The 1998 debris avalanche at Volcan Casita
Abstract
During Hurricane Mitch in 1998, a debris avalanche occurred at Casita volcano, Nicaragua, resulting in a lahar that killed approximately 2500 people. The failure that initiated the avalanche developed at a pre-existing cliff, part of the headwall of a gravitational slide of approximately 1,8 km square in plan view that cuts the southern flank of the volcano.
Structural analysis, primarily based on a high-resolution DEM, has shown that this slide is caused by edifice deformation. This deformation is possibly facilitated by millennia of persistent hydrothermal alteration of the volcano’s core. The gravity slide has some typical features of smaller slumps, such as steep headwalls, an inner flatter area and a pronounced basal bulge fronted by thrusts. The headwall is the source of the 1998 avalanche, as well as several previous mass movements.
The analysis presented here shows how small but highly damaging landslides can occur during the deformation of the volcanic edifice.
Introduction
Heavy rainfall associated with the hurricane contributed to the mobilisation of fractured and hydrothermally altered material on a pre-existin cliff at Casita volcano in northwestern Nicaragua (Ferraro et al., 1999). It resulted in a debris avalanche, which descended the southern flank in pre-existing channels. Several minutes after the initial slope failure the two towns of El Porvenir and Rolando Rodriguez, as well as several smaller hamlets in the lowland south of the volcano, were destroyed by a lahar. A total area of about 12 km square was largely stripped of human life and vegetation (Fig. 1).
The 1980 Mt. St. Helens sector collapse and avalanche event sparked widespread research into volcanic landslides, which have been shown to transform frequently into debris avalanches and debris flows. Fig. 2 shows that the source volume of the primary failure of Casita was comparatively small. However, the landslide immediately disintegrated and entrained older deposits and organic material, transforming into a devastating debris avalanche. It is now known that debris avalanches of Casita-size occur every few years at one volcano or another in the world (Siebert, 1992; McGuire, 1996).
Geology of Casita
Casita is a dormant prehistoric part of the San Cristobal volcanic complex, which rises from the Nicaragua Depression as part of the northern Cordillera de Marabios (Fig. 1). It rises from the western edge of the La Pelona caldera and stands on La Pelona pumice horizons.
Unlike San Cristobal, Casita is not a single cone but an EW-oriented cratered ridge , along which activity has migrated from west to east, creating at least three independent summit craters. Further west, remains of a possible fourth, heavily eroded, crater can be seen (Fig. 6A).
Casita’s summit ridge is formed by cones of lava and tephra, which are cut by normal faults running in a sub-radial fashion from the summit (van Wyk de Vries et al., 2000).
The 1998 debris avalanche
At the Chinandega weather station, at 20km distance the closest to Casita, 1217mm of precipitation were registered over a 3 day period, with 485 mm on October 30 alone.
By 11:00 AM on October 30 more than 700 mm of precipitation had been recorded for the previous 48 h. Between 10:30 and 11:00 a part of the southwestern flank, some 250 meters SW and 100 meters below the summit, composed of hydrothermally altered and highly fractured rock, broke off (Fig. 3A). A rockslide of about 200,000 cubic meters (110x60x30 meters) descended the slope, and was joined by a second flank section, located just on the western edge of the primary source. About 200 meters downslope from the base of the primary source a branching of the mass occurred, where a small fraction was redirected towards the southeast, while the main avalanche continued south (Fig. 4A and inset). It reached a platform about 1.5km farther south (Fig. 6, profile 5 and Fig. 7, point m), spread out, and continued on towards the Panamerican Highway. On the way it destroyed the towns of El Porvenir and Rolando Rodriguez. The first news reports from the disaster area described a large mudflow that had destroyed the towns. This implies a complete transformation of the initial rockslide into a lahar along a 6 km long course.
The branching of the rockslide close to the source indicates early and rapid fragmentation of the source rock. After a descent of about 700 meters in altitude and an average slope angle of 18 degrees, just above the platform there are signs of strong upslope flow along the western side of the valley (Fig. 3b), already indicating flow - rather than slide characteristics of the moving mass, as well as high speed. … Of the dense forest that had covered the valley above and including the platform, only a small island of trees was spared from destruction (Fig. 3B). The trees show snapped-off branches at heights of more than 10 meters, apparently broken off by flying rocks, also indicating a high-speed avalanche.
Survivors of the disaster reported having heard first an explosio (Morales, 1998), within minuted of which the lahar reached the two towns, destroying both without leaving any trace. … Given a nearly uniform slope angle of approximately 18 degrees, the upslope flow marks above the platform, the damage to the trees, and a distance of some 6 km between the source and the destroyed towns that was covered by the avalanche in a few minutes, we suggest a rapid transformation of a rockslide into an uninterrupted, high-velocity lahar.
Description of the deposit stratigraphy
… There are signs of strong erosion into previous avalanche deposits in a deep gully that dominates the western edge of the valley, as well as high-upslope flow along its western slope, indicating that most of the avalanche material was transported through that gully (Fig. 3B and Fig. 4A, inset). The areas outside the ravine course are marked by accumulation of avalanche material. Remaining tree stumps show that, although the avalanche completely destroyed the forest, erosion in these parts was not strong enough to remove the upper soil layer to root depth.
After passing the platform, the ̄ow spread to a width of about 1200 m and continued downslope into settled areas and on towards the Panamerican Highway. The platform became an accumulation area for some of the lower density material (Fig. 3B). While the gully is virtually free of avalanche materials, numerous tree trunks and other debris blanket the platform, suggesting a scum layer of light debris, which could be a result of density stratification. Most of the avalanche continued down the gully into the plain below. Sheridan et al. estimated a central flow height of about 3 meters as it entered the towns of El Porvenir and Rolando Rodriguez.
Incision of the valley above the platform has exposed the stratigraphy below the 1998 deposits. Visual interpretation allowed the identification of at least one previous large landslide that descended the same valley. A cliff just below the landslide detachment exposes the following layers (Fig. 5 and Fig. 7, point p):
(A) Just above the gully base there is a fine-grained sandy layer, with indistinct laminations, suggesting deposition by fluvial processes.
(B) Layer A is covered by an approximately 1.3 m thick deposit, composed of randomly orientated clasts up to 0.3 m across in a coarse, sandy matrix. The clasts are highly irregular and angular, typical for debris avalanche deposits. High colour variation, including white, grey, yellow, red and brown, indicate several different rock types. We suggest that this unit is composed of material deposited during a previous event similar to the 1998 avalanche.
(C) Layer B is overlain by a finer-grained layer, which has the same matrix appearance as the lower unit, but lacks the large blocks. There is no clear boundary between B and C. The upper part of the layer is light grey, made up of about 40% light grey to white, angular fragments in a brown sandy matrix. The layer has a sharp, irregular top.
(D) This layer has a brown, sandy matrix containing about 10% pebbles and cobbles up to 0.15 m in diameter. The clasts are less angular than those in the deposits farther downslope. The top is level and regular apart from where small clasts protrude into the base of the overlying layer.
(E) This is a sandy, banded layer of medium brown colour with an irregular, eroded top.
(F) This top layer consists of the 1998 avalanche deposits. It is composed of a grey/brown sand- to gravel-size matrix supporting about 30% clasts of up to about 0.5 m diameter. The clasts are irregular and of different colours, suggesting a variety of source rocks or levels of hydrothermal alteration. The features of the deposit resemble closely that of layer B, apart from having fewer coarse clasts.
Kerle et. al, The 1998 debris avalanche at Casita volcano, Nicaragua - investigation of structural deformation as the cause of slope instability using remote sensing, Journal of Volcanology and Geothermal Research, June 2000