Speleothem U-series constraints on scarp retreat rates and landscape evolution: an example from the Severn valley and Cotswold Hills gull-caves, UK

Abstract

Abstract:Modeling landscape evolution requires quantitative estimates of erosional processes. Dating 12 erosional landscape features such as escarpments is usually difficult because of the lack of datable deposits. 13Some escarpments and valley margins are associated with the formation of mass-movement caves, sometimes 14 known as 'gull' or 'crevice' caves, which are typically restricted to within 0.5 km of the valley margin or scarp 24Quantifying rates of landscape processes is an essential requirement for constructing, validating and 25 constraining increasingly sophisticated landscape evolution models (Pazzaglia, 2003; Tucker and 26 Hancock, 2010). With quantitative data, rates of landform development can be evaluated, enabling 27 the relative importance of geomorphological processes to be established and facilitating the 28 development of more realistic landscape evolution models. Moreover, quantification is extremely 29 important in predictive work, as it is required for estimating the impact of future global climate 30 change. Whilst some geomorphological processes can be easily quantified, such as the rate of valley 31 incision by dating river terrace sequences (Maddy, 1997; Maddy et al., 32 2000;Maddy et al., 2001), dating cave levels in carbonate terrains (Farrant et al., 1995; Palmer, 33 2007), or by dating other alluvial materials such as tufa (Banks et al., 2012), deducing the timing and 34 rates of other processes such as valley widening and escarpment ( or 'scarp') retreat is harder to 35 determine. However, both rates of valley incision and scarp retreat are required to understand how 36 valleys evolve. Do they develop by progressive incision and valley widening through fluvial channel 37 migration and concurrent hill slope retreat or is the gross relief generated 'in-situ' by the progressive 38 removal of the more erodible lithologies over multiple glacial-interglacial cycles? In the latter 39 scenario, valley width is influenced more by lithological heterogeneity and variable susceptibility to 40 periglacial weathering (Murton and Belshaw, 2011) fissures may contain speleothem deposits which can be precisely dated using uranium series 53 methods (Lenart and Pánek, 2013). As the gull-caves can only develop after a scarp has formed, the 54 basal age of the oldest speleothems within them provide a minimum age for cave inception and 55 hence scarp formation. Moreover, as the gull-caves only form close to the scarp edge, they can be 56 used to determine a chronology of scarp retreat. Taken together with rates of valley incision 57 determined from fluvial terraces, the spatial and temporal pattern of valley development and scarp 58 formation can be resolved and models of regional landscape evolution erected. 59The study area 60In this paper, we use the lower Severn valley and the Cotswold Hills in southern England (Figure 1) Triassic and Lower Jurassic (Lias Group) mudstones ( Figure 1) that occupy the core of this basin. The 77 eastern side is marked by the prominent escarpment of the Cotswol...

Investigation of suspected gulls in the Jurassic limestone strata of the Cotswold Hills, Gloucestershire, England using electrical resistivity tomography

“…Other known occurrences of gulls in the south Cotswolds have little or no surface expression, either because they have a cemented capping, or because they do not penetrate to the surface above (Self, 2008;Hawkins, 2013;Farrant et al, 2015). The latter are mostly known from caving explorations or mine-workings that intercept them, so can be accurately located at depth in cave or mine surveys, and sometimes form intricate grid sets.…”

“…Major phases of uplift in this region occurred in the latest Jurassic to early Cretaceous (Chadwick, 1986), and Palaeogene (King, 2006), and thus the pervasive networks of joints in the Inferior Oolite strata are inferred to predate the development of the Cotswold escarpment in the Middle Pleistocene (Farrant et al, 2015). Following this, the sequence of events that resulted in formation of the gull structures here is thought to be as follows:…”

“…However the relative and absolute timing, the causes and processes, and the resulting structures and deposits are not fully documented or understood. Current research into the age and chronology of the formation of open gulls along the Cotswold escarpment (Farrant et al, 2015) indicates that here they are over 350 000 years old, which has profound implications for any estimates of the age of the escarpment and its landslide phenomena, and the timing and processes of excavation of the Severn Valley. Gulls present potential geohazards -specifically, their disposition, form in 3D and the nature and stability of any fill have engineering implications.…”

“…This would include limestone rubble and soil falling from the walls (colluvium) plus the gradual weathering of limestone debris and in situ development of topsoil. If recent proposals by (Farrant et al, 2015) are accepted, that gull formation along the Cotswold escarpment dates from the Middle Pleistocene and that the current summit surface of the Cotswolds is, or is close to, an exhumed prelate Cretaceous erosion surface, this would preclude the inclusion of material derived from younger Jurassic bedrock that was removed from the Cleeve Hill area in the early Cretaceous. In addition, given the topographic position on a hill top, the potential for later Quaternary input of further material would have been limited, apart from some hillwash and wind-blown fine sediment.…”

Chapter 6 Material properties and geohazards

“…Cambered strata, gulls and valley bulges have been observed at several localities throughout upland Britain and in areas of lower relief. Notable examples have been documented in the Namurian strata of the Pennines (Donnelly 2008, Jones & Weaver 1975, Hill 1949, Lapworth 1911, Miller 1887, Morton 1949, Watts 1905, the Coal Measures of South Wales (Donnelly 2005, Donnelly et al 2000 and Yorkshire (Shotton & Wilcockson 1950), Jurassic strata in North Yorkshire (Cooper 1980), Northamptonshire (Hollingworth et al 1944, Horswill & Horton 1976, Vaughan 1976, along the extensive Cotswold escarpment (Ackermann & Cave 1967, Chandler et al 1976, Farrant et al 2014, Self & Farrant 2013, Worcestershire, the East Midlands (Horswill & Horton 1976) and Dorset (Arkell 1947) and Cretaceous strata in the Weald of Kent (Bristow & Bazley 1972). .56 shows some of the recorded occurrences of cambering in the literature (Donnelly et al 2002, Hawkins & Privett 1981, Higginbottom & Fookes 1970, Hill 1949, Horswill & Horton 1976, Hutchinson 1991, Kellaway & Taylor 1968, Lang 1914, Morton 1973, Sandeman 1918, Self 1985, in the British Geological Survey's National Landslide Database and in the PhD thesis of Parks (1991b).…”