FIRE and MUD Contents

Meteorological Observations of the 1991 Mount Pinatubo Eruption

By J. Scott Oswalt,1 William Nichols,2 John F. O'Hara1

1U.S. Navy.

2U.S. Air Force.


Unusually complete meteorological observations were made of eruptions of Mount Pinatubo by two military weather offices that were located within 40 kilometers of the volcano. Surface-weather and radar observations, weather satellite and NASA space shuttle imagery, and atmospheric pressure and temperature measurements recorded preliminary eruptions of June 12-14, 1991, major eruptions of June 15-16, and subsequent lesser eruptions through early September 1991. Observations also helped to record and define the macroscale and microscale atmospheric processes around Mount Pinatubo.

Weather radar was able to track both eruption-column rise rates and horizontal drift of ash clouds from some preparoxysmal and postparoxysmal eruptions. During the second eruption of June 12, 1991, radar indicated an apparent column rise rate in excess of 400 meters per second. Radar observations suggested that higher eruption columns correlated with greater particle size and density within the column.

High-resolution weather satellite imagery showed the existence of gravity waves within spreading ash clouds. Multichannel NOAA imagery distinctly differentiated between fine, drifting ash and high-level clouds. Radar height measurements were typically 10 to 15 percent lower than ash cloud heights inferred from satellite temperature analyses.

From 1530 on June 14 through the paroxysmal stage of June 15, microbarographs proved to be reliable indicators of explosive eruptions. Atmospheric compression waves created by the explosions caused obvious impulses on the barographic recorders. Compression and rarefaction extremes of 8 to 12 millibars were evident on the barograms during the climactic eruptions.

Typhoon Yunya passed within 75 kilometers of Mount Pinatubo at approximately 1100 on June 15. Copious rainfall from the typhoon increased the weight of tephra accumulations and the destruction they caused. Modification of winds aloft by Typhoon Yunya also spread ash over a much wider range of azimuths than would have otherwise received ash.

Postparoxysmal eruptions and residual heat contributed to moisture condensation and atmospheric instability in the vicinity of Mount Pinatubo. This process increased local rainfall, which in turn increased the occurrence of lahars and secondary phreatic explosions.

Note to readers: Figures open in separate windows. To return to the text, close the figure's window or bring the text window to the front.


Due to the coincidental colocation of two U.S. military weather observing sites within 40 km of Mount Pinatubo, the logistics support capability of their host military bases, and preeruption volcano monitoring, the June 1991 Mount Pinatubo eruption may have been the most closely observed major explosive eruption in history.

The purpose of this paper is to detail the capabilities and observations of the two weather sites, which together provided considerable observational data not normally available by other means. We detail the primary observational tools and methodology used. These include weather-observing technicians, weather radars, infrared sensor, satellite imagery, rawinsondes and pilot balloons, microbarographs, and aircraft pilot reports.

We then relate our combined observations through three eruption stages: the preparoxysmal stage of June 12-15, during which time the observational value of the weather sites became readily apparent, the paroxysmal stage of June 15-16, and the postparoxysmal stage of June 16 to November 15, 1991, during which time significant changes to the microscale and mesoscale climate near Mount Pinatubo were observed. We also discuss a significant tephra-fall forecast error and other remarkable ramifications of the eruption interaction with Typhoon Yunya.

Finally, we consider the characteristics of major eruption ash clouds, the utility of weather radar for volcanic ash detection, the viability of atmospheric pressure sensors as explosive eruption detectors, the use of satellite imagery in posteruption monitoring, secondary phreatic explosions, and minor problems created by windblown, suspended ash.

As military meteorologists and oceanographers fortunate enough to experience the 1991 Mount Pinatubo eruption, the authors seek to provide the geophysics community with our observations of this major event. Because we lack resources to thoroughly research certain topics we address, we recognize possible omissions and ask readers to contact us with questions or suggestions.


Certified military surface weather observers.--Technicians were stationed at Clark Air Base (Clark AB), 15 km east-northeast of Mount Pinatubo, and at Cubi Point Naval Air Station (Cubi Point NAS, part of the U.S. Naval Facilities complex at Subic Bay), 39 km south-southwest of Mount Pinatubo (fig. 1). Both observational sites were manned 24 hours per day, primarily in support of aircraft operations. Hourly and special criteria observations were recorded routinely and disseminated electronically. Recorded weather parameters included time of observation, cloud amount(s) and type(s), horizontal visibility, obstructions to vision, wind direction, average sustained wind speed, wind gust speed, ambient air temperature and dewpoint, atmospheric pressure, and significant weather.

Figure 1. Geography of west-central Luzon Island, showing the three observational sites: Clark Air Base, Pampanga Agricultural College, and Cubi Point Naval Air Station.

Due to the close, unprotected location of Clark AB at the foot of Mount Pinatubo, evacuation of personnel and instruments was required during some periods of volcanic activity, notably at the beginning of the paroxysmal stage of June 15 and lasting until June 19. A log was kept of visual observations from an evacuation site 38 km east of the volcano.

Weather radars.--The AN/FPS-77 at Clark AB and the AN/FPS-106 at Cubi Point NAS were 1950's technology, C-band weather radars with very similar performance specifications (table 1). Both had Plan Position Indicator (PPI) display for horizontal return and Range Height Indicator (RHI) display for vertical return, which provided three-dimensional coverage within the performance range limits. Regrettably, neither the Clark nor the Cubi Point radar could produce hard copy of screen displays; our illustrations of radar signals were traced from the radar screen. The Clark radar had superior coverage due to its unobstructed view of Mount Pinatubo's summit, closer location to the mountain, and greater height range 24,000 m (80,000 ft). Both radar antennae were housed in protective domes, which allowed them to operate (rotation and angle inclination) free from ash intrusion into their drive mechanisms.

Table 1. Military weather radar specifications.


Clark Air Base

Cubi Point Naval Air Station

Frequency (megahertz)



Wavelength (centimeters)



Power Output (kilowatt peak)



Pulse Repetition Rate (pps)



Pulse Length (microsecond)



Maximum Range (nautical miles)



Maximum Height (feet)






Infrared sensor.--The Infrared Detecting Set (AN/AAQ-9 in military nomenclature), while not normally used for weather observations, was utilized by Clark personnel to assist in volcano monitoring efforts. Its spectral response of 7.6 to 11.75 m provided thermal imaging useful for observing volcanic activity at night.

Satellite imagery.--At Clark AB, a Defense Meteorological Satellite Program (DMSP) polar orbiting satellite receiver provided near-realtime high-resolution visual and infrared (IR) imagery. At Clark AB and Cubi Point NAS, geostationary satellite receivers provided near-realtime, low-resolution visual and IR imagery. At the Joint Typhoon Warning Center in Guam, a geostationary satellite receiver, a National Oceanographic and Atmospheric Administration (NOAA) TIROS-N polar orbiting satellite receiver, and a DMSP satellite receiver provided high-resolution, near-realtime visual and IR imagery.

National Aeronautics and Space Administration.--A space shuttle mission underway during the eruption provided realtime visual observation and postevent videography from low Earth orbit.

Rawinsonde and pilot balloons.--The Rawinsonde Set (AN/GMD-5 in military nomenclature) was located at Clark AB, and VAISALA Mini-Rawinsonde (AN/UMQ-12 in military nomenclature) was located at Cubi Point NAS. These systems provided telemetered upper level temperature, humidity, pressure, and wind speed and direction data. Pilot balloons were used to measure vertical visibility and winds aloft by visual tracking.

Microbarograph.--Microbarographs were located at both Clark AB and Cubi Point NAS and provided instantaneous measurement and continuous recording of surface atmospheric pressures. The device is designed for shipboard as well as land use and includes a viscous damping mechanism to prevent motion contamination of the recording.

Pilot reports.--Postevent commercial airline reports detailed ash plume encounters and, in some cases, ash damage. Local military personnel were aware of the ash hazard and associated avoidance areas; therefore, no airborne ash encounters were reported by military aircraft.



At approximately 0851 local time (G.m.t. plus 8 h) on June 12, the Clark weather radar RHI display indicated a strong, coherent echo with well-defined edges located directly over the Mount Pinatubo vent. Visual confirmation of an eruption resulted in the immediate evacuation of the remaining U.S. Air Force security and command personnel, including the weather office, from Clark. In the course of a routine hourly weather observation, the Cubi Point weather observer saw the initial eruption as the tephra column rose above the foreground ridge line, which obscures the line of sight from Cubi Point to the volcano summit. The column rose to a level exceeding the height range of the Cubi Point radar (18 km) and was measured at greater than 19 km by the Clark radar just prior to weather office evacuation. Lightning was observed within the tephra column as well as from the column outward to clear air.

When the column reached a neutrally buoyant level (Sparks and others, 1994), it spread horizontally and radially, as shown by the Cubi Point weather radar PPI display depicted in figure 2 and by visual satellite imagery shown in figure 3. It created a broad umbrella or canopy of ash and resembled a "mushroom" cloud or severe thunderstorm. This ash umbrella spread south-southwest past Subic Bay and east over Clark Air Base (figs. 2 and 3), but only a minute trace of ash fell at Cubi Point, where it was monitored on sheets of black and white paper placed outdoors. Significant tephra fall was reported by townships located directly downwind (west-southwestward) of the low to middle tropospheric winds. The former San Miguel Naval Communications Station (fig. 1) reported a sky darkened by tephra fall, and accumulations of up to 5 cm. The mixture of ash and small, coarse pumice fragments was easily removed from most surfaces.

Figure 2. Cubi Point weather radar Plan Position Indicator (PPI) display of the 0851 June 12 eruption ash cloud. Because the radar is used primarily for aviation service, distances are measured in nautical miles (NM). The Plan Position Indicator was set at 60-nm range gate. Edge contours for the ash cloud are in local time.

Figure 3. Geostationary weather satellite visual image of the 0851 June 12 eruption. The spreading ash cloud (A) covers the entire Subic Bay (B). Picture time is 0931.

Fine ash suspended at higher levels was eventually carried westward over the South China Sea, as reported by commercial aircraft and detected by analysis of satellite imagery. Larger and heavier pumiceous tephra concentrations precipitated relatively quickly, while lighter ash particles remained suspended significantly longer and were transported greater distances before settling out of the atmosphere. Observed winds aloft, radar observations of the ash cloud, and local pilot reports enabled military flight weather forecasters to provide accurate recommendations for safe emergency aircraft sortie (evacuation) operations.

Several additional explosive eruptions between June 12 and the paroxysmal stage of June 15 were observed either visually, by radar, by barogram pressure impulse, or by some combination of these three methods. Typically, eruptions reached the tropopause (~17 km) in 3 to 5 minutes. However, faster rates were observed, such as the eruption detected by both radars at 2252 on June 12 (fig. 4), in which the column rose to 9 km, temporarily slowed its ascent, then rose to above 24 km in approximately 30+-5 s, indicating upward velocity well in excess of 400 m/s. (Note: During this event, we made the assumption that the radar was, in fact, sensing a rising ash column. Other than the anomalous ascension rate, the radar signature was essentially identical to other eruptive events. Though we are confident of the observation, it is difficult to conceive how such a rise rate might be possible. We believe that, ultimately, it is most important to report this observation, despite uncertainty about its explanation.)

Figure 4. Cubi Point weather radar Plan Position Indicator (PPI) display of the 2252 June 12 eruption ash cloud. In this illustration the PPI was set to a 30 nm range gate. Ash cloud edge contours are in local time.

A series of eruptions the afternoon and early evening of June 14 yielded the first major ash fall at Subic Bay Naval Station. As a result of a laterally directed eruption detected at 1530 (Hoblitt, Wolfe, and others, this volume), 2.5 to 5.0 cm of very fine ash was deposited over the base. With local rain showers spawned by approaching tropical cyclone "Yunya" (locally "Diding"), much of the ash fall was wet, had a consistency resembling moist clay, and adhered to exposed surfaces. Having been briefed by U.S. Geological Survey personnel on the effects of volcanic ash fall on aircraft, military authorities directed the closure of the Cubi Point airfield. Closure led to the grounding of a contract commercial cargo plane (DC-10), which subsequently sustained significant damage from prolonged tephra fall on June 15. This ash-fall event also introduced the problem of ash-induced radar attenuation, as the suspended ash caused the leading edges of the ash cloud radar echo to gradually become diffuse and difficult to track. A temporary, simultaneous outage of both the weather radars early that evening, coupled with our uncertainty about the winds aloft (attempts to launch thin latex weather balloons into an ash-filled sky proved futile) resulted in inability to gauge the extent, dimensions, density, and movement of the ash cloud.

Barogram indicators of eruptions on June 12, 13, and early June 14 had been either very subtle or impossible to discern. When the explosions became laterally directed, commencing with the blast at 1530 on June 14, the barograph pressure recorders indicated obvious impulsive pressure fluctuations (compression and rarefaction) in response to atmospheric compression waves radiating outward from the explosions. Thereafter, the barograph was a reliable indicator of explosive events and was used to determine actual times of explosive eruptions.

Weather radar proved useful as verification of eruptive events. Clark AB and Cubi Point NAS noted generally excellent agreement in quantitative and qualitative radar observations. Signal return displays of vertical eruption columns and laterally spreading ash clouds were generally sharp and distinct, being slightly more coherent than a thunderstorm but less than a building or mountain. Both sites also observed that higher eruption columns produced more intense radar returns. This may be attributable to greater particle size and concentration. Use of the iso-echo (contouring of the return-signal strength) display allowed detection of renewed eruptive pulses within the echo display of ash from previous eruptive pulses, and prevented the radar return of new eruptions from being masked by the return of previous events. The largest eruptive events created eruption columns far exceeding the height detection capabilities of the weather radars.

The high-resolution DMSP satellite imagery received at Clark was able to detect moderate to large ash clouds, and enhanced multichannel imagery received at other global sites, was successfully used to differentiate between ash and water or ice (fig. 5). Low-resolution imagery was unable to detect all but the very largest ash cloud of June 15. Ash-cloud height estimates derived from temperature analyses of ash-cloud tops showed only fair correlation with radar observations (comparisons were only possible with smaller post-paroxysmal eruptions that were within radar height range). Satellite height analyses were typically 10 to 15 percent lower than radar height measurements.

Figure 5. NOAA-11 weather satellite multichannel (channels 1-3) image of the ash cloud (A) from the 0851 June 12 eruption drifting over the South China Sea. Mount Pinatubo is at point B. Picture time is 1457 June 12.


An explosive eruption observed by the radars at approximately 1038 on June 15 signaled the onset of vigorous sustained eruptive activity that included the climactic explosions. Reduced light level and suspended ash were observed at Cubi Point shortly after sunrise, and by 1200 the sky was totally obscured and darkened by tephra fall and suspended ash. The evacuation site for the Clark weather office at the Pampanga Agricultural College near Mount Arayat (fig. 1) experienced low light levels as well, as tephra fall began there at approximately 1440. The insolation blockage at Cubi Point was quite dramatic and compounded by a lack of any reflected ground-based light source. The effect was analogous to visibility in a totally darkened room. Without a light source, visibility was reduced to near zero. Insolation blockage also caused a significant decrease in normal ambient air temperature, as measured at Cubi Point (fig. 6). While no measuring equipment was available at the Agricultural College, weather observers there noted that the ambient air temperature felt exceptionally cool that afternoon and evening.

Figure 6. Comparison of observed versus climatological (long-term average) daytime ambient air temperatures at Cubi Point Naval Air Station for June 15. The Fahrenheit temperature scale is used to increase resolution.

Enveloped in rain, suspended ash, and falling pumice, radar could no longer track the outer edges of the ash cloud. Signal attenuation by tephra fall and suspended ash prohibited all but the strongest returns. Therefore, we were limited to radar detection of activity occurring over the vent, where particle size and concentration were greatest.

Tephra fall continued at Cubi Point throughout the day, varying from completely dry ash through a wet, cement-like mud, to muddy water. Particle size gradually increased, with coarse pumice fragments reaching a maximum of approximately 5 cm in diameter at about 1530. Postevent examination of deposits at Clark yielded pumice fall fragments similar size; locally, fragments 15 to 30 cm in diameter were left by volcanic mudflows (lahars).

Unhangared aircraft at Cubi Point NAS suffered damage due to the weight of accumulated wet tephra. For example, both the commercial DC-10 mentioned previously and a military C-130 transport airplane fell back on the ends of their fuselages as the accumulated ash radically altered their centers of gravity. Civil aviation pilots reported ash-cloud encounters that resulted in engine damage, exterior abrasion, and interior contamination by ash (Casadevall and others, this volume). Tephra fall also created considerable problems for ships in port at Subic Bay. The accumulation of saturated ash and pumice on decks created alarming weight increase and caused potential stability problems by shifting centers of gravity. The sulfur content of the debris corroded exposed sensitive electronic equipment. Exposed rotating machinery, such as radars and wind speed indicators, experienced mechanical failure or was damaged from bending or jamming of the drive mechanisms or rotational surfaces. The ability of ships to safely navigate out of the ash cloud was significantly impacted by low visibility and inoperative navigation radars. Tephra was ingested by some engines (four marine diesel engines required replacement), and pumice material floating on the water limited ships' ability to take in engine cooling water and to distill fresh water. Wet ash accumulations caused power transmission lines, insulators, and transformers to short circuit. The power loss rendered the Cubi Point water treatment plant and pumping stations inoperative. Subsequent electrical power restoration was a lengthy and tedious process of thoroughly cleaning and reenergizing individual segments of the basewide grid.

Throughout this tephra-fall episode, frequent lightning was observed across the celestial dome at both Cubi Point and the Agricultural College. The lightning often appeared red, green, or blue and was not a product of cumulonimbus clouds (thunderstorms) but of the frictional effect of ash fall. Satellite imagery and radar observations suggest that thunderstorm activity did not occur. There were no convective cloud formations evident within the ash cloud, and the dynamic processes required for thunderstorm occurrence were disrupted by the enormous volume of ash suspended throughout the tropospheric layer and by tephra-induced subsidence. In the absence of this eruption, the passage of a tropical cyclone taking a track such as Typhoon Yunya (fig. 7) would have very likely brought significant thunderstorm activity to central Luzon.

Figure 7. Best track for Typhoon Yunya, as provided by the Joint Typhoon Warning Center, Guam. Positions are Greenwich Mean Time with the day shown first. Local time is G.m.t. +8 h. Other numbered and lettered symbols are weather stations with various measuring capabilities.

Barograms give a fairly accurate indication of eruptive activity throughout the paroxysmal stage (fig. 8). However, the low temporal resolution of the recordings hampers a detailed analysis of the pressure fluctuation maxima occurring between approximately 1400 and 2200 (0600 and 1400 G.m.t.) on June 15. The maxima shows short-duration pressure extremes of approximately 8 to 12 mbar and corresponded to the onset and duration of considerable sensible seismicity at both observing stations.

Figure 8. Surface atmospheric pressure barograms from Clark AB and Cubi Point NAS for the period June 13 to June 16. Barogram time scales are Greenwich Mean Time (G.m.t.). The onset of the paroxysmal stage is indicated at point (A).

Constant, low-frequency rumble, distinctly different from the nearly continuous thunder, emanated from the general direction of the volcano during the period of frequent magnitude 3 to 5.6 earthquakes and vigorous atmospheric pressure fluctuations.

High-resolution imagery also resolved gravity waves propagating outward through the laterally spreading ash umbrella (figs. 9 and 10). Another feature clearly evident on some high-resolution imagery is the eruption column rising above, or overshooting, the ash umbrella (fig. 9). It appears that the rising eruption column is ascending significantly higher than the level of neutral buoyancy, where the ash umbrella is spreading. This is indicated in figure 9 by the shadow the eruption column is casting on top of the ash umbrella as the column rises above it (Tahira and others, this volume; Lynch and Stephens, this volume)

Figure 9. Geostationary weather satellite visual image of the 0555 June 15 eruption. Picture time is 0630 June 15. Notable features include well-defined concentric gravity waves (A) and the eruption column (B) "overshooting" the broad ash canopy. The exposed low-level cyclonic circulation of Typhoon Yunya (C) is visible to the east of the ash cloud.

Figure 10. Geostationary weather satellite visual image of the climactic eruption series of June 15. Picture time is 1630 June 15. Gravity waves (A) are discernible in the massive ash cloud, which is casting a well-defined shadow (B) to the east.


In anticipation of significant tephra fall accompanying a Mount Pinatubo eruption, volcanologists, meteorologists, and various key decisionmakers were interested in defining a tephra-fall hazard area. On the basis of climatology of the monsoon wind regime of central Luzon, tephra-fall hazard areas were generalized as west-southwest (in response to the late northeast monsoon season) and northeast (in response to the southwest monsoon season) of the tephra source. Numerical forecasts based on observed winds aloft were in general agreement with climatology studies and hazard area delineations. Excellent verification of forecast tephra response to northeast monsoonal flow was observed during pre-paroxysmal eruptions on 12 and 13 June, and this appeared to lend support to defined tephra-fall hazard areas.

On June 13, a tropical disturbance in the southern Philippine Sea was upgraded to Tropical Storm Yunya (fig. 7). Yunya tracked to the northwest, skirting the eastern shores of the Visayan Islands and southern Luzon while intensifying to typhoon strength (fig. 11). By 1400 on June 14, Typhoon Yunya was located over the north coast of Catanduanes Island, close enough to central Luzon to begin inducing a departure from the climatologically normal monsoon wind regime. The typhoon's proximity and cyclonic (counterclockwise) wind flow induced a southward bias in the normal middle to upper tropospheric winds, which was evidenced by the first significant ash fall over Subic Bay during the June 14 eruptions. Typhoon Yunya made landfall over central Luzon near the Dingalan Bay (120 km east-northeast of Clark) at 0800 on June 15 and decreased to tropical storm intensity in response to land interaction and shearing. The low-level circulation of Tropical Storm Yunya then proceeded along the favored track of storms entering Luzon over the Dingalan Bay: northwest over the Pampanga plain region (during which time its center passed within 75 km northeast of the erupting volcano). Yunya exited Luzon at the Lingayen Gulf at approximately 1800 on June 15, proceeding north-northwest along the northern Luzon coast (fig. 9). Meanwhile, the upper-level circulation of the system sheared from the lower level and moved westward, and the associated high-level cloudiness intermingled with the ash cloud as both moved out over the South China Sea.

Figure 11. Geostationary weather satellite visual image of Typhoon Yunya (A) approaching Luzon. Picture time is 1431 on June 13.

Because of Typhoon Yunya's passage, the entire paroxysmal eruption occurred during a radical departure from the usual monsoon wind regime. This anomaly, coupled with uncertainty about ejecta volume (experts had previously warned of this uncertainty), rendered wind climatology-based tephra-fall forecast maps virtually irrelevant. Typhoon Yunya's modification of the seasonal winds caused a southward displacement of light to moderate tephra accumulations, transported by the middle tropospheric to low stratospheric winds, and a northeastward displacement of moderate to heavy accumulations transported by low-tropospheric winds. Due to the highly improbable simultaneous volcanic eruptions and influence of Typhoon Yunya on the wind patterns, plus the enormous amount of volcanic ejecta, Clark, and particularly Subic Bay, received substantially more tephra fall than was generally expected. A comparison between a climatology-based tephra-fall hazard map and an analysis of actual tephra deposits from the June 12-15 eruptions illustrates the effect of Typhoon Yunya in combination with the huge volume of material ejected from Mount Pinatubo (fig. 12).

Figure 12. Ash-fall-hazard planning map versus isopachs of ash fall from the June 12-16 eruptions. The unit of ash depth is centimeters. Due to uncertainty about winds on any given day, other ash fall planning maps declined to attempt specific definitions of hazard areas (Punongbayan and others, this volume). Isopachs for individual eruptions are presented in Paladio-Melosantos and others (this volume).

The tephra accumulation at Subic Bay was all the more significant in view of previous contingency plans and actions by military authorities. With Clark highly susceptible to tephra fall, and possibly even to pyroclastic flows, an emergency evacuation plan had been developed identifying U.S. Naval Facilities Subic Bay as the prime evacuation site for Clark personnel. Subic Bay had been deemed safe from all but the most improbable catastrophic mountain collapse and was south of the expected tephra-fall hazard area associated with the northeast monsoon. Due to increasing eruptive and seismic activity, Clark had evacuated all military dependents and nonessential military and civilian personnel (14,500) to Subic Bay on June 10. The huge increase in the base population exacerbated the problems created by the tephra fall of June 14-15.

The considerable rainfall produced by Yunya saturated most of the tephra fall, and the cement-like consistency of the accumulated tephra made removal very difficult and tedious. Local analysis showed that the moistening of loose, dry tephra increased its weight by 23 percent, and the added weight created by the rainfall certainly contributed to widespread structural failures at Clark, Subic Bay, and communities throughout central Luzon.

Were the eruption and the passage of tropical cyclone Yunya related? A remarkable but probably coincidental correlation exists between the time of Typhoon Yunya's closest approach to Mount Pinatubo (fig. 7) and the pressure fluctuation maxima. Although the Yunya-induced mean atmospheric pressure fall is unremarkable in terms of tropical cyclone passage (approximately 6.3 mbar), there is, nonetheless, an association between the time of pressure minimum (about 1100 on June 15, on the Clark Air Base barograph), the closest approach of Yunya to Mount Pinatubo (within 75 km at approximately 1400, June 15, as tracked by weather satellites), and strong barometric pressure fluctuations associated with eruptions (strongest between 1038 and 2200 on June 15). Both the eruption and the tropical cyclone were well along in their development before they coincided and presumably could have occurred independently of each other. We know of no evidence for any causal relation between the two, but, clearly, the passage of Yunya greatly increased damage from the eruption.


Significant changes to microscale and mesoscale weather patterns near Mount Pinatubo were observed through mid-September. Prominent among these was the formation of "volcanic" thunderstorms. During most of the less vigorous postparoxysmal eruptions, cumulus cloud complexes formed near the top of the buoyant ash plume, where minute ash particles and other material provided abundant hygroscopic nuclei. These cloud complexes frequently developed into cumulonimbus clouds (thunderstorms). The thunderstorms often drifted away from their source region at the top of the plume, producing sometimes significant amounts of localized rainfall, "mudfall," and ash fall along their drift tracks before dissipating. As an example, a deluge at Clark on September 1, which yielded 74 mm of rain in 2 hours, probably was a Pinatubo-induced thunderstorm. New thunderstorms often developed near the top of the ash plume as previous cells drifted away.

As the southwest monsoon (rainy season) took effect in mid-July, visual and radar confirmation of this phenomenon was restricted because of extensive cloudiness and precipitation. Differentiation between buoyant ash plumes and convective cloud formations was then achieved by comparing Clark radar displays with concurrent seismometer data.

Even in the absence of an observable buoyant plume, significantly greater than average occurrences of afternoon thunderstorms and rain showers were observed near the vent and surrounding hot pyroclastic-flow deposits, which together provided the initial buoyancy needed to lift moist, unstable, tropical air to a level where it continued to rise due solely to its own instability (level of free convection). Significant flooding and lahar events were attributable to rain showers and thunderstorms caused by residual heat from Mount Pinatubo.

Rainwater interacting with hot pyroclastic-flow deposits has resulted in steam explosions and jets, ejecting tephra to 15 km and greater, and creating significant localized tephra fall. On some occasions, explosions occurred days after rainfall ended.



Tephra fall distribution from the June 12-15 eruptions was controlled by prevailing winds, which were strongly modified by the approach and passage of Typhoon Yunya. On June 12, Yunya was not yet a factor; on June 14 and 15, it was a major factor. It would be interesting to examine these events with a three-dimensional model of the wind field and attempt to reproduce the observed patterns of tephra fall (fig. 12). For more detailed information about tephra fall, see Paladio-Melosantos and others (this volume) and Wiesner and Wang (this volume).


Weather radars (even 1950's vintage) proved to be extremely useful in detecting initial and secondary explosions and in tracking drifting airborne ash clouds or plumes, whenever the concentration of airborne ash (threshold unknown) was sufficient. Radar information concerning eruptions, tephra-bearing secondary phreatic explosions, and ash-cloud dimensions and trajectories, was critical for ash warnings. Considering the limitations of the radars used, it seems likely that newer Doppler weather radars would be far more capable in detecting eruptions and tracking airborne ash. A Doppler weather radar should be able to differentiate ash particles from water; this capability would make it useful in areas where clouds and precipitation curtail other means of monitoring. Research could be conducted to determine ash-cloud dispersion versus particle size and fall rate. Doppler also presents obvious potential for study of eruption column dynamics. Newer technology may allow remote radar monitoring and provide recording capability for postevent analysis. We think it should be feasible to configure a mobile van with a relatively small and inexpensive commercial Doppler weather radar, which could then be deployed to various areas of interest.


Harris and Rose (1983) and Rose and Kostinski (1994) have noted that quantification of volume, particle size, ash-cloud height, and other pertinent parameters either can be obtained by radar observation or can be reasonably assumed, with the resultant data integrated with the winds aloft for the development of a tephra-fall forecast. Research to improve characterization of ash clouds by radar observations is ongoing (Rose and Kostinski, 1994). We recommend that working algorithms or nomograms that incorporate results from this research be developed for use by weather radar operators.


The ability to readily detect explosive eruptions by barograph was shown. Eruptions from June 14 and 15 produced stronger barograph signals than those of June 12-13, perhaps related to stronger laterally directed components (pyroclastic surges and pyroclastic flows) in the June 14-15 eruptions. A relatively inexpensive monitoring system could be constructed by placing atmospheric pressure sensors near potentially explosive volcanoes. While such a system would not provide eruption warning, it would give instantaneous notification of explosive (especially, laterally directed) eruptions and provide a first guess on the eruption strength as based on the magnitude of the pressure impulse. A digital system could provide convenient long-term data storage but would require a sampling rate high enough to resolve discrete explosions.


Multichannel satellite imagery can distinguish between ash and water droplets or ice crystals (clouds) (fig. 5). Satellite imagery is most useful in posteruption monitoring of ash cloud or plume dimensions and trajectories, in order to warn aircraft that are transiting affected areas. The key to making these warnings effective lies in disseminating information to the global aviation community quickly and updating this information whenever new satellite imagery or other data become available.


Major eruption columns approximated the physical structure associated with "super" thunderstorms. The magnitudes of several eruptions were great enough to allow penetration of the stratosphere, with the "anvil" or "mushroom" top (ash umbrella) forming around a level of neutral buoyancy. Fluctuation of the ash about this level was detected in high-resolution satellite imagery as concentric gravity waves in the ash umbrella (Lynch and Stephens, this volume; Tahira and others, this volume). Analysis of these waves may be useful in determining relative eruption intensities and magnitude changes.


By creating vertical instability and triggering rain shower and thunderstorm formation, the volcano generated its own microscale and mesoscale weather patterns. After the main eruption, and before the onset of extensive cloudiness from the southwest monsoon, it was quite common to see thunderstorms and rain showers develop exclusively over Mount Pinatubo. With the return to the dry northeast monsoon in mid-October, we continued to observe enhanced cumuliform cloudiness over the vent region. We believe the additional heat emitted from both the vent and the deep, proximal pyroclastic-flow deposits created the vertical instability that caused the cumuliform development.

Enhanced rainfall increased the occurrence of secondary phreatic explosions. These explosions are caused by rain water or lahars coming into contact with still-hot pyroclastic-flow deposits. The explosions have lifted tephra to over 15 km, and subsequent tephra fall has locally reduced visibility and ambient light level. These secondary explosions pose a continuing hazard to aviation.


During the dry northeast monsoon, fine ash was lifted and entrained by low-level winds from local ridge lines and large expanses where vegetation had not stabilized tephra deposits. Sufficient wind speed (>8 m/s) was the primary factor in lifting the ash. Although the concentration of airborne ash was enough to significantly lower near-surface visibility (below visual flight rules), no problems with aircraft engine performance were reported. As vegetation recovered and loose ash was washed away by rainfall or was otherwise removed, the problem of airborne ash diminished relatively rapidly. No recurrences of significant wind-blown ash were observed during the late-1992 northeast monsoon.


We are grateful for the encouragement and assistance provided by C. Newhall, W. Duffield, R. Hoblitt, T. Casadevall, and others from the U.S. Geological Survey and by R. Punongbayan and the Philippine Institute of Volcanology and Seismology. This study has also benefited substantially from work and discussions with K. Rodolfo, J. Umbal, M. Paladio, and others from the 1991 Pinatubo Lahar Hazards Task Force. Illustrations provided by L. Carr, U.S. Navy, S. Runco, National Aeronautics and Space Administration, and R. Hudson, U.S. Air Force, were an invaluable addition to this work. Finally, critical assistance by W. Johnson, U.S. Navy, is greatly appreciated.


Casadevall, T.J., Delos Reyes, P.J., and Schneider, D.J., this volume, The 1991 Pinatubo eruptions and their effects on aircraft operations.

Harris, D.M., and Rose, W.I., 1983, Estimating particle sizes, concentrations, and total mass of ash in volcanic clouds using weather radar: Journal of Geophysical Research, v. 88, no. C15, p. 10969-10983

Hoblitt, R.P., Wolfe, E.W., Scott, W.E., Couchman, M.R., Pallister, J.S., and Javier, D., this volume, The preclimactic eruptions of Mount Pinatubo, June 1991.

Lynch, J.S., and Stephens, G., this volume, Mount Pinatubo: A satellite perspective of the June 1991 eruptions.

Paladio-Melosantos, M.L., Solidum, R.U., Scott, W.E., Quiambao, R.B., Umbal, J.V., Rodolfo, K.S., Tubianosa, B.S., Delos Reyes, P.J., and Ruelo, H.R., this volume, Tephra falls of the 1991 eruptions of Mount Pinatubo.

Punongbayan, R.S., Newhall, C.G., Bautista, M.L.P., Garcia, D., Harlow, D.H., Hoblitt, R.P., Sabit, J.P., and Solidum, R.U., this volume, Eruption hazard assessments and warnings.

Rose, W.I., and Kostinski, A.B., 1994, Radar remote sensing of volcanic clouds, in Casadevall, T.J., ed., Volcanic ash and aviation safety: Proceedings from the First International Symposium on Volcanic Ash and Aviation Safety: U.S. Geological Survey Bulletin 2047, p. 391-396.

Sparks, R.S.J., Bursik, M.I., Carey, S.N., Woods, A.W., and Gilbert, J.S., 1994, The controls of eruption column dynamics on the injection and mass loading of ash into the atmosphere, in Casadevall, T.J., ed., Volcanic ash and aviation safety: Proceedings from the First International Symposium on Volcanic Ash and Aviation Safety: U.S. Geological Survey Bulletin 2047, p. 81-86.

Tahira, M., Nomura, M., Sawada, Y., and Kamo, K., this volume, Infrasonic and acoustic-gravity waves generated by the Mount Pinatubo eruption of June 15, 1991.

Wiesner, M.G., and Wang, Y., this volume, Dispersal of the 1991 Pinatubo tephra in the South China Sea.

FIRE and MUD Contents

PHIVOLCS | University of Washington Press | U.S.Geological Survey

This page is <>
Contact: Chris Newhall
Last updated 06.11.99