Mt Pinatubo 1991 Eruption

Ash fall impacts

Figure 22.

Heavy ash falls of >10 cm occurred over 100 km from the volcano in the period June 12-15. Ash fall was directed to the west and south-west (Fig. 22), and fell over the South China Sea, reaching as far as Vietnam, Cambodia and Singapore. A total area of around 840 by 400 km was affected by ash fall. Heavy ash fall resulted in complete darkness in some areas, buried crops and agricultural land, clogged water supply and sewerage systems, and caused many building and house roof collapses. The largest of the several tephra eruptions occurred on June 15 and comprised 3.4-4.4 km3 of tephra, covering around 2000 km2 with 10-25 cm of rain-saturated tephra.

Due to the heavy, wet tephra fall (it fell during Typhoon Diding), 189 people were killed by collapsing buildings.

Plate 9. Residential buildings that withstood 20 cm of pumiceous tephra at the Clark Air Base, from the eruptions of Mt. Pinatubo, the Philippines in June, 1991. - V. E. Neall.

Plate 10. Collapsed vegetable market building at the Clark Air Base, where the weight of wet tephra caused roof collapse and bent vertical steel I-beams. Note jumbled circular ventilators that were previously installed in the roof. - V. E. Neall.

Figure 23.

The tephra eruptions had world-wide effects in addition to those on Luzon. The eruption columns on June 12-15 penetrated the tropopause, blasting millions of tonnes of gases and fine particles into the stratosphere. Within the stratosphere there is very little water vapour and very high winds dispersed the cloud of material rapidly around the world (Fig. 23).

As the cloud travelled around the globe it spread, widening to the north and south, completely encircling the globe within 22 days. The clouds resulted in vivid sunsets in many parts of the world and a daytime haze above others. Satellite observations were hindered by the cloud and cooler and abnormal weather patterns were partly attributed to its presence blocking out sunlight. The SO₂ gases within the cloud also caused some corrosion of the acrylic windows on aircraft.

Pyroclastic flows

Figure 24.

The first pyroclastic flows travelled only 4-5 km from source but later, one on June 12-16 travelled up to 16 km from source at around 80 km/hr. These incandescent flows affected the Marella, Maraunot, O'Donnell and Sacobia river valleys (Fig. 24). Around 5.5 km3 of pumiceous pyroclastic flow deposits were emplaced by the largest eruption on June 15. These affected an area of around 400 km2, drastically altering the landscape by infilling former river valleys with huge thicknesses of deposits (up to 100 m).

Up to 2 years after the eruption, collapse of the still hot (>300 ^oC), thick pyroclastic flow deposits caused secondary pyroclastic flows that travelled up to 10 km. In addition, during heavy monsoon and typhoon rains in the rainy seasons following the eruption, over 200 hot lahars (c. 50^oC) were generated by erosion of the 1991 pyroclastic flow deposits, and more in the two years following. Rainfall runoff interacting with the hot pyroclastic flow deposits caused many secondary phreatic explosions with associated minor ash falls.

Lahars

Figure 25.

Lahars produced during the Pinatubo eruption and in the following years caused the greatest damage and longest-term effects on the surrounding population and landscape. During the eruption, typhoon rains and eruption-related storm cells generated lahars on both the eastern and western volcano flanks. However, the most widespread lahars were generated by monsoon and typhoon rains, after the emplacement of the unconsolidated pyroclastic flow and tephra deposits. Eight major catchments were affected by lahars (Fig. 25). With continuous lahars occurring, channels were filled to their capacity with sediment, causing later lahars to overflow and inundate lowlands surrounding the river channels.

Since 1991, lahars have occurred in the wet seasons of 1992-1994 and the total area affected has progressively expanded, although the frequency of lahars has declined. At the end of the 1994 rainy season, around 2.2 km3 of the 1991 pyroclastic flow deposits and tephra had been eroded and deposited on alluvial fans surrounding the volcano. The 1991-1994 lahars were mostly hot (around 50oC) and steaming. The lahars ranged from dilute, erosive hyperconcentrated streamflows to highly concentrated, viscous debris flows, containing dominantly pumiceous sand. Lahars in 1995 were predominantly cold and contained greater proportions of pre-1991 deposits within them.
Lahar and some secondary pyroclastic flows caused the blockage of several catchment tributaries to form temporary lakes, the largest of these being on the SW flank of Pinatubo. Many floods and cold hyperconcentrated streamflows were caused by breaching of these lakes, three such events along the Passig-Potrero River caused the loss of several lives. Prediction and early warning of these lake-breakout floods and lahars proved difficult because they were not dependant on heavy rainfall and were in areas difficult to monitor due to residual heat and hydrothermal explosions (in the area of pyroclastic flow deposits).

Lahar monitoring and warning system

Figure 26

Most lahars generated on Pinatubo are a result of heavy rainfall on surfaces of unconsolidated pyroclastic flow deposits. The main feature of these surfaces is that they have a very low permeability to rainfall - causing large amounts of runoff. Hence, to identify conditions likely to initiate lahars, telemeted rain gauges were installed in several catchments around the volcano (Fig. 26). Lahars in 1991 and 1992 were initiated following rainfalls of 0.3-0.4 mm per minute, sustained for over 30 minutes. When rain gauges now detect this intensity of rainfall within a particular catchment, lahar watches are set up. Lahar watchers are assigned to designated sites (Fig. 26) and comprise army, police or civil defence personnel. In addition to this network, trip wires and acoustic flow monitors are installed at various sites (Fig. 26). Acoustic flow monitors detect ground shaking near lahar channels caused by the passage of debris flows and hyperconcentrated streamflows.

A series of lahar warning signals are used by civil defence authorities to convey information to the public on the likelihood of lahars, based on data from the detection network (Table 10). At the initial stages of setting up the lahar warning network, some lahars were missed by malfunctioning acoustic flow monitors; however since then every passing lahar has been detected by the equipment.

Although the lahar warning system proved to be a success overall, some problems were experienced. Several early false alarms led people to doubt all alerts, delaying some evacuations. In other areas, lahar dikes and sediment control structures that were built led to a false sense of security and also delayed evacuations.

Other lahar hazard mitigation strategies

Lahars from Pinatubo were expected for several years because of the huge thickness and volume of pyroclastic flow deposits on the flanks of the volcano. To help authorities to manage these, short term (for one-two years) lahar hazard maps were prepared. Initially these used topography and channel morphology to delineate hazard zones, but the maps were continually updated as lahar sedimentation changed channel forms and dynamics, causing the potential for lahars to inundate different areas.

Table 10. Lahar warning signals used by PHIVOLCS and the civil defence authorities around Pinatubo, adapted from PHIVOLCS (1991).

Figure 27

For long term planning a yearly sediment budget was prepared to determine the expected duration and magnitude of lahars in the years following the eruption. This was prepared using an exponential decay model and was continually refined with the addition of information available in successive years (Table 11, 12 and Fig. 27).

In cases of prolonged intense rainfalls associated with typhoons or siyam-siyam (continuous monsoonal rains for more than 9 days), expected sediment yield could be significantly altered. In addition, events such as secondary pyroclastic flows, stream piracy, and lake breakouts greatly affected sediment delivery to the lowlands.

Armoured (by boulders etc.) and unarmoured sand dikes were constructed to protect some communities directly threatened by lahars. However, dikes and sediment retention structures in the rivers draining Pinatubo mostly proved ineffective in containing the huge volumes of eroded sediment. River channels quickly aggraded with sediment from lahars and floods, overtopping almost all of the dikes built. If dikes were effective in one area, greater sediment loads travelled farther downstream, inundating other areas.

It has been estimated that to trap the Pinatubo sediment in retention basins 10 m deep, 120-360 km2 of land would be required. Or alternatively if smaller basins were constantly being excavated, a huge number of trucks and diggers would be required to keep pace with the sediment inputs. Either of these options was very expensive.

Channelisation of lahars, by dikes and levees, although experimented with, is not an effective solution at Pinatubo. The gradient of channels on the lowlands is too low to provide enough energy to move sediment all the way out to sea. No matter how efficient the channels were constructed, they remained filled with sediment.

Table 11. Estimated volumes of sediment to be transported to lowlands surrounding Mt. Pinatubo, adapted from Pierson et al., (1992).

Table 12. Annual sediment delivery from Mt. Pinatubo, estimated by exponential decay models with actual data from 1991-1994, adapted from PHIVOLCS (1995).