A broad river of lava is flowing down the upper western flank of Etna on the evening of 25 October 1999, and lava fountains rise from the source vent in the Bocca Nuova. The crater, which was about 150 m deep in mid-1995 was filled to its rim in October 1999, leading to a series of spectacular lava overflows

The Bocca Nuova from 1995 to 2001

During the first two years of the current eruptive cycle, activity within the Bocca Nuova was moderate, resulting in a fairly slow filling of the crater by intracrater cones and lava flows, and by late July 1997 the crater floor was about 100 m below the northern crater rim. Starting in July 1997, more vigorous activity caused the rapid growth of a large complex cone at the northwestern vents (A) which rose to a maximum height of 50 m below the northern crater rim within the next three months, to become one of the largest cones that have formed within that crater during its 30-year-long life. Activity at this cone was more or less continuous and consisted of lava splashing within one or more vents (at times up to 5 vents were erupting simultaneously), punctuated by larger bursts that produced short-lived lava fountains rising high above the crater rim. During periods of particularly intense activity one single large crater occupied the top of the cone, this was filled to the rim by a pond of seething lava that was in continuous agitation and frequently burst forth with huge lava fountains. As soon as the activity quieted somewhat, most of the lava surface in the crater crusted over while lava splashing and fountaining in various spots maintained open holes in the crust. Bombs and scoriae ejected at these sites were dropped onto the crust and rapidly accumulated to form small mounds that were destroyed when renewed vigorous activity ripped and consumed the crust on top of the magma column.

Cone growth at the southeastern vents (B) was initially much slower, mostly because here the activity was discontinuous, in spite of the violence of the explosions. Two hours of Hawaiian-style lava fountaining on the evening of 25 November 1997 caused the rapid accumulation of ejecta around the vents, building a large half-cone that leaned against the SE crater wall, its highest points being at about the same elevation as the northern crater rim. Vigorous activity from the A vents and a third vent (C) in the NE part of the crater floor continued through late December 1997, contributing to a further voluminous infilling of the crater.

In early January 1998, the magma column below Bocca Nuova began to subside. Incipient collapse at the B vents destroyed the upper half of the cone, and arcuate inward-stepping faults around the base of the cone indicated that slower subsidence occurred over a wider area encompassing the crater floor around the cone. Collapse at the A vents began somewhat later and initially consisted of the enlargement of the vents followed by the slow subsidence of the whole half-cone that had built around the vents. Although the main phase of subsidence occurred in mid-January 1998, the entire crater floor continued to sink slowly through at least March. Subsidence was most pronounced in the southern part of the cone at the B vents where a funnel-shaped depression was formed that engulfed the southern third of the cone. During late January to early March 1998 explosive activity occurred from newly formed vents in this area, possibly reflecting the degassing of the sinking magma column rather than its renewed rise. Little activity occurred in Bocca Nuova until late May 1998 when a new phase of crater filling began. This phase that continued through late July 1998, although essentially similar to the previous one, was characterized by violent explosive activity at cone A while the B vents showed much milder activity than during 1997. Some time before the 22 July 1998 paroxysmal eruption of the Voragine (see below), the collapse depression on the southern side of cone had already been filled and growth was proceeding vigorously at the summit of the cone, being most rapid in the days after the 22 July event. However, it did never fully recover to its pre-collapse size, and renewed collapse began in early September 1998. In October, a pronounced subsidence depression had formed in the same area where subsidence had been most pronounced in early 1998.

Observation of the patterns of subsidence in early 1998 provided new insights into the dynamics and internal structure of a magma column within a wide conduit. While the magma column may be in a more or less plastic state over the full diameter of the conduit at some depth, the top of the column grades towards the surface into more rigid zones consisting of congealed magma and, in its most superficial part, of collapse breccia and products of previous eruptive activity. It is not clear if, and to what degree, this uppermost zone is altered by the rise and intrusion of new magma; in the case of the 1995-1997 activity it appears to have acted as a fairly rigid piston-shaped block sitting on top of the magma column. It was perforated by the two degassing conduits at A and B where magma rose to the surface, first at A which lay closer to the center of the crater, then at B and eventually at C, both situated on the rims of the rigid piston. The causes of the subsidence of the magma column in early 1998 are not clear; intense seismicity during late December 1997 and mid-January 1998 may have been caused by the injection of magma below the western flank of the volcano, triggering the withdrawal of magma from the conduit below Bocca Nuova.

While collapse initially was focused at the central portion of cone A, the center of subsidence later shifted to the southern part of the cone, lying roughly in the center of the crater floor. Similarly, collapse at the B vents initially affected the summit portions of the pyroclastic edifice around the vents, but later the whole structure subsided, being carried downward as the entire piston tapping the conduit sank within the crater by tens of meters. It is clear that the entire crater floor, with a diameter of 250-400 m, had become mobile, pointing to a magma column of equal dimensions at least during early 1998. At the resumption of crater filling in mid-1998, the rising magma column not only caused renewed eruptive activity but also a slight uplift of the crater floor as a whole. This means that the piston that formerly sat fixed within the uppermost part of the conduit has detached from the conduit walls and may more freely move with the vertical fluctuations of the magma column.

From October 1998 until early February 1999, the activity of the Bocca Nuova consisted of discrete explosions from one main vent at A that produced jets of incandescent lava fragments up to 100 m high, and near continuous spattering and occasional small Strombolian explosions from one to three vents at B. As far as is known no new lava effusion occurred on the crater floor. Shortly after the beginning of effusive activity at the base of the SE Crater cone in early February 1999 (see below), the magma level within the conduits of the Bocca Nuova dropped, and collapse of the vent walls produced near continuous and noiseless emission of fine, lithic ash. This kind of activity continued until June and was followed by a slow rise of the magma level, accompanied by small, deep-seated explosions mainly in the A vents, but no eruptive activity of appreciable magnitude occurred at this crater until September 1999.

On the afternoon of 4 September 1999, an extremely violent paroxysm occurred at the nearby Voragine. This event had clear repercussions on the activity not only of the nearby Southeast Crater, which produced a much smaller paroxysm on its own a few hours later, but also on that of the Bocca Nuova. Powerful explosions were observed at this crater for several hours, which showered much of the summit area with large bombs. A small landslide on the south flank of the central summit cone was probably caused by this activity.

During the days after the 4 September 1999 paroxysm the Bocca Nuova produced frequent explosions that ejected bombs outside the crater and thus made any approach to the crater rim extremely dangerous. During an interval of relatively mild activity on 19 September magma was observed to have risen to the lip of the main vent at A, and Strombolian explosions were nearly continuous. During the early morning hours of 20 September a paroxysmal eruptive episode occurred at the A vents: lava fountains jetted hundreds of meters high, and lava covered the entire crater floor to a depth of several tens of meters, raising the crater floor to only 30 m below the lowest point on the W crater rim. Lava spilled into the pit at the B vents and completely concealed them; the cone at A was buried, leaving only an irregular shallow depression as an indicator of where the A vents had been. After 21 September the Bocca Nuova remained totally quiet - the ominous quiet before the storm. During the 2 weeks of inactivity it was possible to look into this crater for the last time. It was clear that in the case of renewed activity on the scale of the 20 September event it would have taken little to fill the crater to its rim and allow for the first overflow over the low point on the W crater rim.

The October-November 1999 eruption

On 5 October, loud detonations shook windows in the towns around the volcano, and plumes rising from the Bocca Nuova indicated that the period of unusual silence had ended. While it was impossible to reach the crater rim, it was evident that most activity occurred in the area of the B vents, and the explosions apparently blasted away the cover of lava of 20 September, allowing fresh magma to reach the surface. During the following week, the activity at the Bocca Nuova continued, involving more and more the area of the A vents, and by the 12th a pulsating lava fountain rose from one vent in the W part of the crater while more sporadic, but far more violent, explosions from the area of the A vents ejected both bombs and lithic blocks, many of which fell several hundred meters from the crater rim. Only rare and relatively weak explosions occurred at the B vents.

Between 13 and 17 October the activity of the Bocca Nuova continued with some fluctuations, but at generally very intense levels, becoming more episodic in character towards the 16th. During the forenoon of 17 October lava fountains rose again from the most active vent in the W part of the crater, marking the onset of yet another paroxysmal episode. Later that day, probably during the late afternoon, the level of lava filling the crater had risen close to the low point on the W rim, and the pressure of the lava accumulated behind that rim caused a small section of it to collapse, in a manner very similar to what had occurred on the SSW rim of the Central Crater on 13 May 1964. Lava rapidly spilled through the newly formed breach, covering the W flank of the main summit cone, burying the dirt road running around the W side of the summit area over a width of about 100 m, and then flowing down the steep upper W flank of the mountain, closely following the path of the lava flows emitted in early May 1964 from the Central Crater. The new lavas reached the slope break at about 2000 m elevation in a few hours but then slowed before reaching the tree limit and the dirt road of the Etnean forest guard ("Forestale road").

During the following two weeks the Bocca Nuova continued to provide an exceptional spectacle to visitors and residents of the Etna area. Eruptive activity varied from isolated, short-lived and very violent eruptive episodes separated by up to 24 hours of nearly complete quiet, to continuous, vigorous explosive-effusive activity (ranging from Strombolian to Hawaiian in character) which could last several days. This relatively irregular activity produced frequent overflows of lava towards W, at times relatively modest, but on some occasions of major proportions, resulting in spectacular surges of lava running down the W flank in many lobes, a display to be enjoyed particularly after sunset when observing the volcano from the area of the town of Bronte. Four main surges of lava occurred on 17-18, 19-20, 25-26 and 27-28 October. The latest of these produced the longest flows of the entire interval, the longest flows erupted from the area of the former Central Crater since 1787, which buried a section of the Forestale road about 100 m long, and destroyed between 100,000 and 500,000 m2 of forest. However, the most vigorous activity, in terms of explosivity and fountain heights, occurred during the forenoon of 22 October when dense jets of lava roared to about 400 m above the crater rim, apparently fed by numerous vents aligned along a fissure trending roughly N-S in the W part of the Bocca Nuova.

Intense, and near continuous eruptive activity continued through 5 November. The main focus of activity was in the NW part of the Bocca Nuova, roughly coinciding with the location of the A vents of 1997-1998, where a conspicuous pyroclastic cone was built which eventually rose several tens of meters above the NW rim of the crater. In the SE part of the Bocca Nuova, the B vents which had shown little activity in late October produced dense fountains of ash and blocks, but most if not all of this material was lithic, since no incandescence was visible at these vents at night. The emission of ash may have signaled the beginning of subsidence of the magma column in the conduits below the Bocca Nuova. However, lava continued to flow from the Bocca Nuova as of late 1 November; no major lava effusion occurred thereafter. Episodes of powerful explosive activity with high jets of lava were observed until the night of 3-4 November after which the activity dropped to relatively low levels.

Map of the lava flows emitted from the Bocca Nuova during October-November 1999, based on photos taken from various locations. Main vents of the Bocca Nova (BN) are shown as orange colored dots. The other summit craters are the Northeast Crater (NE), Voragine (V) and Southeast Crater (SE). The inset at upper left shows the entire Etna area with the location of the new lavas and the towns of Bronte and Catania. VDB in the inset is Valle del Bove.

The activity of October-November 1999 had overwhelming effects on the morphology of the summit area. The Bocca Nuova practically ceased to exist, in the place of the yawning chasm it had been throughout much of its existence, a kind of platform was formed, pierced by the two main vents in its NW and SE parts (and corresponding to the A and B vent areas established as early as 1995). It is worth mentioning, however, that during some of the major eruptive episodes of October-November 1999 (such as that of 22 October), many more vents were active, and it appears that these were aligned along a fissure trending roughly N-S, or NW-SE. The accumulation of pyroclastics around the most active vent (or group of vents) at A gradually built a large scoria cone, whose growth was apparently often interrupted by major surges of lava that swept away portions of the cone. Following the cessation of intense eruptive activity after 4 November 1999, some collapse occurred at the vents, both at A and B, creating new pits and engulfing part of the large cone at A. Each of these pits was about 100 m in diameter, and observations in mid-December 1999 revealed that they were at least 100 m deep, with occasional explosions on their floors too weak to eject pyroclastics as high as their lips.

The W side of the main summit cone was covered by a voluminous succession of lavas, probably with thicknesses of up to several tens of meters. In the course of the activity, overflows occurred practically over the whole width of the W rim of the Bocca Nuova, in a sector about 200 m wide; a few hundred meters downslope the flows fanned out to a width of about 1 km and then divided into dozens of major and lesser branches on the steep W flank of the main edifice of Etna. An extensive lava flow-field was emplaced on this slope, with a maximum width of about 2.5 km and a the longest flow extending more than 5 km from the W rim of the Bocca Nuova. The 1999 lavas were thus far more extensive than those of 1964 in the same area. The volume of the lavas emitted in October-November 1999 from the Bocca Nuova is about 15 million m³ (to which about 2.5 million m³ of lava must be added that filled the crater before the first overflow), corresponding to a mean effusion rate of 10-15 m³ per second, but peak effusion rates during the major eruptive episodes (such as that of 27 October) may have exceeded 100 m³ per second.

The overflow of lava from the Bocca Nuova was at times accompanied by collapse of the unstable crater rim under the weight of the overflowing lava, or caused by the pushing of the lava fill behind the crater rim. Some of these collapse events were highly complex and were accompanied by phenomena not seen before on Etna or anywhere else. Shortly after noon on 25 October, a section of the WNW rim of the Bocca Nuova collapsed, generating a series of avalanches that advanced hundreds of meters down the slope across flowing lavas. This collapse was accompanied by the uplift (by several tens of meters) and inverse rotation of a block on the crater rim caused by the intrusion of lava from the crater below the uplifting block; it was amazing to see that the block was not carried away from the crater on top of the lava passing underneath, but simply pushed upwards and rotated backwards at the same time. Lava was seen to ooze from the base of the block while it was still being uplifted. At the same time, portions of the block crumbled, generating numerous rockslides and avalanches that resembled pyroclastic flows (nu�es ardentes) and advanced at high speed, generating turbulent brownish dust clouds. Although near continuous strong explosions and lava fountains were produced by one or two vents at no more than 150 m distance from the mobile block, the collapse was not caused by them, but rather by the movement of lava at the base of the block.

At about 1300 h on 25 October, a peculiar lava dome rose behind the uplifted block on the crater rim, which generated even longer pyroclastic avalanches (up to 700 m long). This dome probably consisted of material that had filled the crater during the preceding three weeks, which can be imagined as a viscous mixture of lava and pyroclastics. It was the viscosity of this material that made it behave like any silicic lava dome, although its chemical composition was basaltic. Collapse of the steep outer flank of the dome produced rockfalls which transformed into fast (up to more than 70 km/h) pyroclastic avalanches. No one knows how hot they were, because they moved on top of still-advancing lava flows, which luckily prevented that observers stood in their path. Any deposits left by the avalanches were instantaneously remobilized by the moving lava underneath, and fresh lavas covered the area within a few hours so that no studies of the deposits were possible.

The Bocca Nuova in 2000-2001

Mild Strombolian activity occurred during the first weeks of 2000, at times interrupted by episodes of stronger activity. Shortly after the beginning of the incredible series of paroxysmal eruptive episodes at the Southeast Crater in late January 2000, the activity in the Bocca Nuova changed to vigorous degassing without ejection of pyroclastics. Instead, much of the degassing occurred in the form of brief, loud explosions at depth which caused a unique phenomenon known as "smoke rings". Actually there is nothing like smoke coming out of any volcano, and the rings are made of gas, which is why they should be called "gas rings". Between February and late June 2000, the Bocca Nuova produced thousands of gas rings, providing visitors with a rare sight, since such rings do not frequently occur at volcanoes. This, together with the dozens of awe-inspiring episodes of lava fountaining a the Southeast Crater, made the year 2000 a very special one for Etna. Both phenomena had been observed before at this volcano, but never in similar quantities and frequency. Some visitors were even fortunate enough to see gas rings slowly drift over the Southeast Crater while a majestic lava fountain roared hundreds of meters above its summit!

Both the lava fountains and the gas rings essentially ceased between late June and mid-July 2000, but in late September of that year, Strombolian activity resumed within the two pits of the Bocca Nuova and continued without interruptions until mid-December. Some filling of the pits took place, but differently from 1999 this activity did not evolve into a full-scale eruption. A similar period of activity began in late January 2001 and continued through April. The NW pit (A) was filled to within about 50 m of its rim before collapse in its SE part created a new pit-within-a-pit; in a similar manner pit B was seen to be only about 40 m deep in late March 2001. No magmatic activity occurred at the Bocca Nuova after April 2001, and the crater was essentially unaffected by the activity which preceded the July-August flank eruption and by that eruption itself. In late 2001 the morphology of the Bocca Nuova was much the same as in late March, except that pit B was much deeper and its bottom was not visible from the rim.

Copyright © Boris Behncke, "Italy's Volcanoes: The Cradle of Volcanology"