Tuesday, 13 January 2026

Pleistocene Glacial and Periglacial Features in Somerset: Limits of Ice Advance and Local Dynamics in the Levels and Mendips

Map of Somerset Levels

The Pleistocene history of Somerset mirrors that of north Devon in many respects, characterised by peripheral interactions with the Irish Sea Ice Sheet rather than extensive inland glaciation. The Somerset Levels—a low-lying coastal plain prone to flooding—and the adjacent Mendip Hills exhibit a record of ice-marginal deposits, periglacial processes, and interglacial marine incursions, but without evidence of widespread ice override or floating sea ice penetrating deep into the interior. This reflects the region's position at the southern fringe of the British-Irish Ice Sheet, where glacial advances during stages such as the Anglian (Marine Isotope Stage 12, around 478,000–424,000 years ago) or Wolstonian were constrained by topography and climate, leading to localised sedimentation from meltwater and periglacial activity rather than broad ice-sheet coverage.

In the Somerset Levels, glacial deposits indicate limited ingress by the Irish Sea lobe, primarily in the northern areas around Clevedon and Kenn. Here, the Kenn Formation includes tills (diamictons with striated boulders), glaciofluvial gravels, and outwash sands, exposed in col-gullies like Court Hill and Nightingale Valley. These sequences, featuring erratic clasts such as Chalk flints, Greensand chert, and Cretaceous microfossils derived from the west, suggest an ice-marginal environment with proglacial outwash and possible flow tills, but no subglacial deformation indicative of extensive advance. (Note: While major Cretaceous chalk outcrops are concentrated in eastern and southeastern England, such as the North and South Downs, the chalk flints and related erratics in these Somerset deposits were sourced from exposures in the Irish Sea Basin, including the Antrim Chalk in Northern Ireland and possibly seabed sediments or outcrops in northwest Wales. These were entrained by the Irish Sea Ice Sheet and transported southwards and eastwards into the Bristol Channel, explaining their "western" provenance despite the material's geological association with eastern formations.) The deposits are interpreted as dating to a pre-Anglian or Wolstonian event, with the ice front impinging from the Bristol Channel but not progressing far southward; for instance, southern sites like Greylake show only rare glacigenic elements in basal diamictons, and erratic-free gravels dominate further inland. Interglacial marine and estuarine units, such as the Burtle Formation (shelly sands and gravels up to 5–10 metres OD, correlated with Stages 9, 7, and 5e) and Yew Tree Formation (estuarine silts with marine molluscs like Macoma balthica), overlie or interdigitate with these, highlighting episodic sea-level rises during warmer periods without glacial involvement. Periglacial features in the Levels include aeolian coversands, cryoturbated breccias, and colluvial silts, as seen at Holly Lane near Clevedon, where niveo-aeolian loams and frost-shattered limestones reflect cold, dry Devensian conditions (Stages 4–2) with tundra-like vegetation indicated by molluscs such as Pupilla muscorum.

The Mendip Hills, rising to around 300 metres OD, experienced predominantly periglacial activity, with no direct evidence of ice-sheet coverage or independent ice caps akin to those on Dartmoor. Slope deposits and alluvial fans dominate, such as at Bourne and Wookey Station, where fan gravels with cryoturbation, involutions, and cold-stage molluscs (e.g., Pupilla muscorum, Carychium arenaria) indicate mass movement and braided stream deposition under periglacial regimes. These are interspersed with palaeosols reflecting brief interstadials, and aeolian sands derived from distant sources (e.g., Tertiary deposits in Devon) point to wind-blown transport during arid cold phases. At Brean Down on the Mendip fringe, a sequence of rockfall breccias, aeolian silts, and palaeosols spans the Devensian, with fossil mammals (reindeer, arctic fox) and molluscs suggesting steppe-tundra landscapes and minor climatic ameliorations, possibly correlating with Stage 3 interstadials. Glacial influence is marginal at best; enigmatic glaciofluvial gravels at Bleadon Hill contain local Carboniferous Limestone clasts, potentially linked to proglacial lake shores or outwash from Irish Sea ice nearby, but without far-travelled erratics confirming override. Karstic fissures in the Mendips, such as at Bathampton Down, contain recycled erratics (flint, chert) in infill gravels, suggesting periglacial reworking rather than direct glacial emplacement.

Critically, these deposits do not indicate high sea levels facilitating floating ice or marine incursions deep into Somerset during glacial maxima. As in Devon, Pleistocene cold stages coincided with global sea-level drops exceeding 100 metres, exposing the Bristol Channel and facilitating terrestrial periglacial processes rather than glaciomarine environments. Deposits lack marine microfossils in inland contexts, and erratics at higher elevations (e.g., up to 20–30 metres OD in terraces) are attributed to periglacial solifluction or fluvial reworking from earlier events, not to ice-rafting during elevated sea stands. The notion of inflowing sea ice is unsupported, as coastal erratics (e.g., at Weston-in-Gordano) are confined to low elevations and tied to interstadial highstands, not peak glaciation.

In essence, the Somerset Levels and Mendip Hills exemplify a constrained Pleistocene glaciation similar to north Devon's: marginal impingement by the Irish Sea lobe in the lowlands, creating limited outwash and tills without further inland flow, complemented by pervasive periglacial weathering on the hills. This peripheral dynamic, driven by local topography and without coalescence with larger ice sheets, accounts for the observed features through meltwater, solifluction, and aeolian processes, aligning with reconstructions of the British-Irish Ice Sheet's southern limits. Persistent suggestions that the Irish Sea Ice Sheet overrode the Mendips or extended further east across the chalk escarpment to deliver erratics to Salisbury Plain are refuted by multiple lines of evidence. Firstly, there is a complete absence of glacial drift, tills, or subglacial features on the Mendip plateau or its interior; instead, the hills show only periglacial slope deposits, cryoturbation, and aeolian sands, with any recycled erratics in karst fissures attributable to solifluction rather than direct ice emplacement. The ice lobe's contact was limited to the eastern margin of the Mendips at the Somerset Levels, with no evidence of overriding the hills, as confirmed by the lack of striations or erratics on higher ground. Secondly, the easternmost glacial limit in Somerset is marked by scattered deposits in the Bridgwater-Glastonbury area, well short of the chalk escarpment in Wiltshire; beyond this, sediments transition to erratic-free fluvial and periglacial materials. On Salisbury Plain itself, there are no glacial deposits, moraines, or far-travelled erratics from Irish Sea sources that would be expected from an overriding ice sheet capable of transporting large lithologies; Stonehenge bluestones are instead explained by human transport. Geochemical provenancing matches the bluestones to specific Welsh quarries without requiring glacial intervention, and the absence of intermediate glacial drifts between Preseli and Salisbury Plain undermines long-distance ice transport theories. Overall, the southern limits of the Irish Sea Ice Sheet are firmly established in the Bristol Channel and coastal Somerset, with no stratigraphic, geomorphological, or sedimentological support for extensions over the Mendips or to Salisbury Plain. Further optically stimulated luminescence (OSL) dating and sediment provenance studies could refine chronologies, but the consensus highlights distinctly localised phenomena in this extra-glacial landscape.

References

Posted by at

No comments:

Post a Comment

Comments welcome on fresh posts - you just need a Google account to do so.

Subscribe to: Post Comments (Atom)