Floodplain evolution in the East Midlands, United Kingdom: the Lateglacial and Flandrian alluvial record from the Soar and Nene valleys

Abstract

This paper describes and interprets the floodplain stratigraphy of two low-energy rivers in the East Midlands, U.K. and goes on to present an appropriate model of floodplain evolution that may be applicable to other lowland rivers in temperate mid-latitudes. Work has been undertaken at the reach level to try and characterize the entire system. In general only laterally extensive exposures have been used, and detailed stratigraphic and microstratigraphic recording has been used to facilitate process interpretations. All the sites have been radiocarbon dated. The dataset consists of three sites in the Soar valley and five sites in the larger Nene valley. The most common sedimentary architecture is a basal gravel with occasional shallow channels, covered by a mixed loam (sand-silt-clay) of variable thickness with landsurfaces and evidence of bioturbation. This loam is frequently interrupted by palaeochannels, and finally all these sediments are buried by a massive silt-clay unit. Most of the palaeochannels were abandoned in the early Flandrian, or between ca. 3500 years BP and 2000 years BP. The superficial silty clay is diachronous largely because of the irregular topography of the floodplain onto which it was deposited; dates from non-palaeochannel sections range from ca. 3500 BP to ca. 2100 years BP. The floodplain stratigraphy has been profoundly influenced by soil development and sub-aerial processes, especially tree-throw, which has produced distinctive sedimentary structures. The loam unit is interpreted as a soil which developed inbetween both silting palaeochannels and active channels. The landsurfaces are both earlier and contemporaneous with the later phase of channel abandonment. There is some evidence, ca . 5000-4000 BP, of a rise in floodplain watertables. An evolutionary model is proposed which can account for the stratigraphic evidence reported here. It is based upon the development of an anastomosing (stable multiple-channel) system from an initial braided-river topography and its eventual conversion to a predominantly single-channel system due to floodplain and channel siltation. The reduction of channels is compensated for by a change in channel types and capacities while the floodplain aggrades; this is the stable-bed aggrading-banks (SBAB) model, which necessitates no changes in discharges. It is suggested that the sub-meandering or straight to sinuous nature of many lowland U.K. channels may be due to their evolution from an anastomosing pattern where the least meandering channels survived typically with a box-S shape planform and at the edges of the floodplain. The sites also show that the Lateglacial fluvial history of the two catchments seems to have been very different, with incision and subsequent aggradation occurring during the Younger Dryas in the Nene but not in the Soar. Given the proximity and similarity of the two catchments this suggests that relatively minor local factors may have been able to push some catchments across fluvial thresholds. In contrast the Flandrian history of the two rivers has been broadly similar, although there is evidence of greater lateral instability and floodplain reworking in the Soar which may be due to hydrogeological factors or a different landuse history. This work strongly suggests that new process-based interpretations of floodplain stratigraphy, and new models of floodplain evolution may be required before alluvial history can be easily related to the changing Lateglacial and Flandrian climate of lowland U.K.