Abstract
Rain-triggered lahars (RTLs) are a significant and often persistent secondary volcanic hazard at many volcanoes around the world. Rainfall on unconsolidated volcaniclastic material is the primary initiation mechanism of RTLs: the resultant flows have the potential for large runout distances (>100km) and present a substantial hazard to downstream infrastructure and communities. RTLs are frequently anticipated in the aftermath of eruptions, but the pattern, timing and scale of lahars varies on an eruption-by-eruption and even catchment-by-catchment basis. This variability is driven by a set of local factors including the grain size distribution, thickness, stratigraphy and spatial distribution of source material in addition to topography, vegetation coverage and rainfall conditions. These factors are often qualitatively discussed in RTL studies based on post-eruption lahar observations or instrumental detections. Conversely, this study aims to move towards a quantitative assessment of RTL hazard in order to facilitate RTL predictions and forecasts based on constrained rainfall, grain size distribution and isopach data. Calibrated simulated rainfall and laboratory-constructed tephra beds are used within a repeatable experimental set-up to isolate the effects of individual parameters and to examine runoff and infiltration processes from analogous RTL source conditions.Laboratory experiments show that increased antecedent rainfall and finer-grained surface tephra individually increase runoff rates and decrease runoff lag times, while a combination of these factors produces a compound effect. These impacts are driven by increased residual moisture content and decreased permeability due to surface sealing, and have previously been inferred from downstream observations of lahars but not identified at source. Water and sediment transport mechanisms differ based on surface grain size distribution: a fine-grained surface layer displayed airborne remobilisation, accretionary pellet formation, rapid surface sealing and infiltration-excess overland flow generation whilst a coarse surface layer demonstrated exclusively rainsplash-driven particle detachment throughout the rainfall simulations. This experimental protocol has the potential to quantitatively examine the effects of a variety of individual parameters in RTL initiation under controlled conditions.
Highlights
Rainfall on unconsolidated volcaniclastic material, typically pyroclastic density current (PDC) and/or ash-fall deposits, is the primary initiation mechanism of secondary, rain-triggered lahars (RTLs)
Owing to the coarse nature of the tephra, such surface sealing did not occur during simulation Ash Bed 1 (AB1); infiltration rates were sufficiently high that infiltration-excess overland flow did not occur and raindrop impact resulted in rainsplash-driven particle detachment throughout the simulations
Calibrated rainfall simulations have illustrated that both finer-grained surface material and increased antecedent rainfall increase runoff rates and decrease runoff lag time from laboratory-constructed tephra beds
Summary
Rainfall on unconsolidated volcaniclastic material, typically pyroclastic density current (PDC) and/or ash-fall deposits, is the primary initiation mechanism of secondary, rain-triggered lahars (RTLs). These flows pose a significant hazard to downstream infrastructure, with impacts ranging from damage to building contents via flow inundation, to complete destruction and burial of structures (Jenkins et al, 2015). Numerous temporally and spatially variable factors contribute towards these lahar initiation processes including grain size distribution (Pierson et al, 2013), thickness (Janda et al, 1996; Scott et al, 1996; Manville et al, 2000), extent of compaction (Manville et al, 2000) and volatile content (Waldron, 1967) of pyroclastic material; vegetation cover (Yamakoshi and Suwa, 2000; Barclay et al, 2007; Ogawa et al, 2007; Alexander et al, 2010) and type (Capra et al, 2010); rainfall intensity and duration (Rodolfo and Arguden, 1991; Lavigne et al, 2000b; Van Westen and Daag, 2005; Hikida et al, 2007; Okano et al, 2012); slope angle (Pierson et al, 2013) and antecedent rainfall (Lavigne et al, 2000b; Barclay et al, 2007; Okano et al, 2012; Jones et al, 2015) (Fig. 2)
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call