Alaska Volcano Observatory

From Neal and others (2009): "The first known explosive eruption [at Cleveland Volcano] in 2006 occurred on February 6, and was detected in routine Advanced Very High Resolution Radiometer (AVHRR) satellite image analysis by anomaliesAVO staff. Satellite-derived temperatures indicated an initial cloud height of about 6-8 km (22,000-26,000 ft). The eruption appeared to be short lived and had ceased several hours before detection on satellite imagery, but AVO raised the Level of Concern Color Code to RED and NWS issued a SIGMET. The cloud was tracked for more than 400 km (250 mi) southeast of the volcano before it largely dissipated in satellite images. By day's end, with no further reports or images of ash production at Cleveland, AVO downgraded the Color Code to ORANGE. For the duration of condition RED, the FAA imposed a temporary flight restriction (TFR) from the surface to 50,000 ft within a 5 nautical mile radius of the volcano.
"AVO downgraded Cleveland to Color Code YELLOW 5 days later on February 11. Cloud cover persisted during most of this interval, and AVO's information release on February 11 noted that undetected, low-level unrest could continue. No further indication of activity led AVO to further downgrade Cleveland to a Color Code of 'Not Assigned' on February 20. (AVO policy is that a volcano lacking seismic instrumentation cannot be known to be at background, and hence cannot be assigned a Color Code GREEN)."
"On May 5, AVO reported a thermal anomaly and continuous plume of volcanic gas from Cleveland. The plume was visible over the course of 6 hours but traveled only 48 km (30 mi) southwest of the volcano and appeared to be at a relatively low altitude and devoid of ash. Cleveland remained 'Not Assigned.'
"On May 23, an astronaut aboard the International Space Station (ISS) contacted ground control with a report of an eruption from a volcano in the Aleutians. Subsequent communication with AVO, including receipt of a photograph [see fig. 41 in original text, and http://www.avo.alaska.edu/image.php?id=10064 online] and a telephone call from the ISS, confirmed Cleveland as the source. The plume was ash rich and by the time it was detected in AVHRR imagery, it was a detached ash cloud about 130 km (80 mi) southwest of the volcano. Satellite cloud-temperature data indicated a cloud top of about 6,700 m (22,000 ft) ASL. AVO raised the Level of Concern Color Code for Cleveland to YELLOW, and subsequently downgraded to 'Not Assigned' on May 26 after no further activity was detected.
"On August 24, AVO received notice from NWS that a ship had reported an ash eruption from Cleveland volcano. Days later, AVO received video footage from the crew of this fishing vessel showing a definite ash plume reaching about 3 km (~10,000 ft) ASL [see fig. 42 in original text]. Importantly, neither a broadband regional network seismic station in Nikolski [see fig. 1 in original text; M. West, UAFGI, written commun., 2006) nor any time-correlative satellite imagery showed evidence of this eruption. On September 7, after reviewing video footage of the August 24 event and noting an intermittently present thermal anomaly at the volcano, AVO raised the Level of Concern Color Code to YELLOW.
"AVO was alerted by NWS of another Cleveland eruption on October 28 after a pilot report to the Anchorage Air Traffic Control Center. The pilot of a jetliner indicated an initial cloud over the volcano reaching their flight level of 36,000 ft (11,000 m) ASL, and a drifting cloud moving east-northeast at a lower level of 30,000 ft (9,100 m) ASL. Satellite-derived cloud top temperature estimates placed the plume much lower. Utilizing the new warning scheme adopted by United States Volcano Observatories in October, AVO declared Aviation Color Code ORANGE and Volcanic Activity Alert Level WATCH for Cleveland about 2 hours after receipt of the pilot report, and reverted to YELLOW/ADVISORY on October 30 after no indications of further activity. On clear days under optimal satellite viewing conditions, a weak thermal anomaly was detected in the vicinity of the summit crater at Cleveland into November [see fig. 43 in original text].
"Ash explosions likely occurred more often at Cleveland than were detected in either satellite imagery or by pilots or other observers. An infrasonic signal received at the Geophyiscal Institute in Fairbanks on November 6 may have been produced by an explosion at Cleveland (or a nearby volcano); however, with no corroborating evidence, AVO took no action (S.R. McNutt, UAFGI, written commun., 2006)."