Navigation – Plan du site

Volcanic landforms, processes and hazards on volcanoes

Jean-Claude Thouret, David Chester, Angus Duncan et Cliff Ollier
p. 211-215

Texte intégral

1Six years after the special issues No. 2 and 3 (2001) of our Journal were devoted to « volcanic geomorphology »,and two years after the Supplement Band 140 of Zeitschrift für Geomorphologie, this new volume focuses once again on the geomorphology of volcanoes. This special issue contains six of the twelve contributions that were presented during the session entitled ‘Landforms, processes, and hazards on volcanoes’, held at the VIth International Conference of Geomorphologie in Zaragoza (Spain) during September 2005. This volume also includes some articles that have been submitted following the Conference. The variety of case studies reflects the activities of the former Working Group of the International Association of Geomorphologists (1997-2005), whose members have studied the ways in which volcanic landforms are shaped by volcanic rocks.

2Six selected papers have resulted from the research projects which have been led by 17 authors drawn from nine countries (Ecuador, France, Hungary, Japan, Mexico, New Zealand, Portugal, United Kingdom, and the United States). Contributions deal with three themes: (1) how processes of erosion may be measured in volcanic rocks by using volcano landforms and deposits as geomorphic and chronologic tools; (2) how volcanic edifices can be eroded by rapid mass wasting, or by other processes of erosion; and (3) how volcanic, seismic and geomorphic processes may induce risk and produce damage in populated areas on and around unstable, active volcanic edifices.

3Measuring the rates by which processes of erosion operate in volcanic rocks and of the deposits they produce can be undertaken either by means of field observations, or by in situ and laboratory experiments. Two Japanese researchers at the University of Tokyo, Y. Hidetsugu and T. Sugai, examine how large-volume mass movements, such as the 24 ka-old debris avalanche produced by Asama volcano in central Japan, may be emplaced over exceptionally long runout distances. The authors show that this debris avalanche was channeled down valley as far as 90 km from the volcano summit, as a single event without loosing its rheological characteristics as a gravity-driven flow. They emphasise that topographic control was exerted by the V-shaped morphology of the steep valley of the River Agatsuma and that this acted on the rapid emplacement of these mass movement sediments, which had a volume of c. 5 km3. This valley morphology helped to confine and rapidly transfer the gravity-driven flow towards the large and flat lowland of Kanto downstream, where the related debris avalanche had spread out and its deposit had thinned. Based on the geometry of the pre-existing valley and on the lithofacies of the debris-avalanche deposit, this study enables the authors to reconsider the processes by which a volcanic edifice may largely be eroded by sector collapses. They claim that geomorphologists should take account of regional geomorphic settings, particularly of radial valleys that drain volcanoes located on volcanic island arcs in wet environment, such as those found in Japan. This approach can reveal the effects of both valley morphology and fluvial processes on debris avalanches. These effects may also explain how avalanches have been transported and why they reach unusual distances.

4Geomorphologists can measure processes of erosion using experiments either in situ on active volcanoes, or by using simulation analysis carried out in the laboratory. Geomorphologists have long used volcanic deposits, such as lava flows, to reconstruct the topography created by erosion after an eruption and to infer erosion rates over tens of thousands of years to millions of years. It is useful, however, to highlight the pitfalls that should be avoided when applying this well established method. Using their vast experience and knowledge acquired on monogenetic edifices, particularly these on phreatomagmatic volcanoes in New Zealand and in Hungary, K. Nemeth, U. Martin et G. Csillag stress that two methods should be used to reconstruct buried, pre-eruption topographies and to infer erosion rates in volcanic rocks: (1) the lithofacies of volcanic deposits filling recent maars and diatremes, as well as their eroded counterparts; (2) the paleogeographic environment of eruptions, and the evolution of the deposits filling craters created by the phreatomagmatic or magmatic eruptions. It is proposed that researchers who intend to measure erosion rates accurately should use the authors’ model (fig. 10). This model takes into account three essential factors : (1) the type of construction and volcanic products inside a maar, based on the analysis of pyroclastic lithofacies, (2) the process of emplacement of lava flows with respect to the topography surrounding the edifice and the morphology of the vent, which has itself been strongly influenced by the type of host rocks (i.e either soft or hard); and (3) post-eruption processes of compaction, diagenesis and deformation that are recorded in deposits filling the maars. Finally, the textures of the pyroclastic rocks surrounding the vent provide evidence for identifying host rocks. On the basis of this method, the edifice morphology, which is often hidden or eroded, is that of large and shallow maars, or deep ‘glass of champagne’-like maars, above relatively shallow or deep diatremes.

5Common processes of erosion which are acting today on the slopes of active volcanoes are best monitored at experimental field sites. A second group of Japanese researchers, Y. Ogawa, H. Daimaru and A. Shimuzu working at the Institute of Forest and Forestry products in Tsukuba, has measured overland flow and its sediment load on two experimental field sites located on the slopes of Unzen volcano, following the 1991-1995 eruptive episode. They have installed devices at a site that has been reseeded with vegetation, and at a second site which has been artificially covered by fine ash. The latter is comparable in size to that produced by the eruptions.  In order better to understand how and why the particles of the pyroclastic-flow deposits are less remobilised with time once the eruption has ceased, the measurements have been undertaken during three years under both ‘normal’ wet conditions and under heavy rains. The authors demonstrate how two critical factors play a role in decreasing overland flow and its sediment load. The first parameter is the subsurface permeability, which decreases whilst the sieved ash migrates from the top surface towards the subsurface of the field site, because the removal of the thin sieved ash has re-established the infiltration capacity of the bare soil. The second parameter is vegetation cover growing on the seeded field site, which has decreased overland flow and, so, stabilised the loose pyroclastic-flow deposits. The authors claim that the S/Q ratio (overland flow/sediment load) directly reflects the extent to which the soil surface of a field site is stable or unstable: this S/Q ratio depending on the degree of stabilisation of the soil at the surface of pyroclastic-flow deposits.

6Comparisons between a studied volcanic landform, as inferred from in situ measurements, and its model, may be achieved by using numerical tools.  As algorithms become more powerful, mathematics and computer-based techniques, such as numerical codes, can assist the reconstruction of the evolution of volcanic edifices, which encompass a variety of erosion processes acting on cones and domes. Using an algorithm written with a C++ program, Jean-François Parrot working at the National Autonomous University of Mexico, determines key parameters that help to reconstruct the initial shape of a volcanic cone and to unravel its subsequent evolution due to erosion, catastrophic events or to human impact. J.-F. Parrot selects eight essential parameters, which are computed on the basis of an accurate DEM. Using the algorithm C++, he further selects two critical parameters, namely the elevation of the baseline of a cone and the coordinates of a spot taken as the crater centre. The computation based on a DEM creates a landform, which corresponds to that of the volcano prior to erosion and, having computed the initial shape of the edifice, the author is able to measure the volume of eroded material from the cone. He applies this method to the case study of Jocotitlán volcano in central Mexico, which has been eroded by a large landslide, and to two edifices of the area of Chichinautzín near Mexico City. This quantitative tool enables geomorphologists to validate existing models of the evolution of monogenetic edifices or even composite cones, to measure their relative age and to determine crucial parameters that should be measured in the field and scaled up on analogous models.

7Direct hazards and risks are induced by eruptive and seismic activity, whereas indirect hazard and risks stem from a combination of geological and geomorphological processes on potentially active volcanoes. The traditional approach in risk assessment consists in considering all geological processes that can trigger hazards and risk in a densely populated area on active unstable volcanoes located in seismic zones. Pursuing a long-lasting British and Portuguese joint research on the Azores islands, N. Wallenstein, A. Duncan, D.K. Chester and R. Marques describe hazards and risk indirectly linked to the Fogo volcano on the Saõ Miguel island. These hazards are termed indirect because they are triggered when a volcano is quiet. Driven by seismic, hydrothermal, gravity, and hydrological mechanisms, earthquakes, fumaroles, landslides and mudflows occur on that volcanic island due to its recurrent instability. The regional seismic activity represents the most severe hazard, as intensities exceeding IX on the MSK scale have been recorded between the XVIth and XXth centuries. Toxic gas, particularly CO2, stem from several active hydrothermal areas in the island.  Landslides and flash floods can be triggered by earthquakes, volcanic eruptions and by heavy rainstorms as well. A catastrophic mudflow occurred in 1522 due to the IX-intensity earthquake. The authors have mapped the extent of these phenomena and have assessed potential damage, which could occur during mass movements and debris flows in the area hit by the 1522 disaster (Villa Franco do Campo), where 11 000 inhabitants now live. Thus, the indirect hazards are as significant in risk assessment as direct hazards during eruptions and both ought to be the goals of hazard and risk mapping on active volcanoes.

8Another research method in hazard and risk assessment consists of using field survey and GIS techniques to asses risk induced by volcanic flows from a potentially active volcano, on the basis of recognition of deposits preserved in its radial valleys and on its piedmont. In a collective Ecuadorian, American, and Swiss-French endeavour, S. Ettinger, P. Mothes, R. Paris and S. Schilling have studied the distribution and characteristics of the lahar deposits attributed to the 1877 catastrophe triggered by the Cotopaxi volcano in Ecuador. They have also assessed the vulnerability of housing and populations along the valley of the River Jatunyacu to the East. They have superimposed these data in a GIS and used the semi-empirical simulation code ‘LAHARZ’, which allows them roughly to delineate areas prone to floods, if a large magnitude lahar comparable to the 1877 event were to be triggered in this region. These voluminous flows, albeit scarce, are triggered by a combination of factors: interactions between eruptive activity and the snowpack and ice cap, loose volcanoclastic debris, a valley morphology that favours a rapid transport of debris from the Sierra towards the Amazonian lowland. However, the Editor stresses the fact that the present-day ice cap is much smaller than that of 1877 and that semi-empirical simulations are based on mapping deposits assigned to the 1877 eruption, whose distribution and chronology are not assessed with accuracy.

9These articles illustrate some of the present-day research trends in the field of volcanic geomorphology. In particular they focus on: (1) the assessment of topographic control by steep radial valleys around a volcano and how these help to understand the emplacement of voluminous mass movements, which are able to rapidly erode a large part of an edifice and reach excessively long runout distances; (2) the caution that is called for upon when geomorphologists use volcanic deposits, such as lava flows, to reconstruct paleo-topography, because the pyroclastic lithofacies and the paleogeographical evolution of the local environment where the crater has been shaped must be also taken into account; (3) that monitoring techniques and measures of processes of erosion acting on slopes on pyroclastic deposits under a variety of settings, including sites artificially seeded or artificially covered by ash, make experiments closer to reality; a sedimentary budget should be the goal for measurements on annual basis of processes of denudation that are undertaken in field sites on loose pyroclastic deposits; (4) the identification and selection of a small number of critical parameters that aim at calibrating and scaling up analogue and numerical models that become more relevant to in situ measured edifices; (5) in addition to the traditional method of risk assessment on active and unstable edifices, geomorphologists must consider indirect hazards, which depend on the geologic and geomorphic setting of the volcano; (6) the calibration of numerical codes for simulating volcanic flows requires the introduction of the rheological characteristics of flows as critical parameters in numerical programs; and (7) in addition to hazard assessment, a variety of factors and the complex system of vulnerability should be the basis for risk assessment. The final objective of geomorphologists working on and around densely populated volcanoes consists in helping decision makers at times of crisis, and improving awareness of threatened populations and of the civil authorities in charge of emergency management.

10We thank 18 colleagues from many countries for reviewing and improving the selected papers, as well as the editorial committee and three collaborators, who have cleaned the English grammar of this special issue. I am particularly grateful to Dr. D. Chester, A. Duncan and C. Ollier who, over the past 10 years, have helped me so much in editing four special issues devoted to ‘volcanic geomorphology’.

Haut de page

Pour citer cet article

Référence papier

Jean-Claude Thouret, David Chester, Angus Duncan et Cliff Ollier, « Volcanic landforms, processes and hazards on volcanoes », Géomorphologie : relief, processus, environnement, vol. 13 - n° 3 | 2007, 211-215.

Référence électronique

Jean-Claude Thouret, David Chester, Angus Duncan et Cliff Ollier, « Volcanic landforms, processes and hazards on volcanoes », Géomorphologie : relief, processus, environnement [En ligne], vol. 13 - n° 3 | 2007, mis en ligne le 01 octobre 2009, consulté le 29 mars 2017. URL :

Haut de page


Jean-Claude Thouret

Articles du même auteur

David Chester

Articles du même auteur

Angus Duncan

Articles du même auteur

Cliff Ollier

Articles du même auteur

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie