Thomas C. Pierson

Perturbation and melting of snow and ice by the 13 November 1985 eruption of Nevado del Ruiz, Colombia, and consequent mobilization, flow and deposition of lahars

A complex sequence of pyroclastic flows and surges erupted by Nevado del Ruiz volcano on 13 November 1985 interacted with snow and ice on the summit ice cap to trigger catastrophic lahars (volcanic debris flows), which killed more than 23,000 people living at or beyond the base of the volcano. The rapid transfer of heat from the hot eruptive products to about 10 km2 of the snowpack, combined with seismic shaking, produced large volumes of meltwater that flowed downslope, liquefied some of the new volcanic deposits, and generated avalanches of saturated snow, ice and rock debris within minutes of the 21:08 (local time) eruption. About 2 × 107 m3 of water was discharged into the upper reaches of the Molinos, Nereidas, Guali, Azufrado and Lagunillas valleys, where rapid entrainment of valley-fill sediment transformed the dilute flows and avalanches to debris flows. Computed mean velocities of the lahars at peak flow ranged up to 17 m s−1. Flows were rapid in the steep, narrow upper canyons and slowed with distance away from the volcano as flow depth and channel slope diminished. Computed peak discharges ranged up to 48,000 m3 s−1 and were greatest in reaches 10 to 20 km downstream from the summit. A total of about 9 × 107 m3 of lahar slurry was transported to depositional areas up to 104 km from the source area. Initial volumes of individual lahars increased up to 4 times with distance away from the summit. The sedimentology and stratigraphy of the lahar deposits provide compelling evidence that: (1) multiple initial meltwater pulses tended to coalesce into single flood waves; (2) lahars remained fully developed debris flows until they reached confluences with major rivers; and (3) debris-flow slurry composition and rheology varied to produce gradationally density-stratified flows. Key lessons and reminders from the 1985 Nevado del Ruiz volcanic eruption are: (1) catastrophic lahars can be generated on ice- and snow-capped volcanoes by relatively small eruptions; (2) the surface area of snow on an ice cap can be more critical than total ice volume when considering lahar potential; (3) placement of hot rock debris on snow is insufficient to generate lahars; the two materials must be mechanically mixed together for sufficiently rapid head transfer; (4) lahars can increase their volumes significantly by entrainment of water and eroded sediment; and (5) valley-confined lahars can maintain relatively high velocities and can have catastrophic impacts as far as 100 km downstream.

Debris flow rheology: Experimental analysis of fine-grained slurries

The rheology of slurries consisting of ≤2-mm sediment from a natural debris flow deposit was measured using a wide-gap concentric-cylinder viscometer. The influence of sediment concentration and size and distribution of grains on the bulk rheological behavior of the slurries was evaluated at concentrations ranging from 0.44 to 0.66. The slurries exhibit diverse rheological behavior. At shear rates above 5 s−1 the behavior approaches that of a Bingham material; below 5 s−1, sand exerts more influence and slurry behavior deviates from the Bingham idealization. Sand grain interactions dominate the mechanical behavior when sand concentration exceeds 0.2; transient fluctuations in measured torque, time-dependent decay of torque, and hysteresis effects are observed. Grain rubbing, interlocking, and collision cause changes in packing density, particle distribution, grain orientation, and formation and destruction of grain clusters, which may explain the observed behavior. Yield strength and plastic viscosity exhibit order-of-magnitude variation when sediment concentration changes as little as 2–4%. Owing to these complexities, it is unlikely that debris flows can be characterized by a single rheological model.

Initiation and flow behavior of the 1980 Pine Creek and Muddy River lahars, Mount St. Helens, Washington

Two large, high-velocity lahars (volcanic debris flows) were triggered by a pyroclastic surge during the first few minutes of the May 18, 1980, eruption of Mount St. Helens. The initial surge cloud evolved progressively by gravity segregation from a gas-mobilized, highly inflated density flow to a dense, water-mobilized, basal debris flow (lahar) and accompanying ash cloud as it flowed down the east flank of the volcano. The main source of the water for the lahars was probably from eroded snow and ice incorporated into the flow by turbulent mixing, but ground water, expelled together with the rock debris by the initiating volcanic explosions, also may have contributed. Peak lahar discharge from the Pine-Muddy fan, upper Smith Creek, and Ape Canyon probably exceeded 250,000 m 3 /s initially but decreased exponentially in the downstream direction. Total volume of the lahars was in excess of 1.4 × 10 7 m 3 . Initial peak-flow velocities in excess of 30 m/s also decreased markedly downstream. Where flow was not impeded, velocity was strongly related to the depth-slope term (R 2/3 S 1/2 ) from the Manning uniform-flow equation as a power-law function. During much of the route traveled, lahar flow appears to have been supercritical. Deposits left in channels were generally thin relative to flow depth (0 to 2.5 m). Particles up to small boulder size were randomly distributed in the poorly sorted, nonstratified matrix, indicating complete suspension in a fully developed debris-flow slurry; however, much larger clasts were transported as “bed-load.” Computed sediment concentrations of matrix slurry samples ranged from 84% to 91% solids by weight and were similar for the two lahars. Two indirect methods for computing peak-flow velocity, previously only tentatively applied to debris flows, were tested for accuracy by comparing computed lahar arrival times with recorded arrival times at Swift Reservoir. The computed velocities appear to be ∼15% slower than the recorded velocities, which is consistent with the restriction that the velocity formulas produce only minimum values.

Sediment yield following severe volcanic disturbance—A two-decade perspective from Mount St. Helens

Explosive volcanic eruptions perturb water and sediment fluxes in watersheds; consequently, posteruption sediment yields can exceed preeruption yields by several orders of magnitude. Annual suspended-sediment yields following the catastrophic 1980 Mount St. Helens eruption were as much as 500 times greater than typical background level, and they generally declined nonlinearly for more than a decade. Although sediment yields responded primarily to type and degree of disturbance, streamflow fluctuations significantly affected sediment-yield trends. Consecutive years (1995–1999) of above-average discharge reversed the nonlinear decline and rejuvenated yields to average values measured within a few years of the eruption. After 20 yr, the average annual suspended-sediment yield from the 1980 debris-avalanche deposit remains 100 times (10 4 Mg [megagrams]/km 2 ) above typical background level (~10 2 Mg/km 2 ). Within five years of the eruption, annual yields from valleys coated by lahar deposits roughly plateaued, and average yields remain about 10 times (10 3 Mg/km 2 ) above background level. Yield from a basin devastated solely by a blast pyroclastic current diminished to background level within five years. These data demonstrate longterm instability of eruption-generated detritus, and show that effective mitigation measures must remain functional for decades.

Flow characteristics of large eruption-triggered debris flows at snow-clad volcanoes: constraints for debris-flow models

Abstract Theoretical modelling is not yet adequate to predict the behavior of debris flows, which can be an extremely hazardous hydrologic process commonly associated with volcanic eruptions, particularly at snow-clad stratovolcanoes. To provide a realistic basis for modelling the behavior of large (> 1000 m3/s) debris flows, this paper summarizes kinematic, volumetric and hydraulic characteristics of ten large historic volcanic debris flows from four different volcanoes. Although debris flows larger than these are known to have occurred in the past, the ten summarized here define a practical upper range in magnitude of more typical flows to be considered for future hazard prediction. Peak flow velocities of the ten debris flows studied were indirectly measured to be between about 5 and 20 m/s on gradients of 0.005 to 0.25 m/m, but locally they were as great as 40 m/s. Hydraulic (average) depths were as great as 25 m, but were more commonly between 5 and 15 m in channels up to 400 m wide. Computed peak discharges (volumetric flow rates) were as high as 105 m3/s, and total volumes were as much as 108 m3. Total flow volumes increased by as much as four times in relatively steep channels as eroded sediment was incorporated into the debris flows. Flows generally achieved supercritical flow and deposited minimal volumes of sediment on gradients steeper than 0.02 m/m. Subcritical flow and active deposition predominated on gradients less than 0.01 m/m, although flows travelled tens of kilometers on such low gradients while laying down deposits. Total distances travelled (as debris flows) were as far as 120 km.

Magnitude and timing of downstream channel aggradation and degradation in response to a dome-building eruption at Mount Hood, Oregon

A dome-building eruption at Mount Hood, Oregon, starting in A.D. 1780 and lasting until ca. 1793, produced dome-collapse lithic pyroclastic flows that triggered lahars and intermittently fed 10 8 m 3 of coarse volcaniclastic sediment to sediment reservoirs in headwater canyons of the Sandy River. Mobilization of dominantly sandy sediment from these reservoirs by lahars and seasonal floods initiated downstream migration of a sediment wave that resulted in a profound cycle of aggradation and degradation in the lowermost reach of the river (depositional reach), 61–87 km from the source. Stratigraphic and sedimentologic relations in the alluvial fill, together with dendrochronologic dating of degradation terraces, demonstrate that (1) channel aggradation in response to sediment loading in the headwater canyons raised the river bed in this reach at least 23 m in a decade or less; (2) the transition from aggradation to degradation in the upper part of this reach roughly coincided with the end of the dome-building eruption; (3) fluvial sediment transport and deposition, augmented by one lahar, achieved a minimum average aggradation rate of ∼2 m/yr; (4) the degradation phase of the cycle was more prolonged than the aggradation phase, requiring more than half a century for the river to reach its present bed elevation; and (5) the present longitudinal profile of the Sandy River in this reach is at least 3 m above the pre-eruption profile. The pattern and rate of channel response and recovery in the Sandy River following heavy sediment loading resemble those of other rivers similarly subjected to very large sediment inputs. The magnitude of channel aggradation in the lower Sandy River, greater than that achieved at other volcanoes following much larger eruptions, was likely enhanced by lateral confinement of the channel within a narrow incised valley. A combination of at least one lahar and winter floods from frequent moderate-magnitude rainstorms and infrequent very large storms was responsible for flushing large volumes of sediment to the depositional reach. These conditions permitted a sedimentation response in the Sandy River that approached the magnitude of channel aggradation resulting elsewhere from large explosive eruptions and high-intensity rainfall regimes, despite the fact that the Sandy River aggradation was in response to an unremarkable dome-building eruption in a climate dominated by low to moderate rainfall intensities.

Unusual ice diamicts emplaced during the December 15, 1989 eruption of redoubt volcano, Alaska

Abstract Ice diamict comprising clasts of glacier ice and subordinate rock debris in a matrix of ice (snow) grains, coarse ash, and frozen pore water was deposited during the eruption of Redoubt Volcano on December 15, 1989. Rounded clasts of glacier ice and snowpack are as large as 2.5 m, clasts of Redoubt andesite and basement crystalline rocks reach 1 m, and tabular clasts of entrained snowpack are as long as 10 m. Ice diamict was deposited on both the north and south volcano flanks. On Redoubts north flank along the east side of Drift piedmont glacier and outwash valley, ice diamict accumulated as at least 3 units, each 1–5 m thick. Two ice-diamict layers underlie a pumice-lithic fall tephra that accumulated on December 15 from 10:15 to 11:45 AST. A third ice diamict overlies the pumiceous tephra. Some of the ice diamicts have a basal ‘ice-sandstone’ layer. The north side icy flows reached as far as 14 km laterally over an altitude drop of 2.3 km and covered an area of about 5.7 km 2 . On Crescent Glacier on the south volcano flank, a composite ice diamict is locally as thick as 20 m. It travelled 4.3 km over an altitude drop of 1.7 km, covering about 1 km 2 . The much higher mobility of the northside flows was influenced by their much higher water contents than the southside flow(s). Erupting hot juvenile andesite triggered and turbulently mixed with snow avalanches at snow-covered glacier heads. These flows rapidly entrained more snow, firn, and ice blocks from the crevassed glacier. On the north flank, a trailing watery phase of each ice-diamict flow swept over and terraced the new icy deposits. The last (and perhaps each) flood reworked valley-floor snowpack and swept 35 km downvalley to the sea. Ice diamict did not form during eruptions after December 15 despite intervening snowfalls. These later pyroclastic flows swept mainly over glacier ice rather than snowpack and generated laharic floods rather than snowflows. Similar flows of mixed ice grains and pyroclastic debris resulted from the November 13, 1985 eruption of Nevado del Ruiz volcano and from eruptions of snowclad Mount St. Helens in 1982–1984. Such deposits at snowclad volcanoes are initially broad and geomorphically distinct, but they soon become extensively reworked and hard to recognize in the geologic record.

Interdisciplinary Studies of Eruption at Chaitén Volcano, Chile

High-silica rhyolite magma fuels Earths largest and most explosive eruptions. Recurrence intervals for such highly explosive eruptions are in the 100- to 100,000-year time range, and there have been few direct observations of such eruptions and their immediate impacts. Consequently, there was keen interest within the volcanology community when the first large eruption of high-silica rhyolite since that of Alaskas Novarupta volcano in 1912 began on 1 May 2008 at Chaiten volcano, southern Chile, a 3-kilometer-diameter caldera volcano with a prehistoric record of rhyolite eruptions [Naranjo and Stern, 2004semi; Servicio Nacional de Geologia y Mineria (SERNAGEOMIN), 2008semi; Carn et al., 2009; Castro and Dingwell, 2009; Lara, 2009; Munoz et al., 2009]. Vigorous explosions occurred through 8 May 2008, after which explosive activity waned and a new lava dome was extruded.

Reducing risk from lahar hazards: concepts, case studies, and roles for scientists

Lahars are rapid flows of mud–rock slurries that can occur without warning and catastrophically impact areas more than 100 km downstream of source volcanoes. Strategies to mitigate the potential for damage or loss from lahars fall into four basic categories: (1) avoidance of lahar hazards through land-use planning; (2) modification of lahar hazards through engineered protection structures; (3) lahar warning systems to enable evacuations; and (4) effective response to and recovery from lahars when they do occur. Successful application of any of these strategies requires an accurate understanding and assessment of the hazard, an understanding of the applicability and limitations of the strategy, and thorough planning. The human and institutional components leading to successful application can be even more important: engagement of all stakeholders in hazard education and risk-reduction planning; good communication of hazard and risk information among scientists, emergency managers, elected officials, and the at-risk public during crisis and non-crisis periods; sustained response training; and adequate funding for risk-reduction efforts. This paper reviews a number of methods for lahar-hazard risk reduction, examines the limitations and tradeoffs, and provides real-world examples of their application in the U.S. Pacific Northwest and in other volcanic regions of the world. An overriding theme is that lahar-hazard risk reduction cannot be effectively accomplished without the active, impartial involvement of volcano scientists, who are willing to assume educational, interpretive, and advisory roles to work in partnership with elected officials, emergency managers, and vulnerable communities.

Extraordinary sediment delivery and rapid geomorphic response following the 2008–2009 eruption of Chaitén Volcano, Chile

U.S. Geological Survey Volcano Science Center SERNAGEOMINs Programa de Riesgos Volcanicos Conicyt Fondecyt grants 1110609 1141064 11130671 Conicyt Fondap 15090013 Vamos Research Centre