From Beget and others (2005): "There is evidence that a lake filled the caldera to a depth of 150 meters over a period of several hundred years. The caldera lake drained catastrophically after failure of part of the caldera rim. The resultant flood caused deep erosion of the landscape on the north side of the volcano (Wolfe, 2001). Lahar deposits are present at low elevations in many of the stream valleys draining the volcano, and can be traced up to 12 kilometers down valley from the caldera rim. At most sites only a single lahar deposit is exposed, but at least three lahars, each more than 1 meter thick, occur in sea cliffs near the mouth of Ginger Creek on the west flank of the volcano, and are exposed intermittently for more than 2 kilometers along the beach. Radiocarbon dates suggest these lahars were emplaced 300 to 400 years ago."

Hantke (1951) wrote that there was "an eruption on Umnak" in 1805. It is difficult to say if this eruption occurred and was from Okmok. Grewingk (1850, translated 2003 by Fritz Jaensch) who compiled accounts of many travelers to the region during the time period 1760-1850, makes no mention of this eruption. Langsdorff (1812), who was present in the area from 1803 through 1807, also writes "The active volcano on Umnak * * * has been totally still for a fairly long time" - apparently discrediting the later report from Hantke.

Grey (2003) has compiled and studied information about the 1817 Okmok eruption. From her thesis: "The first confirmed activity at Okmok was a large explosive eruption around 1817. There is a discrepancy as to the exact date and duration of the eruption. Veniaminov (1840, translated by Lydia T. Black and R.H. Geoghegan, 1984) maintains it was March 2, 1817; Postels (in Lutke, 1836) alleges it occurred on March 1, 1820. Grewingk (1850, translated 2003 by Fritz Jaensch) favors Veniaminov's assertion and 1817 is the date most often cited. However, Hantke (1951), who never visited the Aleutians, implies that Okmok was active during the entire four-year period and this is reflected in Simkin and Siebert (1994), though Hantke cites no specific references for his information. Whenever this eruption occurred (or perhaps there were two more separate eruptions?), there is general agreement about what happened during the explosive phase. During a storm with heavy SW winds, 'the range lying on the NE side of Umnak Island exploded * * * hurling great rocks for distances of up to 5 versts (~5 km)' (Veniaminov, 1840; Grewingk, 1850). The strong earthquake accompanying the eruption frightened the inhabitants of Unalaska, 120 km ENE of Okmok, who reportedly woke in the morning to find up to a foot of ash on the ground in some places (Lutke, 1836; Veniaminov, 1840; Grewingk, 1850). Iliuliuk Creek, which flows through Unalaska village, was reportedly so clogged with ash that it supported no fish for almost a year thereafter (Lutke, 1836; Grewingk, 1850). Though these reports all mention heavy ash fall in Unalaska, today no such thick ash layer can be found to substantiate this claim. It is quite possible that these eyewitness accounts overestimated the thickness of the deposit. It is also likely that any ash that was deposited in Unalaska has long since been eroded by the wind and rain action notorious in the Aleutians. Fieldworkers on Chuginadak Island in 2002 reported that the 2001 Cleveland ash fall on the island was approximately 20 cm thick up to 5 km from the vent, but still, no coherent ash layer can be found now, just some windblown pockets (J. Dehn, personal communication, 2003).
"The Aleut village of Egorkovskoe (or Adus), located on Cape Tanak (formerly called Egorkovskoi) on the northernmost end of Umnak, was destroyed by the 1817 eruption while its inhabitants were hunting in the Pribilof Islands (Grewingk, 1850). The villagers relocated to the Inanudak isthmus and in 1830 to the present site of Nikolski (Grewingk, 1850; Veniaminov, 1840). Because of the distance from the eruption source within the caldera, it is unlikely that the explosion itself was responsible for depositing the large boulders found near Cape Tanak. Wolfe and Beget (2002, and Wolfe, 2001) note a distinct absence of a thick enough tephra fall deposit at this site to bury a village. Rather, it is more likely that the boulders and gravel were carried in a syneruptive ash-laden outburst flood from Okmok caldera, either by disruption of an intracaldera lake or by melting of snow during the eruption. Radiocarbon date ranges for a 50-75 cm sand and gravel deposit found at Cape Tanak bracket the date of an outburst flood from the caldera between 1636-1951 AD from one soil sample, with a 55% probability between 1726-1813 AD, and between 1806-1931 AD for a second soil sample (Wolfe, 2001). These dates correspond well with the date of this eruption and support this hypothesis.
"It is possible that this explosive eruption originated from Cone E within the caldera because of the large, fresh pit crater in that cone that must have been formed by a forceful explosion quite recently. Alternatively, field evidence suggests that the eruption may have been through the intracaldera lake, originating at the site of Cone B, which is located close to the breach in the caldera wall (J. Beget and T. Neal, personal communication, 2003)."

From Grey (2003): "According to Grewingk (1850, translated 2003 by Fritz Jaensch), Lutke reported an eruption on the northeast end of Umnak Island, during which a 'mountain ridge opened up' in 1824 and was still 'smoking' in 1830. He also says that Postels (in Lutke, 1836) erroneously reported the activity as stretching along the northwest coast, but does not mention on what basis this presumption of error is made - perhaps the eruption was merely viewed from off the NW coast and appeared nearer than it was. I interpret these two reports as describing a linear (fissure) eruption and postulate that it may have taken place along the chain of small vents within the caldera that extends west from Cone B. This assumes fire fountains reaching above the caldera rim (500 m or more in height)."
Grey (2003) explains that fire fountains this high are unlikely, and then continues as follows: "Perhaps the observer saw the red glow of the eruption reflected on the underside of a low cloud deck. The relative freshness of the deposits from this arcuate fissure supports a historic time frame. There is also a young lava flow from Cone B that is so fresh it was originally mistaken for 1958 lava during field work in 2000 and could have been erupted in this time frame as well."
Powers (1958) also states that he believes a lava flow likely occurred during this eruption.

From Grey (2003): "A new crater is reported to have formed within the caldera in 1878 (Hantke, 1951), accompanied by earthquakes and a tsunami (location unknown). This could refer to the birth of Cone A, but since Hantke offers no additional information or references to support this notion, the report is apocryphal at best. It is intriguing to note that eruptions are also reported at Vsevidof, on the southwestern end of Umnak Island, in 1817, 1830, and 1878 (Miller and others, 1998). Of particular interest is the suggestion by Miller and others that the 1878 activity of Vsevidof 'may have been from a radial fissure eruption on the west flank; this event may have produced the youngest dacite flow, which extends west-southwest to Cape Kigushimkada.' Such a fissure eruption would probably produce earthquakes and could account for Hantke's 'new crater,' though not located inside a caldera. It is thus possible that Hantke misinterpreted the location of the 1878 activity, and there may not have been an eruption at Okmok in that year at all."
Sapper (1917), compiling information from Petroff (1884) and Fuchs, reports "a brief flame and ash eruption and a new crater developed from which came steam and hot cooking mud. This new crater appeared between Vsevidof and the settlement." This statement seems to support Miller and others's (1998) assertion that there was a radial fissure eruption at Vsevidof in 1878.

From Grey (2003): "An explosive eruption is said to have occurred in 1899 (Dunn, 1908). Robert Dunn (1908) details his visit to Umnak Island while 'vulcaneering' in the Aleutians. After landing on the beach south of Cape Aslik [see figure 4.2 in original text], he hiked up the slope past Jag Peak to the top of the ridge (the WSW rim), where he first laid eyes on the caldera: 'Below, yawned simply one titanic crater, five miles from far side to far side, if one single inch. Strewn on its floor, like toys perfectly carved, rose seven ash-cones; cones varied from symmetrical mounds that towered upon quite circular steep terraces, up to 500 feet and more, to the broken, chaotic black thing, like a big sand dump right under me. And that was the living soul of the discovery. Out of some vague cavern in its midst, undulated a column of white steam, a serpent-like Atlas, buoying the world's cloud cover.'
"Dunn also observed the ash spread to the west and south of the caldera: 'The devastation reached between four and five miles from the crater edge. This, of course, was the eruption of 1899.' Judging from Dunn's description of his location and his map, his black 'sand dump' is likely Cone A, thus implicating the then undeveloped cone as the source of this event [see figure 4.3 in original text]. Dunn is the first white man known to recognize the existence of the caldera and name it."

From Grey (2003): "Jaggar (1931) relates the journal entry of Mrs. Esther Wendhab, who tended seismographs in Dutch Harbor, on March 23, 1931: 'Captain Nelson of the Eunice reported a volcano smoking terribly with thick black fumes, where before he had only observed white steam, at Tulik on Umnak Island. He passed it March 21 and so unusual was its action he believes it is about to erupt. There was no record on the seismograph.'
"This eruption is reported to have continued at least until May 13 (Jaggar, 1932; Coats, 1950), when 'smoke' was still observed. Because activity is recorded as lasting nearly two months, this may have been a time of development of the pre-1943 Cone A lava field. Byers (1959) noted that the pre-1943 lavas to the northeast of Cone A were nearly identical in appearance to the 1945 lavas and would have been nearly indistinguishable had the 1945 eruption not been so well observed. Also, Freiday (1945) reported that parts of what he called the 'WAVE SPAR WAC' flow (the Cone A lava field, see figure 4.4 in original text) were melted out of the snow when he flew over the caldera [see also figure 5.2 in original text]. The date of his visit is not specified, but probably during the winte rof 1942-1943 and again in early summer 1943. Hantke (1951) calls for simultaneous eruptive activity at Okmok and Tulik in 1931, but the deeply glaciated state of Tulik precludes it having been active in historic time (Coats, 1950; Miller and others, 1998)."

A UP article recounts a probable eruption of Okmok, on December 20, 1935. "Seward, Alaska - Dec. 30 - (UP) -- Violent eruption of Mt. Tulik, active volcano on nearly uninhabited Umnak Island in the Aleutian Archipelago was reported today by Capt. Chris Trondsen of the steamer Starr, who reached Seward after a cruise along the islands.
"Capt. Trondsen said great streams of flaming lava were pouring out of the cone of Mt. Tulik Dec. 20, and the sky for miles was red with the glare. There are no villages near the mountain."

Coats (1950) reports smoke at Okmok in 1936.

The Anchorage Daily News reported on October 20, 1938, that U.S. Coast and Geodetic Survey crews of the ships Pioneer and Surveyor had reported "Mount Tulik on Umnak Island" active during the summer of 1938. The deeply glaciated nature of Mount Tulik (Coats, 1950; Miller and others, 1998) indicates that it has not been active within historic time. Eruptions attributed to Tulik are almost certainly eruptions at Okmok instead. Coats (1950) reports this eruption as occurring in October; this date probably references the date of the newspaper article and not the date of the eruption, which seems to have occurred sometime during the summer months. Powers (1958) reports a lava flow associated with this eruption, as well.

From Grey (2003): "Servicemen stationed at Ft. Glenn [U.S. Army base established on Umnak Island in 1942 and decommissioned in 1950) would occasionally trek into Okmok caldera looking for adventure as evidence by the occasional spent shells found on the caldera floor. One sunny afternoon in June 1943, two GI's drove up to the caldera rim near the base of Tulik cone and rappelled down to the crater floor, probably between the sites of the small intracaldera glacier and Cone F (Keller, 1976 and 1991). After spending the day hiking across the caldera, one of the men twisted his ankle on their way back and they were forced to spend the night inside the 'Devil's Brewpot.' Mr. Keller's account of their overnight experience follows:
"'A trembling of the earth woke me, and the whole crater was alight with a rosy glow. That was a horrible awakening! Right inside a volcano, and it was coming to life! There was a deep, rumbling sound, and a nearby cone was hissing. About that time another cone some distance away burst forth with a magnificent display of pyrotechnics * * * By the time we reached the cliff the shaking had abated, but rocks were rolling down the steep incline to the bottom * * * A fine black ash began drifting down upon us, and as suddenly as it had begun, the shaking stopped. The rosy light faded away and the crater resounded with a deafening silence.'
"Though this was a minor eruption, it was enough to terrify the two men, who based on Keller's description were approximately 4 km away from Cone A, the vent that had the brief fountaining (again, from Keller's description) event. The cone they heard hissing was probably Cone C, which is near the location where they descended into the crater and still has active fumaroles at its summit (field observations, 2001). Alternatively, it could have been sound from Cone A reverberating off of Cone C or the caldera wall."

From Snyder (1953): "On March 2 [1953] Richard McDonald, U.S. Geological Survey seismologist stationed at Adak, rode a Navy plane from Kodiak to Adak and back. He reported steam rising from Pavlof, Shishaldin, Makushin, and a small cone within Okmok Caldera on this day."

From Grey (2003): "An eruption that began in October, 1960 produced enough ash to reach Chernofski Sheep Ranch on western Unalaska Island [see figure 4.8 in original text] (Holmes, 1994). According to John Reeder of the Alaska Division of Geological & Geophysical Surveys, who interviewed Milton Holmes, resident Chernofski rancher, 'there were times when Milton could see lava shooting above the Okmok caldera only to fall back in. When the wind was not blowing, a red glow could be seen from the crater. At one time, he had to wear goggles while riding range because the ash was so thick' (Reeder, 1986). A concurrent report in the Anchorage Daily Times on October 17, 1960, states that Reeve Aleutian Airways passengers and crew 'were treated to an unexpected show of fireworks - the eruption of Tulik volcano' on October 15, involving a 'black cloud 9,500 feet into the air' and 'a 15-mile lava flow.' The flight attendant was quoted as seeing 'red lava and chunks of rock flying through the air at the base of the volcano.' This report is perplexing because in distinguishing between Tulik and 'Okmak Crater,' it limits the probability of misidentifying the active vent; yet as mentioned earlier [in other eruption descriptions], there is no geologic evidence to suggest an eruption from Tulik in historic times.
"Furthermore, there are no young lava flows associated with Tulik, and the maximum length of any lava flow exposed within the caldera is 7.8 km (the 1958 flow); there is no lava flow reaching 15 mi (25 km) in length anywhere on Okmok. Either the report was exaggerated or the number is a misprint and should have read '5-mile,' which would be reasonable. However, despite the vigorous level of activity implied by Holmes' observations and the Times report, no lava flow has been identified associated with this activity. It is probable that this was a brief burst of lava fountaining and the flow the Reeve passengers saw was the fresh 1958 flow.
"There is also one report of light ash fall in the community of Unalaska, 120 km ENE of Okmok in the late fall of 1960 (Reeder, 1986). Since neither Makushin nor Akutan volcano was erupting at the time, nor were those volcanoes immediately west of Okmok, that ash probably came from Okmok. Simkin and Siebert (1994) list the end date for this eruption as April 15, 1961. Based on measurements of tephra thickness within Okmok caldera by Reeder in 1980 and on western Unalaska Island by Holmes in 1961, the estimated total bulk volume of tephra from this eruption is 'greater than 1 x 10^7 cubic meters' (Reeder, 1986). It is possible that Reeder may have included tephra from the 1958 eruption in his calculations and overestimated the 1960 eruptive volume, since no lava flow is identified associated with this event."
From Black (1981): "Stanley Holmes, Unalaska, reports that when Tulik (Okmok) caldera erupted in 1960, lava ran across Crater Creek. The temperature of the water in the creek remained above normal for a year, disrupting the salmon run (personal communication, March 22, 1978)."

From Reeder (1987): "During the morning of 24 March 1981, pilot Harald Wilson observed a growing ash and steam eruption plume over Okmok Volcano. Harold was flying his Peninsula Airways, Inc. Navaho from Cold Bay to Dutch Harbor of Unalaska Island. His initial observation of the eruption plume was about 1000 local time (=GMT -10 h.) when he was about 80 km NNW of Dutch Harbor at a 2,600 m altitude. The plume of mixed steam and ash had reached an approximate 2,800 m altitude and was still growing. The visibility was good with a light wind from the ESE and with fairly cloudless sky. Before landing at Dutch Harbor 20 minutes later, Harold was able to watch the eruption plume grow to about a 5,500 m altitude, which was drifting to the NW.
"At 1100 on 24 March, Harold left Dutch Harbor for Nikolski of Umnak Island. During this flight, Harold flew over Okmok Caldera and circled the cinder cone that had just erupted (Cone A in the SW part of Okmok Caldera, see U.S. Geological Survey Bulletin 1028-L). The eruption had stopped and only steam was rising out of the cinder cone. Some ash was still in the air, but most of the plume had dissipated. A thin but very noticeable layer of black ash was deposited from the eruption on snow to the NW and WNW of the Cone A over the caldera floor as well as beyond.
"At 1500 of 24 March, Harold returned to Dutch Harbor from Nikolski. Only a small steam plume was being emitted from Cone A. Harold continued to fly to Nikolski about two to four times a week for Peninsula Airways, Inc. until late 1983. During these trips, Harold has only observed nearly continuous steam emission from Cone A, which is normal activity for this cone."

From Grey (2003): "A 100 km long plume was recognized in one NOAA 7 image at 1716 local time on July 8 [1983] (0216 UTM on July 9), but was not seen on images 12 hours earlier or later. This was apparently only a brief puff of ash and steam, accompanied by sulfurous gases, as reported by Aleutian Air pilot Tom Madsen (Reeder, 1986)."
From Reeder (1986), describing what Tom Madsen saw on his flight to Atka from Dutch Harbor on July 8: "Upon reaching Driftwood Bay, which is about 23 km WNW of Dutch Harbor, Tom could clearly see a hazy yellowish-brown smoke horizon straight west in the Okmok region. This smoke horizon was at an altitude of about 1,000 m and higher. The smoke was densest at Okmok, and the smoke horizon extended north across the Bering Sea as far as Tom could see (about 100 km). Due to thick fog and rain, Tom could not see the extent of the cloud to the south. Because Tom could not recognize the top altitude of this smoke cloud and because it appeared to extend over a very broad region, he elected to continue his normal route along the Bering Sea side of the Aleutian Island at an altitude of 500 m. Upon reaching Cape Idak of the most eastern part of Umnak Island, which is about 25 km NE of Okmok caldera, he was beneath the eastern edge of the cloud. Tom could detect sulfur gases. The cloud above him had a definite yellow tinge and at a distance was yellowish brown. He did not reach the other end of the cloud until he reached Inaudak Bay along the central Bering Sea side of Umnak Island, which is about 25 km SW of Okmok caldera and about 50 km SW of Cape Idak. No damage occurred to the Twin Beechcraft. Tom continued his normal route to Atka, arriving at about 1100. He then returned to Dutch Harbor by the same route, leaving Atka at about 1400 and arriving at Dutch Harbor at about 1600. During this return, he again had the same experience of flying under the Okmok eruption cloud, which started at about Inanudak Bay and ended at Cape Idak. During Tom's previous 3 years of flying in the Aleutian Islands, he has never had such an experience."
From Reeder (1986): "On 9 August, J.W. Reeder, while flying with Tom Madsen, made examinations of Okmok caldera. No new lava flows were recognized within Okmok caldera. Cinder Cone A, which is located in the SW region of Okmok caldera was teaming and some ash was noticeable over remaining winter-snow up to a 2 km radius from Cone A. A nearly transparent blueish smoke was also detected drifting to the east from the cone. J.W. Reeder and Tom Madsen have made several independent observations of the caldera during the previous 3 years. Cinder Cone A appeared to be similar in physical appearance, except it was steaming and smoking above average."

From Miller and others (1998): "A commercial pilot reported a steam and ash plume above a cinder cone (probably the 1945 cone) in the southwestern part of the caldera on November 18, 1986. Another small ash eruption occurred from the same cone on January 5, 1987 about 13 hours after a shallow, magnitude 6.6 (Ms) earthquake struck 130 km south of Okmok. A small pyroclastic flow was produced on the southeast flank of the cone and ash emission from the cone continued intermittently from January 1987 to February 1988 (Smithsonian Institution, 1986, 1987, 1988)."

From Grey (2003): "The latest eruption at Okmok began on or shortly before February 11, 1997 and originated at Cone A. An excellent synopsis of the eruption is found in U.S. Geological Survey Open-File Report 99-448 (McGimsey and Wallace, 1999) and is paraphrased here. Possible precursory activity included a steam plume observed by a passing pilot on November 11, 1996. The next report was of a dark plume, witnessed by a pilot and passengers en route from Atka to Dutch Harbor on February 11, 1997. Because these observations were not immediately reported, and because AVO did not yet have seismic instruments deployed on Umnak Island, AVO first became aware of unrest at Okmok with the detection of a thermal anomaly in Advanced Very High Resolution Radiometer (AVHRR) satellite imagery on February 13, 1997. The beginning of the eruption was confirmed with a phone call from ranchers at Ft. Glenn, 15 km E of Cone A, who reported a dark plume rising to 3000 m ASL and drifting to the SW.
"Inclement weather prevented further observation until February 18, when pilots reported an ash plume to 4600 m, and ranchers at Ft. Glenn reported a red glow reflected on the underside of clouds over the caldera. This was the first verification of lava at the surface. The thermal anomaly detected in AVHRR imagery continued to grow to 22 pixels by February 22. At this time, A Ft. Glenn rancher who had climbed to the caldera rim reported a lava flow extending 0.8 to 1.5 km northeast from Cone A. On February 28, a National Oceanic and Atmospheric Administration (NOAA) research crew passing by in a helicopter spent about twenty minutes flying inside the caldera, capturing six minutes of video footage and several still photographs of strombolian fountaining activity at Cone A. The photos document that the first (NE) lava flow lobe had reached its full length of 5.5 km from the base of Cone A by February 28 [ see figure 4.9 in original text], and the second (N) lobe had not yet begun to emanate from the cone.
"The satellite thermal imagery suggest that the second (N) lobe began to form sometime between March 1-4 (Moxey and others, 2001). Over the next three weeks, several pilot reports (PIREPs) were issued and satellite imagery continued to show high thermal activity and occasional ash plumes (AVO, unpublished data). On March 11, a PIREP reporting ash to 9000 m prompted the Federal Aviation Administration (FAA) to issue a NOTAM (Notice to Airmen; contents unknown). The largest thermal anomaly in AVHRR was on March 12 with 19 saturated pixels (Patrick, 2002). This probably coincides with the maximum areal extent of the lava flow, while low-level effusion, perhaps reactivated on the first (NE) lobe, likely continued until sometime between March 26-31 (Moxey and others, 2001; Patrick, 2002; Patrick and others, 2003). The presence of multiple flow units within the distal portion of the first (NE) lobe, as seen on Ikonos imagery and in the field suggest multiple pulses of effusion on this part of the flow. Three units were initially recognized, but field investigations suggest that the third (top) unit is perhaps an apparent additional unit, the result of lava flowing over a drop in the underlying 1958 flow. Intermittent ash bursts and low plumes continued for the next several months (McGimsey and Wallace, 1999).
"When the 1997 eruption of Okmok was over, the resulting lava flow covered a total area of 8.8 square km or about 10% of the caldera floor, based on the lava flow map produced in this study. Previous estimates range from 7.5 square km (Moxey and others, 2002) to 9.8 square km (Lu and others, 2003). With thickness varying from 5-45 m (Lu and others, 2002 and Lu and others, 2003), the calculated bulk volume is 1.5 x 10 ^8 cubic m (Lu and others, 2003)."

McGimsey and others (2004) report on an earthquake swarm detected May 11, 2001, at Okmok Volcano: "At 8:00 ADT (16:00 UTC) on the morning of May 11, 2001, AVO seismologists detected on the Makushin seismic network a swarm of earthquakes occurring under Okmok volcano prompting AVO to release an Update at 17:00 ADT (01:00 UTC). The events were ML=2.0-3.6 - too small to be felt by nearby residents - and determined to be possibly of volcanic origin. Satellite observations revealed no volcanic activity nor thermal anomalies. On-going satellite-based deformation measurements (SAR interferometry) show that the center of the caldera has inflated 20 cm between the 1997 eruption and September 2000. In April, 2001, AVO scientists observed an area of snowmelt in the caldera; although possibly indicative of heat flux, the area corresponded to the thickest part of the 1997 lava flow, which may still be hot (Patrick and others, 2003). By May 15, 2001, the earthquake swarm had greatly diminished. Okmok was last mentioned in the May 25, 2001 AVO Weekly Update."

From McGimsey and others (2014): After the 2008 eruption, "activity at Okmok remained at background levels until the morning of March 2, 2009, when a series of relatively strong, tremor-like events were recorded on the local seismic network. The largest tremor burst occurred at about 21:11-21:18 UTC. These events were recorded on average about once per hour during a 24-hour period, but were not detected on adjacent networks. A clear satellite view on February 28 revealed a 1-pixel thermal anomaly in the caldera near the location of Cone D, on the northeastern floor of the caldera. Clouds obscured satellite view of the area on March 2, and any surface manifestation accompanying the seismic activity went unobserved. In response to this abrupt, anomalous seismicity, AVO upgraded the Aviation Color Code/Volcano Alert Level to YELLOW/ADVISORY on the evening of March 2, 2009. A pilot flying low over the caldera on March 3 made observations and took several photographs. Several of the photographs appear to show an indistinct flowage deposit down the flank of New Cone [Ahmanilix], a 400-m (1,300 ft) high tephra cone (Schaefer and others, 2012) developed inside the caldera at the primary eruption site (see fig. 25 of Neal and others, 2011 for location). A clear satellite view on March 9 revealed an area of broad, but apparently shallow, slope failure on the west-northwest side of New Cone, which partly filled the pair of shallow pits located between the cone and the new lake west of Cone D. Material - likely pumice - also spread out into the lake but had been windblown to the eastern shore.
"Deformation (uplift) of the caldera floor began in August 2008 following cessation of the eruption. The GPS data from September 1, 2008 to March 1, 2009, show a linear progression of 9 cm of uplift in the center of the caldera (J. Freymueller and T. Fournier, UAFGI, written commun., May 3, 2009, AVO internal log entry 29436). Although seemingly rapid, the uplift was not considered unusual compared to the considerably more rapid inflation pulse that occurred in 2002-03 during a non-eruptive episode (J. Freymueller, UAFGI, written commun., May 3, 2009, AVO internal log 29436).
"Seismicity remained at low, to near background levels for the next several weeks, and satellite views showed no activity, thus prompting AVO to lower the Aviation Color Code/Volcano Alert Level to GREEN/NORMAL on March 20, 2009. No significant changes or activity occurred for the next 2 months. Then, on May 24-25, 2009, a series of tremor bursts were recorded, including a 15-minute-long high amplitude signal that registered across the entire Okmok network. The activity ceased almost as quickly as it began, and the network returned to near background seismic levels.
"Activity at Okmok remained at background levels through the remainder of 2009. The only point of interest was that satellite imagery in June compared with earlier imagery revealed extensive erosion of the 2008 eruption features within the caldera (J. Larsen, UAFGI, written commun., 2009, AVO internal log entry). Many of the collapse pits had become filled with erosional debris, the flanks of New Cone became heavily rilled, and the crater floor enlarged as sediment began filling and raising the bottom. The two lakes adjacent to New Cone also increased significantly in size."
From Neal and others (2014): "The rapid inflation [during 2009] slowed somewhat by the middle of 2010. In 2011, in inflation at Okmok continued, increasing over recent (2010) rates (5-6 cm over a 12-month period ending in September 2011), but still at a lower rate than was recorded following the 2008 eruption (M. Kaufman, UAFGI, written commun., September 7, 2011). In summary, the inflation at Okmok continues - albeit in pulses - with rates in the 5-7 cm/yr range (J. Freymueller, UAFGI, written commun., September 28, 2013)."

From Dixon and others (2015): "No eruptive activity was reported at Okmok Volcano, but seismic and geodetic observations of note were made in 2013. Sporadic tremor episodes and three swarms of earthquakes caught the attention of duty personnel in 2013. Geodetic instruments within the Okmok caldera detected a mid-year pulse of inflation. The Aviation Color Code and Volcano Alert Level remained at GREEN/NORMAL throughout the year.
"On March 7, a 36-hour long swarm of over 1,000 low-frequency earthquakes was recorded on seismograph station OKTU, a station on Mount Tulik just outside the caldera (fig. 37 in original text). These earthquakes were too small to be recorded on adjacent stations and could not be located. The earthquakes formed two groups of earthquakes with similar waveforms (or earthquake families) that began at the same time, with the first earthquake family continuing for the duration of the swarm and the second family lasting for about the first 6 hours (fig. 38 in original text). Family 2 (short-duration family) contained larger events than family 1. After the family 2 (short-duration family) events ceased, the events in family 1 became larger and more infrequent than events earlier in the same earthquake family. Towards the end of the swarm, the event interval became more erratic, and the swarm ended abruptly.
"Beginning in May 2013, the geodetic network at Okmok detected a pulse of rapid inflation, one of the steepest rises at Okmok since the 2008 eruption (fig. 39 in original text). The spatial pattern of the deformation is similar to past inflation events at Okmok, and points to an inflation source beneath the center of the caldera. In a study of ambient noise correlations between the Okmok stations OKNC and OKCE, evidence was found of 0.2-percent decrease in seismic velocity during late August and September within the caldera, indicating a change in composition of the crust sampled by the ray paths (Matt Haney, USGS/AVO, written commun., 2013). The geodetic and seismic evidence suggests an infusion of fluid or gas and, although it is certain that this was a change, it is not clear whether this change was magmatic or hydrologic.
"A swarm of earthquakes began at 01:55 UTC on September 28 (17:55 AKDT on September 27) southwest of Okmok and northeast of Mount Recheshnoi in an active geothermal area. A second swarm occurred at Geyser Bight on October 9, forming a continuous zone of seismicity that extends from Recheshnoi towards Okmok (fig. 40 in original text). Neither swarm has led to eruptive activity and has continued to occur into 2014."
From Cameron and others, 2017: "Okmok volcano continued to inflate during 2014-a
general trend that started no more than 3 weeks after the volcano’s 2008 eruption (Freymueller and Kaufman, 2010). More specifically, analyses of GPS and InSAR data from 2008 to present (2016) show evidence for two major pulses of post-eruptive inflation (Qu and others, 2015). Inflation of the volcano began at a rapid rate immediately after eruption and slowed with time, largely ceasing by mid-2013. A new pulse of rapid inflation began at that time, its rate slowly decreasing over time. Inflation continued through 2014, although at a much slower rate."

From Cameron and others (2020): "AVO seismologists noted seismic tremor at Okmok Caldera (herein called Okmok volcano to include associated volcanic features exterior to the caldera) in September 2016. This activity did not lead to significant unrest, and the Aviation Color Code and Volcano Alert Level remained GREEN and NORMAL, respectively. Tremor episodes began on September 19 and were frequently noted during the following week. Tremor episodes occurred at a rate of 5-10 per day and were detected on seismic stations within the caldera but not on those located outside the caldera. The locatable tremor episodes occurred in the east side of the caldera, and relative lag times suggested a source near Cone D. Although depths were not well constrained, they often seemed to be shallower than 5 km; however, occasional and more broadly occurring events seen across the network often had depths of greater than 25 km in the same region of the caldera.
"Inflation coincident with the tremor was not initially noted with the GPS network, but re-evaluation found a westward motion on GPS sites OKCE and OKNC. This motion is not consistent with the typical inflation signal at Okmok volcano (Jeff Freymueller, University of Alaska Fairbanks Geophysical Institute, written commun., 2017). Seismic and infrasound alarms were crafted for the volcano to assist with monitoring, but no further activity ensued. Pulses of tremor were sporadically noted through October, and by early November the instances of tremor were no longer observed."

From Cameron and others, 2023: "In 2018, Mount Okmok (a name which herein includes associated volcanic features outside Okmok Caldera, such as Jag Peak and Tulik Volcano) continued the long-term reinflation that began immediately after its last eruption in 2008. This deformation takes place in discrete pulses that appear modulated onto a lower-rate, steady background deformation pattern. In 2018, monitoring stations OKCE and OKNC recorded a complete pulse, with a total horizontal displacement of about 10 centimeters (cm) [4 in] and a vertical displacement of as much as 12 cm [4.7 in]. Past analyses of geodetic data (GPS and InSAR) indicate a shallow magma reservoir exists underneath the caldera floor (for example, see Freymueller and Kaufman, 2010; Lu and Dzurisin, 2014), and the continued volcanic inflation is consistent with an ongoing accumulation of shallow melt. Mount Okmok’s Aviation Color Code and Volcano Alert Level remained at GREEN and NORMAL throughout 2018."
From Orr and others, 2023: "In 2019, Mount Okmok continued the long-term reinflation that began after its last eruption in 2008. This deformation takes place in discrete pulses that appear modulated onto a lower-rate, steady background deformation (for example, Xue and others, 2020). Like in 2018, a complete pulse took place in 2019 and was visible in the time series of monitoring station OKCE. The total displacement values were much less for the 2019 pulse, however, producing amplitudes of 4-5 centimeters (cm) [1.6-2 inches] in the horizontal and ~5 cm [2 in] in the vertical components. These are roughly half the amplitudes of the horizontal and vertical components recorded in the 2018 pulse. Past analyses of geodetic data (GNSS and InSAR) suggested a magma reservoir lies 2-3 km [1.2-1.9 mi] below sea level beneath the caldera floor (for example, Freymueller and Kaufman, 2010; Lu and Dzurisin, 2014), but more recent work suggests the existence of a shallow sill at 0.9 km [0.6 mi] and a pressure point source at 3.2 km [2 mi] below sea level (Xue and others, 2020). Regardless, continued inflation of the volcano is consistent with ongoing accumulation of melt at shallow depths.
"AVO scientists identified several seismic tremor episodes at Mount Okmok in September 2019. These began with a short tremor burst recorded on September 4 following observations of a few deep, low-frequency earthquakes on August 21 and 24. Several more tremor bursts were recorded on September 5. A series of longer (2-3 minute) tremor bursts took place on September 6, with bursts recurring every 10 minutes for about 90 minutes. AVO observed similar seismic activity again on September 9. The tremor episodes were not formally locatable, but the difference in tremor amplitudes between seismic stations suggests the tremor took place near cone A. Note that cone A and the other cones in Okmok Caldera have no formal names; the names used herein are informal.
"In addition to monitoring this activity through daily seismic checks, AVO implemented internal seismic and infrasound alarms to detect any increases in unrest. Satellite data showed no signs of unrest at Mount Okmok during these events, and the local infrasound array recorded no acoustic emissions. A retrospective analysis, however, found that the tremor onset coincided with a stop in the long-term inflation signal typically seen in geodetic data. Intermittent tremor bursts continued after the early September activity before finally subsiding in late November. The tremor bursts of 2019 did not lead to greater unrest, so the Aviation Color Code and Volcano Alert Level of Mount Okmok remained at GREEN and NORMAL throughout the year."
From Orr and others, 2024: "In 2021, Mount Okmok continued the long-term deformation trend that began immediately after its 2008 eruption. This deformation takes place as discrete inflationary pulses superimposed onto a lower-rate, steady background inflation and is consistent with ongoing accumulation of melt at shallow levels. One such pulse was recorded in 2021, appearing on the time-series plots for GNSS stations OKCE, OKNC, and OKSO. The total displacements of the 2021 pulse were similar to the totals of 2019 but larger than those of 2020 (see, for instance, the OKCE time series in figure 31 [in the original reference]), with amplitudes of 4-5 cm [1.6-2 in] in the horizontal and ~5 cm [~2 in] in the vertical components. This inflation is roughly half that of 2018, when the volcano underwent ~10 cm [~4 in] of horizontal displacement and as many as 12 cm [4.7 in] of vertical displacement (Cameron and others, 2023).
"From roughly the beginning of October through November 2021, Mount Okmok departed from its common deformation pattern by producing an additional inflationary pulse, prompting AVO to release an Information Statement. This pulse, which was observed at stations OKCE and OKNC, was consistent with a pressure increase at less than 1 km [0.6 mi] depth that had a source located near Cone D, south of Ahmanilix. Past analyses of GNSS and InSAR geodetic data suggest a shallow magma reservoir exists underneath the caldera floor of Mount Okmok (for example, Freymueller and Kaufman, 2010; Lu and Dzurisin, 2014)."