Anne Duperret

Coastal chalk cliff instability in NW France: role of lithology, fracture pattern and rainfall

Abstract Coastal retreat has been studied along 120km of French Channel chalk coast from Upper Normandy to Picardy. During the investigation period, 1998–2001, 55 significant collapses were recorded. Of these 5.5% were very large-scale, 34.5% large-scale, 34.5% medium-scale and 25.5% small-scale collapses. Observations indicate that the larger the collapse size the greater the coastal cliff retreat. Four types of cliff failure were observed: (1) vertical failures in homogeneous chalk units; (2) sliding failures where two superimposed chalk units were present; (3) wedge and plane failures mainly recognized in the UK in formations with stratabound fractures; (4) complex failures in cliffs with more than one style of fracturing. Rainfall in relation to the timing of cliff collapse indicates two periods that trigger a collapse. The first occurs about one month after heavy rainfall within poorly fractured chalk and the second occurs when a dry period is interrupted by sharp rainfall in cliffs with major karst features (pipes etc). Medium to small-scale cliff collapses were, in some cases, caused by marine erosion at the base of the cliff creating a notch. A key factor controlling the type of collapse is the lithostratigraphic unit, while the extent of the collapse scar may be controlled by fracture type.

Coastal cliff geohazards in weak rock: the UK Chalk cliffs of Sussex

Abstract Geohazards related to chalk coastal cliffs from Eastbourne to Brighton, Sussex are described. An eight-fold hazard classification is introduced that recognizes the influence of chalk lithology, overlying sediments and weathering processes on location, magnitude and frequency of cliff collapses. Parts of the coast are characterized by cliffs of predominantly a single chalk formation (e.g. Seven Sisters) and other sections are more complex containing several Chalk formations (Beachy Head). Rock properties (intact dry density or porosity) and mass structure vary with each formation and control cliff failure mechanisms and scales of failures. The Holywell Nodular Chalk, New Pit Chalk and Newhaven Chalk formations are characterized by steeply inclined conjugate sets of joints which lead to predominantly plane and wedge failures. However, the dihedral angle of the shears, the fracture roughness and fill is different in each of these formations leading to different rock mass shear strengths. In contrast the Seaford and Culver Chalk formations are characterized by low-density chalks with predominantly clean, vertical joint sets, more closely spaced than in the other formations. Cliff failure types range from simple joint controlled conventional plane and wedge failures to complex cliff collapses and major rock falls (partial flow-slides) involving material failure as well as interaction with discontinuities. Other hazards, related to sediments capping the Chalk cliffs, include mud-slides and sandstone collapses at Newhaven, and progressive failure of Quaternary Head and other valley-fill deposits. Weathering, including the concentration of groundwater flow down dissolution pipes and primary discontinuities, is a major factor on rate and location of cliff collapses. A particular feature of the Chalk cliffs is the influence of folding on cliff stability, especially at Beachy Head, Seaford Head and Newhaven. A new classification for cliff collapses and a new scale of magnitude for collapses are introduced and used to identify, semi-quantify and map the different hazards. Climate (and climate change) and marine erosion affect the rate of development of cliff collapse and cliff-line retreat. This was particularly evident during the wet winters of 1999–2000–2001 when the first major collapses along protected sections of coastline occurred (Peacehaven Cliffs protected by an undercliff wall; Black Rock Marina the Chalk cliffs and the Quaternary Head). It is the geology, however, that controls the location and scale of erosion and cliff failure.

Chalk physical properties and cliff instability

Abstract Physical properties such as porosity and intact dry density (IDD) are compared with strength testing in relation to the Chalk formations in the cliffs of the English Channel. Natural moisture contents are close to saturation moisture contents for chalks with intact dry densities above 1.70 Mg/m3. Below this IDD, the natural moisture contents show a much greater range and greater divergence from the saturation line. There is also an indication that certain types of chalk retain water at saturation level while others gain and lose water more readily. Strength tests (Point Load Index, Brazilian Crushing Strength and Uniaxial Compressive Strengths) show up to four times reductions in strength between dry (higher strength) and saturated (lower strength) samples. Absence of a strong correlation between density and strength is interpreted as resulting from either mineralogical differences in the samples and/or textural differences between different chalks. The variation in physical properties and strength in the different chalks forming the cliffs indicates the strong stratigraphical and sedimentological controls on mechanical performance of the material and mass in cliff failures.

Multiscale fracture analysis along the French chalk coastline for investigating erosion by cliff collapse

Abstract Coastal cliffs of Upper Normandy and Picardy are eroded by cliff collapses of various sizes. This paper presents a multi-scale analysis of the pre-existing fractures embedded within the Cretaceous chalk. About 20 representative sites equally spaced along the 120km long coastal section were analysed and compared to a continuous structural analysis of the coast derived from aerial photographs taken in 1986. Ancient collapses interpreted on the aerial photos were compared to the pre-existing fracture content. Regional faults, pre-1986 collapse location and fracture density are spatially correlated. However, recent collapses observed on the field between 1998 and 2001 did not systematically correlate to the pre-existing fracture occurrence and therefore, there is no clear link between recent collapse and the regional faults.