahar. Activity increased somewhat in December, with sporadic appearances of lava fountaining amidst the ash clouds. It was reported that 18:15 GMT 28 December saw a 30 sec. burst of lava from Pavlof that resembled a blow torch, was visible in broad daylight. From 03:55 to 04:10 GMT on 31 December a continuous series of lava pulses surged to heights of at least 150 m above the cone. One observer likened the activity of Pavlof to the chugging of a locomotive, rather than a steady state ash emission.
"Harmonic tremor has been recorded during eruptive periods, as well as explosion signatures. Some of the explosions have been strong enough to rock fishing boats 10-15 km away, in Pavlof Bay."
Matthew Sturm sent AVO an eyewitness account of the October 1975 lava flow, which he observed on October 7. His journal for that day reads: "During the night run the watch saw numerous red distress flares south of the ship and diverted to investigate. As they worked south, it became clear the flares were being sent up well inland on the Alaskan Peninsula, at which point it dawned on someone that perhaps it wasn’t flares at all. A little bit of trigonometry and it was soon clear that we were seeing Pavlof Volcano erupting. But just to be sure we stayed around and as the sun rose we could see a huge cloud above the 8200’ peak with red sparks embedded in the black cloud, while on the northern slopes below the summit we could make out black ash and lava spilling over the snow and glaciers on the upper reaches of the mountain. The volcano would be peaceful for a few minutes, then suddenly a cloud would jet up and billow out for three or four thousand feet above the mountain. The entire crew was mesmerized by the sight and we stayed there for over an hour watching."
Shackleford (1978) summarizes the 1976-1977 activity as follows: "Strombolian eruptions which began in September 1975 continued throughout 1976 and were continuing in 1977. In 1976 activity was at a somewhat lesser level than the preceding year. Possible lava flows reported in February and December." Please see Shackleford (1978) for a month-by-month description of activity.
Shackleford (1979) continues to summarize the activity in 1977: "Reports of the 1977 Pavlof activity are extremely sketchy and incomplete, in part due to very poor weather conditions. Since March 1977 there have been no reports of activity (which may well have continued beyond March) due to a lack of reporting personnel, rather than a lack of activity." Please see Shackleford (1979) for a month-by-month summary of 1977 activity.
McNutt (1999) calculates an eruptive volume of greater than 14.6 x10^6 cubic meters (dense rock equivalent) during the time period from September 13, 1975 through November 10, 1977.

From McNutt and others (1991): "The visual observations of the 1986 eruptions, both aerial and on the ground, provide information generally lacking from previous historic eruptions. The physical characteristics of of the 1986 eruption are probably similar to other historic summit eruptions, although the 1986 activity was more explosive and of longer duration.
"The 1986 activity was chiefly Strombolian, characterized by sporadic emissions of dark ash to heights of up to 5 km; one exceptionally strong (probably Vulcanian) eruptive event sent an ash column to over 15 km on 18-19 April. The initial phase of the eruption appears to have involved the summit vent on the north side of the volcano which has been the site of all Pavlof eruptions since the mid-1960s. This eruptive phase lasted from 16 to 26 April and included a hot rootless agglutinate flow that extended down the northwest flank of the volcano. The deposit was actively degassing steam along its entire length in late June. It had an estimated volume of about 4x10^6 cubic meters and, at an elevation of 760 meters, was 20-30 meters thick and 40-50 meters wide. The deposit was clast-supported and fines-depleted, consisting entirely of glassy, slightly vesicular andesitic basalt (SiO2 = 53.5%) bombs and irregular masses of spatter with a maximum diameter of about 1.2 meters. It was probably emplaced during the initial and relatively violent vent-clearing phases of the eruption on 16-28 April. Associated hot lahars caused melting of snow and ice and extensive flooding in the Cathedral River drainage north of the volcano. Heavy ash fall occurred north and west of the volcano during the April activity; 2-3 mm of ash fell on the nearby communities of Cold Bay, 35 km to the west, and King Cove, 48 km to the southwest, but caused no damage.
"Eruptive activity began again in late May and was highlighted by sporadic, but vigorous Strombolian eruptive activity and by the formation of a new vent high on the east flank of the volcano. Comparison of aerial photography indicated that this is the first major change in the vent geometry of Pavlof since the early 1960s. The period of activity was characterized by repeated small bursts of ash and cinder to a few hundred meters above the vent and spatter tossed a few tens of meters all accompanied by explosive, thunderlike reports. The explosions characterizing the Strombolian eruption occurred at intervals of 5-15 seconds during the half dozen times the new vent was observed between 14 June and 30 June, suggesting a moderately rapid rate of magma rise in the conduit. Only minor ash emission was observed associated with the explosive ejection of incandescent bombs from the new vent. Activity from the old north vent during this time consisted of steam emission with little or no ash.
"Close airborne examination of the new vent on 15 June revealed that a steep-sided, asymmetrical spatter rampart, 50-75 m across, had been constructed on the east, or downslope, side of the vent. A steep chute, directly below the spatter rampart, contained a steaming, rootless rubble flow. Further down the volcano at an elevation of about 750 m, this flow was about 100 m wide and consisted of bread-crust-like bombs and irregular masses of lava and spatter, up to 4 m in diameter, in an ash-rich, moderately inflated matrix (in contrast to the flow on the northwest flank). It was actively degassing, exhibiting abundant steam fumaroles and occasional phreatic eruptions. The flow, with an estimated volume of 3.8 x 10^6 cubic meters, also generated a number of mudflows that continued downslope to about 600 m where the rubble flow-mudflow complex widened into three broad lobes. Below this elevation, debris-laden water from the flows was contained in a steep-sided small canyon.
"The nature of the material in the flow at this elevation and the physical characteristics of the spatter rampart at the vent strongly suggest that much of the lower part of the flow may have resulted from similar pyroclastic flow activity following partial collapse of the oversteepened spatter rampart. Throughout the course of the eruption, the steep spatter rampart may have periodically become unstable and collapsed, either in whole or part, forming hot, disaggregated pyroclastic flows cascading down the chute. Where ice and snow were overrun by the hot debris, mudflows were generated that continued to travel down and fan out on the volcano's lower slopes. In support of this hypothesis, a pyroclastic flow was observed on 19 June moving down the same chute from about 1400 m elevation to about 900 m; the upper 1100 m of the volcano was cloud-covered at this time.
"An alternative scenario is that the pyroclastic flow observed on 19 June resulted from a 'boil-over' of the magma column at the vent following a larger-than-usual explosion. Although no seismic evidence for such an explosion was observed, the obscuring cloud cover prevents an exact explanation for the origin o fthe observed pyroclastic flow.
"The April-August eruption is similar to, but somewhat stronger than most of the other Pavlof eruptions of this century (Simkin and others, 1981; McNutt 1987a) in terms of its Strombolian character. The duration of the eruption, as indicated by lava fountain activity and spatter ejection, however, appeared to be much longer than for most modern eruptions. This eruption also altered the physiography of the summit area by forming a new vent. In contrast to most recent eruptions (i.e., 1973-1983) that occurred in the fall of the year, the 1986 eruption was concentrated in the spring and summer."
McNutt and others (1991) also give detailed information about the seismological data collected during this eruption.
Information on the continuing volcanic activity in 1987 and 1988 is in other sources. From Reeder (1990, pg. 53): "Numerous observations of tephra and steam emissions were made during most of 1987 from NE and SE near summit vents and less from a NE flank vent. Lava flows occurred on both the NE flank and the SE flank of the volcano from the summit vents down to at least a 1,000 m a.s.l. elevation in January and February, form the NE summit vent in May and June, and again from both summit vents in August." Please see the rest of this text for detailed eruption observations in 1987.
From Reeder (1991): "Pavlof volcano has been fairly active during 1986 and 1987 with tephra and lava emissions occurring form several summit and flank vents. During 1988, eruptive activity was restricted to only small tephra and steam emissions from NE near summit vent up through August 13, except for one minor steam emission from a NE flank vent on March 2. Since August 13, 1988, no eruptive activity was observed for the rest of the year." Please see the rest of this text for detailed eruption observations in 1988.

From Neal and McGimsey (1997): "Pavlof Volcano, historically the most active volcano in the Wrangell-Aleutian volcanic arc, began a vigorous strombolian eruption in mid-September, 1996 (Neal, 1996). The eruption, which continued into early 1997, occurred only two months after a 6-station seismic network was established near the volcano.
"A NWS observer in Cold Bay noted steam and incandescent ejecta above the volcano at about 0830 ADT on September 16. Analysis of seismic data and NOAA-12 and -14 AVHRR satellite images suggest that the eruption likely began at a very low level by September 11. Over the next few weeks, nearby residents observed intermittent strombolian eruptions from near the summit of the volcano. Pilots reported incandescent bombs the size of pick up trucks accompanied by minor ash clouds alternating with steam plumes rising from a few hundred meters to approximately 2 km above the volcano.
"Photographs from overflights on September 23 and AVO video from September 27-30 showed lava fountains emanating from two vents (figs. 4A and 4B). One vent was located on the east edge of an ~150-m diameter crater that indented the northwest summit of the volcano. A second, more active locus of fountaining was perched on the west edge of this crater 100-150 m below the summit. The two loci of fountaining were about 100 m apart and were generally not synchronous in activity. The east vent was less vigorous overall, producing intermittent puffs of gray to dark gray ash and steam tens of meters high. The west vent was the source of intermittent bursts of incandescent spatter up to 300 m high.
"By September 23, a small spatter cone was forming at the west vent and a collar of spatter, spatter-fed flows, and small lahars extended about 500 meters down the ~30 degree northwest flank below the summit crater. A lava flow formed by the coalescence and remobilization of heavy spatterfall and direct spill over from the west vent plunged down the steep flank, melting a narrow channel through seasonal snow and glacial ice. By September 29, the lava flow had reached the base of the cone, about 3.5 km from its source, and was beginning to widen into a lobate fan. Dark lahar deposits extended beyond the toe of this lava flow across the gently sloping ground northwest of the volcano, coming within about 40 m of AVO's seismic station PV6. By late October, a second lava flow issued from the east vent and on December 2, when videotaped by Alaska State Troopers, this flow was the more active of the two and had nearly reached the base of the cone in the saddle between Pavlof and Pavlof Sister.
"Eruptive activity became intermittent during the month of December. Seismicity decreased abruptly early on December 4 and ash was not visible above the regional cloud cover that obscured the summit of Pavlof for several days. Brief episodes of heightened seismicity occurred on December 10 (accompanied by at least one pilot report of ash) and December 27. The last reliable observation of ash emission occurred on January 3, although pilots and observers in Cold Bay reported possible minor ash in the steam plume over the volcano on a few occasions through February 6. Collapse of unstable agglutinate and hot fragmental debris on the steep upper cone may well account for some of these small ash plumes.
"During the first two weeks of the eruption, occasional elongate clouds (<50 km long) containing minor amounts of ash were detected on NOAA AVHRR satellite images (fig. 5). During the third week, both pilot reports and satellite image analysis documented larger but still diffuse ash clouds trailing as far as 175 km downwind, but they rarely reached more than ~6 km above the sea level. These clouds varied in length from a few tens to several hundred km and were observed intermittently, weather permitting, through late December. On November 4, accompanying some of the strongest seismicity of the eruption, a plume was visible in Bands 4-5 extending 350 km northeast of the volcano.
"In addition to elongate plumes, thermal anomalies associated with high temperature material were also recorded near the volcano's active vents and along the two main lava flow paths. The number of saturated pixels on AVHRR images varied from 1 - 15 indicating areas of up to about 18 km2 above 37 degrees C (A.L. Roach, oral communication, 1997). The last significant thermal anomaly was recorded in late December, however "warm" pixels were noted during daily analysis of AVHRR data into mid-February. Pilot reports and observations from Cold Bay confirm continued warm ground around the summit of the volcano as inferred from areas of snow-melt.
"As in the 1986 eruption, the 1996 activity produced rubbly, fragmental lava flows that extend in two main lobes down the northwest flank of the volcano (fig. 6). Early in the eruption, these flows occupied, at least in part, channels cut into the seasonal snow and glacial (?) ice on the volcano's flank. Melting of this snow and ice produced water and rock mixtures of unknown consistency that flowed out onto the more gently sloping terrain northwest (and possibly northeast) of the volcano. As of this writing, we do not know how far these lahars traveled or what impact they had on the Cathedral River and other drainages around Pavlof.
"Very light ashfall was reported in King Cove on the night of October 5-6, Sand Point on October 19, and Nelson Lagoon on October 28."
More from Neal and McGimsey (1997): "On October 3, based on observed plume heights, the FAA issued a Notice to Airmen (NOTAM) restricting flight below ~7 km and within 10 nautical miles of Pavlof. Higher levels of seismicity and more energetic ash plumes began on October 15 and in response, the FAA increased the altitude of restricted air space to approximately 8 km and the size of the restricted zone to a 25 mile radius around Pavlof. The FAA continued to enforce this restriction until January 27, 1997. Although Pavlof ash plumes reached altitudes of 30,000 feet or more on a few occasions, there were no serious disruptions in the North Pacific airways.
"There were, however, impacts on local air traffic. On November 4, a United States Coast Guard (USCG) C-130 operating at low level over the Bering Sea was struck by lightning. The flight crew also reported a "smoky" smell in the cockpit and a fine dust throughout the plane. Subsequent discussion with the USCG failed to positively identify the source of this material. However, based on NWS forecast winds during the time of this report, it appears unlikely that primary ejecta from Pavlof could have been the culprit; rather, it is possible that low level winds remobilized fine ash from the ground. No sample of the material was recovered for analysis.
"On November 27, 1996, a Reeve Aleutian Airways flight aborted landing into Sand Point when the pilot detected a brown haze that he interpreted to be ash from Pavlof."
McNutt (1999) calculates a dense rock equivalent volume of material erupted during this eruption as ~7 x10^6 cubic meters.

From McGimsey and others, (2011): "Pavlof, one of the most frequently active volcanoes in the Aleutian arc, abruptly erupted on August 14, 2007, following an 11-year repose [see fig. 25 in original text]. The 31-day Strombolian eruption was preceded by less than 1 day of increased seismicity detected on the AVO seismic network, and produced a spatter-fed lava flow, minor ash clouds, and lahars that extended down the south flank into the sea. The following account is drawn in part from Waythomas and others (2008).
"The previous eruption of Pavlof Volcano was September 1996-January 1997 (Neal and others, 1997; McGimsey and Wallace, 1999). Minor activity (snowmelt, ash dustings, steam plumes, sulfur emissions) was noted in 1999 and 2001, and in 2005, steaming was observed at adjacent Hague volcano [see table 4b in original text]. Nothing unusual was observed during the summer of 2007 and the seismicity was at background levels through August 13. Abruptly on the morning of August 14, the 5-station seismic network on Pavlof began recording low-frequency earthquakes occurring at a rate of 2 to 7 events every 10 minutes, a pattern that had preceded eruptions in 1996, 1986, 1983, and 1981 (McNutt, 1987; McNutt, 1989; Roach and others, 2001). Although NWS observers in Cold Bay (37 mi southwest) with clear views of the volcano could see no anomalous steaming or other activity, and clear satellite views of the summit that morning also showed no signs of activity, the dramatic increase in seismicity prompted AVO to raise the Aviation Color Code/Volcano Alert Level to YELLOW/ADVISORY.
"During the night, an intense thermal anomaly (TA) was visible in satellite images (Advanced Very High Resolution Radiometer-AVHRR), and seismic activity continued to increase in both number and duration of events per hour, clear signs that the unrest was escalating. On the morning of August 15, based on observations of the TA and increasing seismicity, AVO elevated the Aviation Color Code/Volcano Alert Level to ORANGE/WATCH and announced that an eruption was expected. With the upgrade in color code, AVO began 24-hour surveillance of the volcano. Later in the day, AVO received eyewitness accounts from mariners of incandescent blocks rolling down the eastern-southeastern flank of the volcano during the previous night, beginning around midnight. Pilots reported a thin, low-level ash plume extending a few kilometers southwest from the summit. After receiving these reports, AVO established that the volcano was in eruption. Aerial photographs taken on August 15 show lava fountaining from a vent located about 200 m (650 ft) below the summit [see fig. 26 in original text].
"On August 16, strong seismic signals recorded at a single station (PVV), located 8.5 km (5.3 mi) southeast of the summit, heralded the passage of lahars down the south flank; more than 41 lahar events would be recorded by this station over the next 29 days. Satellite observations of a strong thermal anomaly (TA) [see fig. 27 in original text] and nighttime incandescence at the summit reported by local residents were indications of vigorous lava eruption at the summit vent [see fig. 28 in original text]. The seismic network recorded long periods of volcanic tremor with repetitive explosions that indicated nearly continuous Strombolian eruption. In addition to the generation of lahars, this activity produced low-level ash clouds (5-6 km ASL; 3.1-3.7 mi), and a spatter-fed lava flow that descended the southeastern flank. By August 18, AVO personnel in the field reported that vigorous eruption of lava at the summit continued. Using a Forward Looking Infrared (FLIR) camera, they determined that a 20- to 50-m-wide, 65- to 165 ft-wide) 600 °C (1,112 °F) lava flow extended 565 m (1,850 ft) from the vent down the southeast flank [see figs. 29 and 30 in original text]. Thermal data collected the next day indicated that the outer part of this flow was about 140 °C (284 °F) and had cooled considerably. The vent crater for the last eruption of Pavlof, in 1996, was located on the upper northwestern side of the summit. For this eruption, the active vent migrated to the upper southeastern side, about 200 m (650 ft) below the summit [see figs. 31-33 in original text].
"Seismicity at Pavlof was elevated and steady throughout the remainder of August and then began waxing and waning for the first week of September. A strong TA was present in satellite images, even through clouds, during this time. During the second week of September, the seismicity began showing signs of a steady decrease [see fig. 34 in original text], and by September 13, seismicity decreased to low levels and only a minor steam plume was visible above the volcano. A TA was last seen on September 15, and AVO declared that the activity had reached a lull by September 17. An AVO field crew with clear views reported that all eruptive activity had ceased during their visit on September 19, and the Aviation Color Code /Volcano Alert Level was downgraded to YELLOW/ADVISORY on September 20. The next 2 weeks of low seismicity and no further signs of activity or unrest prompted AVO to declare the eruption over (ending on September 13), and the Color Code/Volcano Alert Level was downgraded to GREEN/NORMAL on October 5.
"Ash, a blocky lava flow, and multiple lahars were generated by this eruption. Mixed ash and steam clouds produced during the most energetic eruptive period, mid-August to mid-September, reached altitudes of 5-6 km (about 20,000 ft) ASL. The plumes were diffuse, drifted primarily to the southeast over the North Pacific Ocean, and many could not be detected in satellite imagery. No ash reportedly fell on nearby communities and there were no significant impacts to aviation. AVO deployed a DRUM aerosol impactor (particle collector) in Sand Point, 90 km (56 mi) east of Pavlof, and collected fine ash (2.5-0.1 µm). Although no visible ash fallout was observed during aerosol sampling, these results demonstrate that volcanic ash was present in respirable size fractions downwind of the volcano even during periods of low ash emissions (Peter Rinkleff and Cathy Cahill, AVO/UAFGI, written commun., 2010).
"Analyzed samples from the lava flow are basaltic andesite in composition (53% SiO2), which is similar to the products of previous Pavlof eruptions (McNutt and others, 1991; Neal and McGimsey, 1997). Lahars were produced by interaction of hot blocks and spatter from the lava flow with snow and ice on the southeastern flank. The lahars inundated an area over 2 km2 (0.78 mi2) and formed a debris fan that extended 3.6 km (2.2 mi) from the base of the volcano into Pavlof Bay [see fig. 35 in original text]."

From Dixon and others (2015): "Pavlof Volcano erupted in May 2013 and was characterized by Strombolian explosions and periods of continuous tremor. Eruption plumes deposited trace amounts of ash in nearby communities during the first 2 weeks of the eruption and again in early June. Activity and observations are summarized in table 7 [in original text]. Pavlof was upgraded from the Aviation Color Code and Volcano Alert Level of GREEN/NORMAL to ORANGE/WARNING on May 13, remaining at ORANGE/WARNING for 17 days during the 3-week-long eruption, with remainder of the time at YELLOW/ADVISORY. On August 8, the Aviation Color Code and Volcano Alert Level were downgraded to GREEN/NORMAL, where it remained for the rest of the year.
"The 2013 eruption of Pavlof Volcano began on the morning of May 13, 2013, following a 6-year period of repose. Eruption onset was characterized by subtle, low-level seismicity beginning about 16:00 UTC (08:00 AKDT) and continuing for the first 24-48 hours of the eruption. A strong, persistent thermal signal from the Pavlof summit area was first observed in mid-infrared AVHRR satellite imagery at 15:17 UTC (07:17 AKDT) on May 13, and AVO upgraded the Aviation Color Code and Volcano Alert Level to ORANGE/WATCH, stating that an eruption was likely to progress. A pilot report at 03:00 UTC on May 13 (19:00 AKDT, May 14) confirmed the eruption, with numerous dark streaks on the upper northern flank of the volcano that appeared to be a lava flow and lahars initiated by melting of snow and ice in the summit area (fig. 28 in original text). Similar flows also were observed farther down the northern flank and that were initiated by the ejection of hot debris onto snow and ice. Residents of Sand Point, 85 km (53 mi) east of the volcano, reported seeing a distinctive glow at the summit of Pavlof during the evening of May 13, indicating likely lava fountaining. By the next day, satellite observations showed that a lava flow extending down the northern flank well beyond the summit vent. The flow was estimated to be about 600 m (2,000 ft) long and 30 m (100 ft) wide, and originated from a vent within a small crater just north of the summit.
"AVO received numerous observations on May 14 confirming eruptive activity underway at Pavlof Volcano. Pilot reports and Web camera views of the volcano indicated that ash emissions as high as about 4 km (13,000 ft) ASL were occurring intermittently. Views of Pavlof, such as that in figure 29 [in original text], indicated several light-colored plumes rising off the lower northern flank, suggesting flowage of hot debris over ice and snow. Throughout the day on May 14, numerous strong bursts of tremor coincided with similar observations of large steam plumes rising off the northern flank of the volcano. Light-colored plumes from some of these events reached as high as 6 km (20,000 ft) ASL. Incandescence associated with strong lava fountaining at the summit also was observed throughout the evening of May 14. The lava fountaining was robust enough that relatively continuous infrasonic tremor was produced and recorded on infrasound arrays on Akutan Island (290 km [180 mi] southwest of Pavlof) and Okmok (460 km [285 mi] southwest of Pavlof) Volcanoes and at Dillingham (455 km [283 mi] northeast of Pavlof).
"Noticeable fallout of fine ash occurred as far as 80 km (50 mi) downwind of the volcano on May 14, 15, and 18, and was reported to AVO. During May 15-20, sulfur dioxide (SO2) from Pavlof was detected in OMI satellite data; the estimated SO2 mass during May 15-16 was 1,000-2,000 metric tons and at least 4,000 metric tons on May 18-20 (Simon Carn, written commun., 2013). Sulfur dioxide also was detected on multiple days by the GOME-2 (Global Ozone Monitoring Experiment-2) and IASI (Infrared Atmospheric Sounding Interferometer) instruments.
"From May 15 to 21, the eruption was characterized by nearly continuous tremor and explosions. A relatively continuous ash plume was apparent in AVHRR and MODIS satellite images and also was observed by pilots. Ash emissions during this period reached as high as 7 km (23,000 ft) ASL and extended about 400-500 km (250-310 mi) southeast of the volcano on May 18 (fig. 30 in original text). On May 19, trace amounts of ash fall occurred on the communities of Sand Point and Nelson Lagoon, 90 km (56 mi) southwest and northeast of Pavlof, respectively. AVO scientists examined an ash sample collected in Sand Point that consisted almost entirely of dark angular glass shards. A single electron microprobe analysis of the glass indicated that it is compositionally andesite (58 percent silicon dioxide, SiO2) and similar to ash deposits associated with previous historical eruptions (K. Wallace, USGS-AVO unpub. data; microprobe analyses by L. Hayden, USGS).
"From May 22 to June 4, the volcano was relatively quiet, and the seismicity during this period was characterized by episodic, discrete bursts of tremor lasting from 30 seconds to approximately 1 minute. During May 22-23, the Pavlof seismic network detected distinct ground-coupled airwaves. Infrasonic arrays at Dillingham and Okmok Volcano also recorded these probable explosion signals as impulsive infrasonic waves.
"From May 27 to June 4, seismic tremor and small discrete explosions were no longer detected in seismic and infrasound data. Satellite observations during this period showed no evidence of elevated surface temperatures, volcanic gas (SO2) or ash emissions. During periods of clear weather, no visual observations of ash emissions and Web camera views of the volcano were noted, indicating eruptive activity had paused. These observations prompted AVO to downgrade the Aviation Color Code and Volcano Alert Level to YELLOW/ADVISORY at 20:50 UTC (12:50 AKDT) on May 27.
"On June 4, 2013, AVHRR, MODIS, and GOES satellite data detected ash emission from Pavlof; passing pilots reported ash plumes as high as 5.7 km (18,700 ft) ASL. Slightly elevated levels of seismic tremor also were observed by midday (local) on June 4, roughly coincident with the observations of ash emissions, prompting an upgrade to ORANGE/WATCH at 12:15 UTC (20:15 AKDT). From June 14 to 19, seismic activity was characterized by periods of intermittent volcanic tremor and slightly more robust and more frequent explosions compared to the character of the seismicity from May 13 to 24. During this period, ash plumes generally were smaller and did not extend more than about 50 km (30 mi) downwind of the volcano. The maximum plume height reported by pilots was approximately 6 km (29,000 ft) ASL on June 10. Residents of Cold Bay reported barely perceptible trace ash fall during June 6-7.
"From June 20 to 24, the Pavlof seismic network recorded moderate levels of relatively continuous tremor and small explosions. Several low-level ash plumes (generally less than 3.5 km or 11,500 ft) ASL were generated, although cloud cover occasionally inhibited observations.
"Beginning around 07:00 UTC on June 25 (23:00 AKDT on June 24), tremor amplitudes at the volcano increased significantly and were characterized by high levels of continuous tremor and frequent explosions associated with robust episodes of lava fountaining and ash emission. The level of seismicity on June 25 was the strongest detected during the entire eruption. Observers in Sand Point reported incandescence and ash plumes as high as 6 km (20,000 ft) ASL on the morning of June 25, and around midnight on June 25, ash fall was reported in King Cove 50 km (30 mi) southwest of the volcano. Analysis of satellite images and pilot reports confirmed ash plumes as high as 7-8 km (23,000-26,000 ft) ASL. Sulfur dioxide emissions from Pavlof were detected by the Joint Polar Satellite System Ozone Monitoring Profiler Suite from 01:00 to 23:10 UTC on June 25. Automated analysis of this data indicated a mass of SO2 of 6,000-7,000 metric tons near the volcano.
"Between June 25 and 27, intermittent ash plumes rose to 6 km (20,000 ft) ASL. After June 27, the seismicity became less energetic, and occasional low-frequency events and low levels of increasing discontinuous tremor characterized the seismicity. Pilot reports on June 28 indicated no activity at the volcano, and over the next several days, seismic tremor and small discrete explosions were no longer detected in seismic and infrasound data. Satellite observations after July 1 showed no evidence of elevated surface temperatures, volcanic gas, or ash emissions. On July 2, AVO downgraded the Aviation Color Code and Volcano Alert Level to YELLOW/ADVISORY, where it remained until August 8 when the volcano returned to its normal background state and the Aviation Color Code and Volcano Alert Level were downgraded to GREEN/NORMAL.
"The lahars and ash plumes generated during the eruption did not pose any serious hazards for the area. However, numerous local airline flights were cancelled or rerouted, and trace amounts of ash fall occurred at all local communities surrounding the volcano, including Cold Bay, Nelson Lagoon, Sand Point, and King Cove. Observations by AVO scientists during July 16-17 indicated that only the upper part of the Cathedral River drainage (fig. 31 in original text) had been inundated by lahars. However, a fountain-fed lava flow, about 5.8 km (3.6 mi) in length, covering an area of about 730,000 m2 (180 acres) on the northern flank of the volcano was observed (fig. 32 in original text). It was only possible to collect a few samples of the lava. These samples have not yet been analyzed (as of July 2015), but appeared similar to other andesitic lava flows produced by historical eruptions of Pavlof."

From Cameron and others, 2017: "The first eruptive episode at Pavlof in 2014 began at about 19:00 UTC (11 a.m. AKDT) on May 30, based retrospectively on a pulsating tremor-like signal in 1-5 Hz bandpass seismic data. The seismicity was visible across the Pavlof seismic network. By 07:22 UTC on May31 (11:22 p.m. AKDT, May 30), satellite data showed a strong thermal signal, suggestive of lava at the surface. AVO upgraded the Aviation Color Code and Volcano Alert Level from GREEN/NORMAL to ORANGE/WATCH at 19:36 UTC (11:36 a.m. AKDT) on May 31. Subsequent analysis indicated that the opening phase of the eruption on May 31 was recorded on an infrasound network on Akutan Island about 278 km (173 mi) southwest of Pavlof. AVO received the first pilot reports of ash emission on June 1, indicating distinct ash clouds as high as about 3km (2 mi) ASL, drifting north-northeast as much as 80 km (50 mi) beyond the summit of the volcano. Lahar (mud flow) signals were evident in seismic data from station PV6 by late in the day on June 1. These initial flows lasted for 15-30minutes each, and were preceded by sustained low-level pulsatory tremor.
"At about 01:30 UTC, June 3 (5:30 p.m. AKDT on June 2) , the amplitude of the seismic tremor increased significantly; Web camera, satellite views, and several pilot reports all indicated that a period of robust ash emission was underway. At this point, AVO upgraded the Volcano Alert Level and Aviation Color Code from ORANGE/WATCH to RED/WARNING, where it remained for about the next 24 hours. During this period, the volcano erupted almost continuously, and observers in Cold Bay reported incandescence at the summit, seismic stations recorded lahar signals, seismic tremor remained at high levels, and strong thermal signals were evident in satellite data. The highest ash plume generated during this period of heightened activity reached as high as 6,700 m (22,000 ft) ASL (based on pilot reports) and extended about 100 km (60 mi) to the southeast over Sand Point and Unga Island. Although there were no reports of ashfall in nearby communities on June 2-3, significant SO2 emissions were detected in infrared atmospheric sounding interferometer (IASI) satellite data on June 3.
"Beginning at about 06:30 UTC on June 3 (10:30 p.m. AKDT on June 2), the overall level of seismicity began to decrease slowly, and at 01:54 UTC on June 4 (5:54 p.m. AKDT on June 3), AVO downgraded the Aviation Color Code and Volcano Alert Level to ORANGE/WATCH in response to the decreased levels of seismicity. Although the tremor level was reduced, the overall level of seismic activity remained relatively steady throughout the day on June 3, and was associated with a mostly steam-and-gas plume, containing only minor amounts of ash reaching as high as 5,900 m (19,600 ft) ASL and drifting to the south. Trace amounts of fine ash were reported at the Sand Point airport on June 3-4, resulting in cancelation of flights to Sand Point on the morning of June 4. As many as five flights to Cold Bay and nine flights to Dutch Harbor also were canceled on June 3-4 because of the risk of encountering ash. Local commuter air service was suspended as well during this time, affecting air travel to King Cove and False Pass. Flight cancellations continued into June 5, disrupting flights to and from Dutch Harbor and Cold Bay.
"The activity on June 2-3 was characterized by periods of vigorous lava fountaining that resulted in the accumulation of lava spatter on the upper northern flank of the volcano. This accumulation resulted in the formation of a spatter-fed lava flow that eventually extended about 4.7km (2.9 mi) downslope. At other times during the eruption, accumulations of spatter grew and built unstable piles of hot, fragmental material that occasionally collapsed and formed hot granular avalanches that flowed rapidly down the northern flank of the volcano for several kilometers. These hot granular rock avalanches swept across snow and ice, producing impressive steam plumes and generating melt water that led to the formation of lahars in some of the main drainages on the northern flank of the volcano. No ash or steam plumes were evident in satellite data after June 4, although seismic data recorded two explosions on the morning of June 5, and the World Wide Lightning Location Network (WWLLN) detected lightning near Pavlof from 10:16-10:59 UTC (2:16-2:59 a.m. AKDT) on June 5. Meteorological lightning is unusual in this part of Alaska, so the lightning likely was related to volcanic ash generated by the explosions. Emission of SO2 also was detected in OMI satellite data on June 5, and the SO2 plume extended from the volcano west about 100 km (60 mi).
"On June 6, the level of seismic activity decreased appreciably, relative to the previous several days. From June6-25, the level of unrest at the volcano continued to decrease further, and local observations of low-level lava fountaining were reported intermittently until about June 14. By June 25, satellite and Web camera data showed no evidence for lava fountaining or ash emissions. Only weakly elevated surface temperatures near the new lava flows on the northeastern flank were evident in satellite data. Consequently, AVO downgraded the Aviation Color Code and Volcano Alert Level to YELLOW/ADVISORY on June 25. By July 29 (AKDT) seismicity at Pavlof had returned to its normal background status and AVO downgraded the Aviation Color Code and Volcano Alert Level to GREEN/NORMAL."

From Cameron and others, 2017: "Pavlof remained quiet and at background levels of seismicity until November 12, when seismic tremor increased sharply and increased surface temperatures were detected in satellite data, suggesting that lava had reached the surface. This was corroborated by reports of minor ash emissions and low-level lava fountaining from observers in Cold Bay. Observers also reported flows of rock debris and ash descending the northern flank of the volcano, and incandescence was observed in Web camera images. As a result of these observations, AVO upgraded the Aviation Color Code and Volcano Alert Level to ORANGE/WATCH on November 12.
"After November 12, the level of seismic activity continued to increase gradually and the thermal signal at the summit became persistently visible in satellite data. On November 13, satellite data showed a 200-km-long (124-mi-long) ash plume extending northwest of the volcano.Pilot reports estimated the ash cloud top at about 2,400-2,700m (about 7,900-8,900 ft) ASL. For the next 24 hours, all Pavlof seismic stations recorded nearly continuous seismic tremor, and satellite data showed strongly elevated surface temperatures, consistent with sustained lava fountaining. A narrow ash plume extending as high as 200 km (124 mi) from the volcano continued to be visible in satellite data, and information from passing pilots indicated that the ash plume eventually reached an altitude of about 4,800 m (15,700ft) ASL. Minor SO2 emissions were detected on November 14 in satellite data.
"On November 15, the intensity of seismic tremor increased significantly over a 6-hour period and satellite data indicated that the ash cloud, visible for the previous several days, had expanded and reached an altitude of about 7,600m (25,000 ft) ASL, and extended at least 200 km (124 mi) northwest of the volcano. In response to this increase in eruptive activity, AVO upgraded the Aviation Color Code and Volcano Alert Level to RED/WARNING. This plume eventually reached an altitude of at least 11,000 m (36,000ft)
ASL and extended about 385 km (240 mi) northwest of the volcano. Although the volcano was obscured by clouds, observers in Cold Bay reported rumbling and thunder-like sounds coming from the direction of Pavlof Volcano. Infrasonic tremor levels detected by the infrasound array in Dillingham increased steadily through the day on November 15 and were as high as, or higher than, infrasound levels recorded at any time during Pavlof’s May-June 2014 or 2013 eruptions (David Fee, UAFGI/AVO, written commun., 2014). The SO2 plume generated by this phase of the eruption extended west over the Bering Strait and into eastern Russia.
"Seismicity at the volcano decreased significantly on November 16 and remained at low levels, and ash plumes were not observed in satellite data thereafter. In response to thedecrease in seismicity and ash emission, AVO downgraded the Aviation Color Code and Volcano Alert Level to ORANGE/WATCH on November 16.
"Satellite observations indicated that eruptive activity ceased by about November 17. The intensity of thermal signals decreased gradually and the levels of tremor fluctuated slightly, but the overall amplitude decreased steadily. Thermal signals were occasionally observed in satellite datathrough November 26, but were the result of the still hot lava and debris on the northern flank of the volcano. As a result of the diminished levels of unrest, AVO downgraded the Aviation Color Code and Volcano Alert Level to YELLOW/ADVISORY on November 26, UTC (November 25, AKST).
"No further eruptive activity occurred at Pavlof Volcano in 2014, although unstable accumulations of cooling lava spatter occasionally collapsed, generating small ash emissions. The volcano gradually returned to normal background status, and on January 15, 2015, AVO downgraded the Aviation Color Code and Volcano Alert Level to GREEN/NORMAL."

From Cameron and others (2020): "Pavlof Volcano erupted on March 27, 2016, about 500 days after the end of the previous eruption in 2014 (table 5). The March 27-28 eruption was a brief but intense eruptive event lasting about one day (fig. 16). This event generated minor ashfall at Nelson Lagoon 77 km (48 mi) northeast of Pavlof Volcano and trace ashfall at Port Heiden and Dillingham 263 km (164 mi) and 453 km (281 mi) northeast of Pavlof, respectively. More than a hundred flights were cancelled between March 27 and March 29 because of ash from the eruption. Two additional eruptive periods occurred in May and July 2016, but these events produced only small ash clouds and minor ash fallout confined to the upper flanks of the volcano. The March 28 eruption generated significant ash clouds reaching as high as 12.5 km (41,000 ft) ASL, and pyroclastic flows and lahars on the flanks of the volcano, one of which destroyed seismic station PV6 on the north flank of the volcano.
"The first indication of unrest in 2016 began with a late morning telephone call on March 25 from National Weather Service personnel in Cold Bay, Alaska, who reported a steam plume rising from the summit of Pavlof Volcano (fig. 17). Although steam plumes at Pavlof are common, this particular plume was reported as more robust than usual. A review of web camera images from a Federal Aviation Administration (FAA) web camera in Cold Bay showed a prominent steam plume visible from 17:47 until 19:57 UTC (9:47 to 11:57 AKDT), when it became obscured by clouds. No unusual seismic activity was noted on March 25 or the following day on March 26.
"The volcano remained quiet throughout most of the day on Sunday, March 27, and a relatively clear satellite view of the volcano summit above the cloud deck at 22:33:15 UTC (14:33 AKDT) showed no evidence of any unrest. At 00:18 UTC, March 28 (16:18 AKDT), an ash cloud reaching about 6 km (20,000 ft) ASL and moving north was reported by a nearby pilot. At about 01:33 UTC March 28 (17:33 AKDT March 27), AVO received a pilot report of ash emissions from Pavlof Volcano reaching an altitude of about 9.1 km (30,000 ft) ASL and observations of lava at the surface. Retrospective analysis of seismic data indicated that seismicity began to increase from background levels at about 23:53 UTC (15:53 AKDT) denoted by an abrupt increase in real-time seismic amplitude measurement (RSAM) levels and the appearance of continuous tremor on all operating stations of the Pavlof Volcano network. The tremor and RSAM levels observed on station PS4A illustrate seismicity during the eruption (fig. 18). At about 00:00 UTC March 28 (16:00 AKDT) tremor and RSAM levels continued to increase until about 06:38 UTC March 28(22:38 AKDT) after which the RSAM values leveled off but continued to fluctuate. At approximately 09:46 UTC March 28 (01:46 AKDT) they increased toward a peak value of 3200 and then began to decline in a saw tooth pattern until 20:40 UTC (12:40 AKDT) when the values fell to about pre-eruption levels (fig. 18).
"The abrupt increase in RSAM and tremor observed around 00:00 UTC March 28 (16:00 AKDT March 27) prompted AVO to raise the Aviation Color Code and the Volcano Alert Level to RED and WARNING at 01:12 UTC March 28 (17:12 AKDT), respectively. The rapid increase in seismicity is characteristic of several recent Pavlof Volcano eruptions, and it is not uncommon for the Aviation Color Code and the Volcano Alert Level to move from background status to the highest level with minimal precursory seismic activity.
"At 04:05 UTC March 28 (20:05 AKDT March 27), AVO received a pilot report of lava fountaining near the summit and a flowage feature on the north flank of the volcano. According to the pilot, the flowage feature had apparently reached the Bering Sea coast. The pilot referred to the feature as a pyroclastic flow, and it is possible that pyroclastic flows produced by collapse of the eruption column formed during the eruption. It is also possible that meltwater generated by the interaction of pyroclastic flows with snow and ice may have flowed well beyond the volcano to the north as lahars. The specific drainage containing the flow was not identified by the pilot, but lahars from previous historical eruptions have inundated both the Leontovich and Caribou River drainages on the north side of Pavlof Volcano.
"Pyroclastic flows or hot granular mass flows associated with collapse of spatter accumulations on the upper part of the edifice likely destroyed seismic station PV6 on the lower north flank of the volcano (fig. 19). Flowage signals were evident at PV6 starting about 01:06 UTC March 28 (17:06 AKDT March 27), and the station stopped transmitting 12 minutes later at 01:18 UTC March 28 (17:18 AKDT March 27).
"Satellite observations from 04:15 UTC March 28 (20:15 AKDT March 27) indicated that the ash cloud from the eruption extended 180 km (108 mi) northeast beyond the volcano over the Bering Sea (fig. 20). Ash cloud heights from pilot reports indicated a maximum altitude of about 9.1 km ASL (30,000 ft). By 15:10 UTC March 28 (07:10 AKDT), the Pavlof Volcano ash cloud formed a narrow, continuous plume that extended for about 885 km (550 mi) from the volcano over interior Alaska as detected in a Himawari-8 false color image (fig. 21).
"Ash fallout on March 27–28 was reported in several communities northeast of Pavlof Volcano, including Nelson Lagoon, Port Heiden, and Dillingham, Alaska. In Nelson Lagoon 77 km (48 mi) northeast of Pavlof, 3–17 millimeters (mm; 0.125–0.66 inches) of dark ash fell, covering roofs and surfaces (fig. 22). Trace amounts of ash (<0.8 mm) were reported in Port Heiden and in Dillingham 263 km (164 mi) 453 km (281 mi) northeast of Pavlof Volcano, respectively.
"The March 27-28 ash plume caused the cancellation of 41 Alaska Airlines flights to and from Barrow, Bethel, Anchorage, Fairbanks, Kotzebue, Nome, and Prudhoe Bay, Alaska, on Monday, March 28, 2016. Regional flights operated by Bering Air were cancelled on the morning of March 28, PenAir suspended service to Dutch Harbor, Alaska, in the afternoon, and Ravn reported numerous flight cancellations (FOX59, 2016). Flight cancellations continued on Tuesday, March 29, and Alaska Airlines reported 28 cancelled flights to Barrow, Bethel, Kotzebue, Nome, and Prudhoe Bay, representing about 57 percent of Alaska’s flights to the most northern region it serves. Alaska Airlines additionally reported more than 6,200 travelers were affected on March 28 and 29 (Alaska Airlines, 2016).
"Lightning in the vicinity of Pavlof Volcano was reported by the WWLLN, which consisted of 16 lightning flashes detected over a 6-hour period between 13:10-19:13 UTC (05:10–11:13 AKDT) on March 28. Sulfur dioxide (SO2) emissions were detected by Infrared Atmospheric Sounding Interferometer (IASI) satellite sensors on the European Space Agency MetOp series of polar orbiting satellites on March 28 and 29. The SO2 cloud extended over interior Alaska and northwestern Canada (fig. 23) and eventually reached the southern Hudson Bay area of central Canada. The cumulative SO2 mass determined from the Ozone Mapping Profiler Suite (OMPS) satellite data obtained on 28 March was on the order of 20–30 kilotons (Simon Carn, Michigan Technological University, written commun., 2016).
"The March 28 Pavlof eruption was clearly recorded by infrasound instruments 453 km (281 mi) to the northeast at Dillingham (fig. 18). An infrasound signal was first evident at about 02:00 UTC March 28 (18:00 AKDT March 27), which corresponds to an emission time at the vent of about 01:40 UTC March 28 (17:40 AKDT March 27). The magnitude of the infrasound signals gradually rose until about 04:30 UTC March 28 (20:30 AKDT March 27) and then stabilized at high levels for several hours before declining significantly with no clear signals detected at Dillingham after about 21:00 UTC (13:00 AKDT) March 28 (David Fee, University of Alaska Fairbanks Geophysical Institute, written commun., 2016).
"By 20:40 UTC (12:40 AKDT) March 28, ash emissions were no longer evident in satellite data, and seismic activity had declined to nearly pre-eruption levels. AVO lowered the Aviation Color Code to ORANGE and the Volcano Alert Level to WATCH at 02:01 UTC (18:01 AKDT) March 29 because of the decline in robust eruptive activity.
"After March 28, 2016, unrest at Pavlof Volcano gradually declined. By April 6, 2016, ash emissions were no longer detected, and only weakly elevated surface temperatures associated with cooling deposits were observed in satellite data. At this point, AVO lowered the Aviation Color Code to YELLOW and the Volcano Alert Level to ADVISORY. Unrest continued to decline throughout the month of April 2016, and on April 22 AVO lowered the Aviation Color Code to GREEN and the Volcano Alert Level to NORMAL.
"Pavlof Volcano remained at background levels of unrest until May 13, 2016. At about 18:35 UTC (10:35 AKDT) May 13, seismic activity increased to levels commonly associated with low-level eruptive activity, suggesting that an eruption may have started. AVO responded by raising the Aviation Color Code to ORANGE and the Volcano Alert Level to WATCH at 20:04 UTC (12:04 AKDT) May 13. No volcanic activity was observed in satellite data or in web-camera views of the volcano on May 13. Minor ash emissions reaching as high as 6 km (20,000 ft) ASL were observed in images from the Cold Bay FAA web camera beginning around 03:46 UTC May 15 (19:46 AKDT May 14). Elevated surface temperatures were observed in satellite data on May 15, and SO2 emissions were detected in Ozone Monitoring Instrument (OMI) satellite data at 23:28 UTC (15:28 AKDT) May 17. AVO received several pilot reports of ash clouds rising to about 4.6 km (15,000 ft) ASL on May 17. This brief period of low-level eruptive activity lasted only a few days, and on May 20 AVO lowered the Aviation Color Code to YELLOW and the Volcano Alert Level to ADVISORY. After a few more weeks of no activity, AVO moved the Aviation Color Code to GREEN and Volcano Alert Level to NORMAL on June 17.
"The volcano remained quiet until July 1, 2016, when seismicity at the volcano began to increase, and minor steam emissions were observed in web-camera images. The increase in unrest prompted AVO to raise the Aviation Color Code to YELLOW and the Volcano Alert Level to ADVISORY.
"Beginning around 21:00 UTC (13:00 AKDT) July 12 web-camera images showed minor ash emissions rising just above the roughly 2.6 km (8,500 ft) high summit vent and extending a few miles to the southwest. Clear satellite views of the volcano from about this same time showed no evidence of ash emissions or elevated surface temperatures, and there was no anomalous seismicity associated with this low-level activity. After July 12, Pavlof Volcano remained in a state of low-level unrest with occasional slightly elevated surface temperatures and minor steam and diffuse ash clouds observed. The Aviation Color Code remained YELLOW and the Volcanic Alert Level ADVISORY.
"On July 28, minor eruptive activity was observed in web-camera and satellite images of Pavlof Volcano and was characterized by vigorous, steam-rich degassing and minor ash emissions. Pilots also reported seeing ash emissions as high as 2.7–3.6 km (9,000–12,000 ft) ASL. These observations coincided with a minor increase in seismicity to levels high enough to warrant raising the Aviation Color Code to ORANGE and the Volcano Alert Level to WATCH at 19:55 UTC (11:55 AKDT) on July 28. Observations of minor steam and ash emissions as high as 2.4 km (8,000 ft) ASL were made by pilots on July 31. Only weakly elevated surface temperatures were observed on August 2, and by August 4 there was no further evidence of unrest at Pavlof Volcano. Thus, AVO reduced the Aviation Color Code to YELLOW and the Volcano Alert Level to ADVISORY.
"Unrest at Pavlof Volcano remained slightly above background levels for about six months after early August 2016. Throughout this period, low-level steam and gas plumes from the summit and weakly elevated surface temperatures were observed occasionally when viewing conditions were favorable. The elevated surface temperatures were likely associated with still cooling pyroclastic deposits on the north flank of the volcano. By February 2017, unrest had declined to background levels although occasional, small, low-frequency events were observed in seismic data, consistent with an open, degassing system. AVO lowered the Aviation Color Code to GREEN and the Alert Level to NORMAL on February 2, 2017."

The Aviation Color Code for Pavlof Volcano was raised to YELLOW on Wednesday, June 7 following an increase in low-frequency seismic activity and a pilot report indicating a possible ash cloud to 4000 ft asl. Active degassing from the summit was observed in web camera images and by local observers in Cold Bay yesterday and this morning. Infrasound data from local instruments on Pavlof and a more distant network in Sand Point show no evidence of significant explosive activity during the past week. Seismic activity has been at background levels since Thursday, June 8. Precursory activity leading up to previous explosive eruptions at Pavlof have been subtle and while some episodes of increased seismic activity have preceded eruptive episodes other increases have died back down without explosive activity.

Weakly elevated surface temperatures and vapor plumes continued to be observed at Pavlof, but declined through June, July, and August. On August 30, 2017, citing normal, background levels of activity at Pavlof, AVO lowered the Aviation Color Code and Volcano Alert Level to GREEN/NORMAL.

From Cameron and others, 2023: "The number of located earthquakes at Pavlof Volcano increased in 2018 compared to previous years. Although this increase was at least in part due to improvements to the local seismic network, there are indications that seismic activity in 2018 was nonetheless above background levels. Shallow LP earthquakes are common at the volcano, and the relative amplitudes of their waveforms, as observed by the Pavlof Volcano seismic network, indicate they tend to be located at or near the volcano summit. Although these events are commonly too small to be located by AVO’s routine earthquake location procedures, they are observable on spectrograms and are still noted in AVO’s routine seismic checks (Power and others, 2020). Similar activity has been observed at other open-vent systems in the Aleutian Arc, such as Shishaldin Volcano (Petersen and others, 2006; Pesicek and others 2018) and Mount Veniaminof (Pesicek and others, 2018), and appears to be typical for such systems (Quezada-Reyes and others, 2013).
"The Pavlof Volcano seismic network [in 2018] consists of three broadband digital seismic stations (PN7A, PS1A, and PS4A), three short-period 3-component stations (PV6A, PVV, and HAG), and one short-period vertical-component station (BLHA). The current configuration is fairly new; AVO carried out substantial upgrades to the network in the summer of 2017, during which PN7A, PS1A, PVV, and PS4A were upgraded to broadband digital stations, PV6 was replaced by PV6A, and HAG was upgraded from a vertical-component analog station to a 3-component station (Dixon and others, 2019).
"AVO located 152 earthquakes within a 20-kilometer radius of Pavlof Volcano in 2018...Throughout 2018, the volcano remained at Aviation Color Code GREEN and Volcano Alert Level NORMAL. Of the events that occurred in 2018, 111 were designated as LP earthquakes and 41 as volcano-tectonic (VT) earthquakes...Moderately deep LP earthquakes (10-20 km [6-12 mi] deep) took place in a broad swath east of Pavlof Volcano and Pavlof Sister throughout 2018, and although these events appear to be well-located based on reported root mean square errors (0.05-0.33), they often had small magnitudes.
"On March 27, 2018, an unusual cluster of VT earthquakes took place about 6-8 km [3.7-5 mi] north-northwest of Pavlof Sister and 12-13 km [7.5-8 mi] north of Pavlof Volcano. The earthquakes were 6.6-9.0 km [4.1-5.6 mi] below sea level with local magnitudes (ML) between 0.3 and 1.5. Later, in early April, a sequence of deep LP earthquakes took place near Pavlof Sister. These events were 10.1-29.3 km [6.3-18.2 mi] below sea level with ML values between −0.37 and 1.99, but only two events from the sequence were above ML 1.0. On September 11, another series of deep LP earthquakes took place in the same location as the March LP earthquake activity. The September 11 sequence had nine events, the two largest being ML 1.3 and 1.4. The hypocentral depths of these nine events ranged from 26 to 32 km [16-20 mi]. The sequence was also accompanied by a short tremor episode lasting about four minutes, though the tremor may have actually been closely spaced LP earthquakes.
"Volcano-tectonic earthquakes are not as common at Pavlof Volcano as at other volcanic systems, making the March 2018 brittle failure sequence notable. These events are assumed to have been distal VT earthquakes on the basis of their location (12-13 km [7.5-8 mi] from Pavlof Volcano), though they also may have been tectonically generated. AVO has not located any similar cluster in the same area since it began monitoring the volcano in 1996. The fact that there were more located events in 2018 than in previous years is not surprising given the network improvements made in 2017. Many of the 2018 events were of small magnitudes, and the increase in earthquake locations is partly due to the improved network. However, the additional presence of the March VT earthquake cluster and the LP earthquakes above ML 1.0 indicate the 2018 activity may have been above background levels for the upgraded network, suggesting increased magmatic activity at Pavlof Volcano in 2018. Alternatively, this behavior may be typical for the volcano but is only observed now because of the improved network.
"Observations made by Power and others (2004) of deep LP earthquake activity throughout the Aleutian Arc suggest a link between magma movement in the lower to middle crust and eruptive activity, although the timing between these parameters appears to vary. The deep LP earthquake activity at Pavlof Volcano may indicate the presence of continued magma supply to the volcano, with 2018 a slightly more active year. Further monitoring of earthquake activity at Pavlof Volcano will likely shed light on the magmatic system that feeds eruptive activity at the volcano."

From Orr and others, 2023: "In 2019, Pavlof Volcano showed signs of weak activity that caused AVO to raise the Aviation Color Code and Volcano Alert Level to YELLOW and ADVISORY three times...The first change was made on May 15, 2019, in response to an increase in seismic tremor and webcam images of robust summit degassing. The Aviation Color Code and Volcano Alert Level were lowered back to GREEN and NORMAL on June 12, after activity declined...
"A total of 90 earthquakes were located within 20 km of Pavlof Volcano during 2019. Of these events, 44 were classified as volcano-tectonic (VT) earthquakes, with a collective depth range of −2.85 to 31.5 km and a local magnitude (ML) range of −0.41 to 1.65. The other 46 were classified as LP earthquakes, with a depth range of 8 to 34 km and a ML range of −0.36 to 1.65. Of the 46 LP events, 21 were located ~5 km northeast of Pavlof Volcano beneath Pavlof Sister, a pattern also seen in the seismicity of previous years (Power and others, 2004b). AVO recorded several tremor episodes at Pavlof Volcano in 2019. There is no obvious temporal relationship between the tremor and the earthquake activity...The activity in 2019 started with weak, intermittent tremor on May 15-19. Two brief tremor signals (~3 minutes each) were then observed on May 29, followed by several brief tremor bursts on May 31...
"The annual number of earthquakes located at Pavlof Volcano increased between 2017 and 2019. This trend is probably influenced by improvements made to Pavlof Volcano’s seismic network in the summer of 2017, but the fact that AVO located significantly fewer events in 2019 than 2018 suggests a change in activity level unrelated to the network improvements."

AVO raised the Aviation Color Code and Volcano Alert Level for Pavlof to YELLOW/ADVISORY on September 21, 2020 (AKDT), citing ongoing increased seismicity over the previous 24 hours. Seismic unrest declined in October, and on October 14, 2020, AVO lowered the Aviation Color Code and Volcano Alert Level at Pavlof to GREEN/NORMAL.

From Orr and others, 2024: "After a 9-month quiescent period that started in late 2020, seismic activity at Pavlof Volcano increased on July 9, 2021, in a distinct change from its background levels. Tremor periods observed on that date prompted AVO to increase the Aviation Color Code and Volcano Alert Level from GREEN and NORMAL to YELLOW and ADVISORY. From July 9 to August 5, seismic activity at the volcano was characterized by more volcanic tremor periods and occasional low frequency events, but no outward signs of eruptive activity were observed.
"On August 5, 2021, clear webcam views of Pavlof Volcano showed episodic, low-level ash emissions. These intermittent ash bursts came from a new vent on the upper southeast flank of the volcano, which produced diffuse ash clouds that rose just above the summit and drifted roughly 10-15 km [6-9 mi] southeastward before dissipating. Seismic and infrasound data associated this activity with occasional small explosions and tremor. The ash emissions clearly indicated that an active eruption was in progress, so AVO raised the Aviation Color Code and Volcano Alert Level to ORANGE and WATCH later that day.
"Seismic and infrasound sensors regularly recorded small explosions through the rest of the year, and when viewing conditions permitted, minor ash emissions appeared in webcam views and were observed by pilots. Diffuse ash clouds intermittently rose as high as ~12,000 ft (~3,700 m) ASL and were visible in satellite data as far as 10-15 km [6-9 mi] beyond the vent. Ash fallout during this period settled mainly on the upper south-southeast flank of the volcano, within 2-3 km [1.2-1.9 mi] of the vent. Occasional stronger ash bursts and farther-traveling ash clouds may have produced trace amounts (less than 1/32 inch or 0.8 mm) of ash fall as far as 10-15 km [6-9 mi] southeast of the vent, but this amount of ash is difficult to observe in satellite data and so could not be confirmed.
On August 25 and 26, mid-infrared satellite images showed slightly elevated surface temperatures at the summit of the volcano for the first time. The detection of elevated surface temperatures in satellite data usually indicates that lava is at or near the surface; however, lava flows at Pavlof Volcano were not confirmed until November 8, when the thermal output at the vent increased considerably. The presence of shallow subsurface magma and hot gases may have contributed to the slightly elevated thermal signals observed between late August and early November.
"Satellite images acquired on November 11 showed an active lava flow or spatter accumulation, about 200 m in length, and associated lahar deposits (extending ~2 km [~1.2 mi] beyond the vent) on the upper southeast flank of the volcano. Many historical eruptions of Pavlof Volcano have been characterized by lava fountaining or jetting that result in the accumulation of spatter around the vent. Occasionally, these growing spatter piles become unstable and collapse, forming hot particulate flows that are capable of eroding and melting glacier ice and snow on the volcano. The formation of meltwater by this process is a primary mechanism for lahar generation at Pavlof Volcano (Waythomas and others, 2017). The eruptive activity in early November may have been characterized by such periods of lava fountaining, spatter accumulation, and the extrusion of hot, granular flows that produced the lahar deposits observed in satellite imagery. A Sentinel-2 short-wave infrared (SWIR) image from November 12 showed a circular area of hot material around the active vent, consistent with the notion that spatter accumulation had been occurring. However, none of the webcam views or occasional pilot reports from November and December confirmed incandescence or lava fountaining at the south flank vent.
"On November 17, satellite observations indicated both the presence of ballistic clasts around the vent (some located as far as 2.5 km [1.6 mi] away) and continued lahar development that was likely associated with ongoing lava-ice-snow interaction. The ballistic clasts were ejected during energetic explosions that were recorded in seismic and infrasound data during the week of November 12-18. Minor explosions and small ash emissions took place occasionally during the last two weeks of November 2021, although the volcano was obscured by clouds for much of this period.
"From late November through December, the Pavlof Volcano seismic network detected elevated seismicity consisting of episodic, sustained tremor periods and discrete low-frequency events. Many explosions were detected in seismic and infrasound data—these may have produced localized ballistic ejecta fallout around the active vent, as indicated by satellite data acquired on December 1, 2021. On December 4, a webcam and a passing aircraft recorded minor ash emissions that rose from the summit of the volcano to an altitude of ~10,000 ft (~3,000 m) ASL. Satellite data from this period also commonly indicated moderately to strongly elevated surface temperatures. High-resolution satellite imagery collected during clear weather commonly showed lava effusion on the upper southeast flank of the volcano, as well as continued interaction with snow and ice that subsequently produced small lahars downslope from the lava flows. For instance, satellite images obtained on December 19, 2021, showed both an ~600-m-long [~2,000-ft-long] lava flow and minor lahar deposits that extended ~1,300 m [~4,300 ft] beyond the flow front."
Pavlof remained at Aviation Color Code ORANGE for much of 2022, with intermittent lava effusion, small explosions, and low-level ash emissions. On December 17, 2022, AVO lowered the Aviation Color Code and Volcano Alert Level at Pavlof from ORANGE/WATCH to YELLOW/ADVISORY, stating that no explosions had been detected since December 11, 2022, and no elevated surface temperatures or incandescent lava since December 2, 2022. On January 19, 2023, AVO lowered the Aviation Color Code and Volcano Alert Level to GREEN/NORMAL.