Physical‐scale modelling of jökulhlaups (glacial outburst floods) with contrasting hydrograph shapes

Abstract

AbstractJökulhlaups (glacial outburst floods) are hazards in many glaciated environments. Jökulhlaups generated by different trigger mechanisms have different hydrograph shapes. Field evidence indicates that jökulhlaups with different hydrograph shapes produce different impacts. However, little attention has been paid to understanding the controls on jökulhlaups with differing hydrograph shapes and their subsequent impacts. Historically, flume studies have been used to investigate fluvial processes under controlled laboratory conditions. Nevertheless, few flume studies have replicated scaled hydrographs, and none have attempted to model field jökulhlaups. This study documents the technical procedure by which field prototype jökulhlaups were replicated using physical‐scale modelling techniques: an ‘exponentially rising’ jökulhlaup characterized by a prolonged rising stage and a rapid falling stage, and a ‘linearly rising’ jökulhlaup characterized by a rapid rising stage and a prolonged falling stage. The parameters of water discharge, time, discharge acceleration and deceleration rate, sediment size and sediment discharge were scaled to a 1:20 ratio.Better‐sorted sediment, more uniform downstream sediment thickness and more gradual aggradation and deposition rates resulted from the prolonged rising stage of the exponentially rising hydrograph. More time was available during the prolonged rising stage of the exponentially rising hydrograph for bedload sheets to migrate, and for sand ripples to develop. Rapid rising stage deposition during the linearly rising hydrograph produced thick proximal deposits, limited bedload sheet migration, massive sand sheets and structureless and poorly sorted sediments. During the prolonged falling of the linearly rising hydrograph, sufficient time was available to support sand ripple development.The study highlights that jökulhlaup hydrograph shape is important for controlling rates of sediment acquisition, transport, deposition, erosion and bedform development. The outcomes of these experimental investigations indicate that modelling of high magnitude floods is feasible in a laboratory flume. This technical approach has proved very beneficial for developing a greater understanding of flood impacts and the processes that operate during jökulhlaups. Copyright © 2006 John Wiley & Sons, Ltd.