Geomorphological and sedimentological evidence for variations in Holocene flooding in Welsh river catchments

https://doi.org/10.1016/j.gloplacha.2009.11.010Get rights and content

Abstract

Valley floor geomorphology, sedimentology and radiocarbon dating have been used to reconstruct variations in flooding during the Holocene at six sites within the upper Severn, Teifi and Dee catchments in Wales, UK. High-magnitude palaeoflood deposits, for which ages were estimated, were identified at three floodbasin sites in the upper Severn and Teifi catchments. At the remaining sites fluvial discontinuities, indicative of episodes of increased flooding, were identified. Data from these six sites, and six additional sites studied previously, indicate that seven probably centennial length episodes of major flooding have occurred in Wales since c. 7000 cal BP. These flooding episodes occur at similar times to peaks in flood frequency within Great Britain as a whole.

Introduction

Long records of riverine flooding are required to establish the magnitudes and frequencies of high-magnitude flood events for floodplain land-use planning, flood warning and flood defence design. In addition, long flood records are required to elucidate temporal variations in flood frequency and to determine whether flood-generating mechanisms can be regarded as temporally constant, a key assumption underpinning conventional flood frequency analysis (Baker, 1991, Knox, 2000, Jain and Lall, 2001). Gauged flood records, which in Wales generally do not exceed 50 years in length, are too short for these purposes, because assessment of the frequency of high-magnitude floods would involve considerable extrapolation beyond the available data range (Baker, 1989, Baker, 2003, Benito et al., 2004). This problem is replicated in other countries (Table 1). In certain circumstances, historical flood records can be used to supplement gauged data and hence extend flood records (e.g. Mudelsee et al., 2003, Kondrashov et al., 2005, Macdonald et al., 2006). However, historical records are frequently unavailable (Rumsby, 1991); this is a particular problem in the sparsely populated rural Welsh catchments investigated in this study. In such catchments, alternative approaches, for example, palaeoflood hydrology, need to be used or developed to provide information to supplement gauged flood records (Baker, 1989, Baker, 1991, Baker, 2003).

Palaeoflood hydrology is the study of floods which occurred prior to the gauging, or historical recording, of flood events within the catchment that can be reconstructed from geomorphological, sedimentological or botanical evidence (Baker, 2003). Depending on the type of evidence available, palaeoflood records may extend over centuries or millennia (Wohl, 1992, Ely et al., 1993). Various methods of palaeoflood reconstruction have been developed for use in different types of channel (bedrock or alluvial), climatic setting (e.g. semi-arid or humid) (Baker, 2003, Knox, 2003) and catchment size (Knox and Daniels, 2002). Palaeoflood magnitudes and frequencies are reconstructed separately using different sedimentary components and characteristics (e.g. Kochel and Baker, 1982, Kochel et al., 1982). In certain settings, such as alluvial stratigraphic sequences, palaeoflood frequency information is more readily extracted than information on palaeoflood magnitude (Knox and Daniels, 2002).

Methods of palaeoflood reconstruction applicable to alluvial rivers in temperate, maritime climates, such as those in Wales, are at present considerably less well developed than those applied to bedrock rivers in arid and semi-arid areas. In Britain the relatively few attempts at palaeoflood reconstruction generally fall into two categories: first, a number of palaeoflood histories, that is, chronologies of major flood events, have been constructed for individual catchments which span the period from the Little Ice Age to the present; these studies have frequently involved application of lichenometry in multiple small catchments in upland regions (Rumsby, 1991, Macklin et al., 1992a, Macklin et al., 1992b, Merrett and Macklin, 1999, Werritty et al., 2006, Macklin and Rumsby, 2007). Second, records of variation in the frequency of flooding in Britain as a whole have been constructed for the whole Holocene using analysis of a database of radiocarbon-dated alluvial units (Macklin and Lewin, 2003, Macklin et al., 2005, Johnstone et al., 2006). Thus far no attempts to reconstruct flood histories using geomorphological and sedimentological evidence have been made in Welsh alluvial rivers. This study aims to address this data gap by using geomorphological and sedimentological evidence from Welsh floodplains to reconstruct variations in flood occurrence in Wales during the Holocene and to assess the degree to which the reconstruction of palaeoflood histories is possible in Wales using this type of data.

Section snippets

Study sites

Wales is a particularly suitable region for conducting a study of this nature for three reasons. First, in spite of Wales' small size and predominantly upland character, there is a high degree of geomorphological diversity in Welsh river systems. Second, in piedmont and lowland reaches, Welsh river systems frequently have well-developed alluvial valley floors which have high potential for alluvial archive preservation. Third, previous work (Gittins, 2004) has shown that Welsh river systems are

Methods

Geomorphological mapping, sediment coring and radiocarbon dating were used to reconstruct Holocene flood histories at the six sites. A combination of three approaches was used to produce a geomorphological map of each site. At Caersws, the map produced by Maas et al. (2001) and later work by Johnstone (unpublished) were combined to produce a single geomorphological map. At the Roundabout site, mapping was undertaken by a walkover survey. In both cases available topographic data and information

Results

The results of the geomorphological and sedimentological investigations at the six study sites are presented in the form of a reconstruction of the alluvial history at each site, followed by the evidence for the occurrence of high-magnitude palaeofloods, where it exists. The principles used to derive palaeoflood records from these alluvial histories, and presentation of these palaeoflood histories, follow in Section 5.

Episodes of flooding in Welsh rivers

In order to compare the sequence and timing of events at each of the six sites used in this study, with the five previous studies in the Severn, Vyrnwy, Dyfi and Rheidol catchments (Macklin and Lewin, 1986, Taylor and Lewin, 1996, Taylor and Lewin, 1997, Macklin et al., 2002, Johnstone, 2004), an age–height diagram (Passmore and Macklin, 1997) was produced (Fig. 8(a)). The ages and elevations of individual radiocarbon-dated samples at each site have been plotted together with a simplified

Conclusion

The geomorphological and sedimentological data from six sites in the upper Severn, Teifi and Dee catchments, coupled with data presented in previous studies of six additional sites in the Severn, Vyrnwy, Dyfi and Rheidol catchments indicate that there have been seven episodes of increased flooding within Welsh river catchments since c. 7000 cal BP. The ages of these seven episodes of increased flooding have been constrained using dated fluvial discontinuities and the estimated ages of major

Acknowledgements

This study formed part of a project jointly funded by the Welsh Assembly Government, Environment Agency Wales, Countryside Council for Wales, British Geological Survey and Aberystwyth University and was also supported by an Aberystwyth University studentship to AFJ. LiDAR data for the Dee and Teifi catchments were supplied by the Environment Agency. Thanks are due to Gez Foster, Matt Rowberry, Catherine Swain, Eric Johnstone, Ian Dennis, Marco van de Wiel, Mike Marshall, Paul Coombes and Huw

References (51)

  • P.A. Brewer et al.

    A Geomorphological Re-appraisal of the River Severn Roundabout GCR site

  • C. Bronk Ramsey

    Radiocarbon calibration and analysis of stratigraphy: the OxCal program

    Radiocarbon

    (1995)
  • C. Bronk Ramsey

    Development of the radiocarbon calibration program OxCal

    Radiocarbon

    (2001)
  • L.L. Ely et al.

    A 5000-year record of extreme floods and climate change in the southwestern United States

    Science

    (1993)
  • Gittins, S., 2004. Changes in Sediment Storage in Welsh Rivers, 1890–2002. Unpublished PhD Thesis, University of Wales,...
  • Global Runoff Data Centre, 2008. Tabular Summary by Country per WMO Region (Status: December 2007)....
  • A.M. Gurnell et al.

    Channel planform change on the River Dee meanders, 1876–1992

    Regulated Rivers: Research and Management

    (1994)
  • R.D. Hey

    Upper River Severn between Dolwen and Penstrowed, Powys

  • G. Higgs

    River Severn between Welshpool and the confluence of the Vyrnwy and the Severn

  • S. Jain et al.

    Floods in a changing climate: does the past represent the future?

    Water Resources Research

    (2001)
  • Johnstone, E., 2004. River Response to Late Quaternary Environmental Change: the Dyfi Catchment, mid-Wales. Unpublished...
  • Jones, A.F., 2007. Reconstructing Historical and Holocene Flooding in Welsh River Catchments. Unpublished PhD Thesis,...
  • A.F. Jones et al.

    High-resolution geomorphological mapping of river valley environments using airborne LiDAR data

    Earth Surface Processes and Landforms

    (2007)
  • J.C. Knox

    Responses of river systems to Holocene climates

  • J.C. Knox

    North American palaeofloods and future floods: responses to climatic change

  • Cited by (18)

    • Paleoflood evidence for an upper limit of the maximum flood magnitudes along the Gold Gorge, the upper Hanjiang River, China

      2022, Catena
      Citation Excerpt :

      Long-term flood information (timing, magnitude and frequency) is of crucial importance for understanding climate change impacts, assessing water resources and mitigating flood risks (Baker, 2008; Greenbaum et al., 2014a; Benito and Díez-Herrero, 2015; Fuller et al., 2019; Benito et al., 2020). Considering that available gauged or historical records are generally too short, paleoflood hydrology has proven to be useful in extending flood datasets and improving our understanding of rare, large magnitude floods over long-time scales (Kochel and Baker, 1982; Baker, 1987; Jones et al., 2010; Benito and O′Connor, 2013; Saynor et al., 2020). Paleoflood slackwater deposits (SWDs) are the most accurate paleostage indicator (PSI) for reconstructing paleoflood hydrological characteristics (e.g., Kale et al., 2000; Benito et al., 2008; Sheffer et al., 2008; Huang et al., 2013; Lam et al., 2017a; Baker et al., 2019; Liu et al., 2020; Thorndycraft et al., 2005).

    • Investigating extreme flood response to Holocene palaeoclimate in the Chinese monsoonal zone: A palaeoflood case study from the Hanjiang River

      2015, Geomorphology
      Citation Excerpt :

      The rapidly growing number of palaeoflood hydrology studies in a global context will help to interpret Holocene climatic variations (Baker, 2006). Through field-based analyses such as stratigraphic sedimentary and geomorphologic and hydrologic investigations, Holocene palaeoflood hydrological research can provide long-term data for the chronology and magnitude of floods (Thorndycraft and Benito, 2006; Jones et al., 2010). The sedimentary record of high-stage flood deposits is accumulated in backwater areas (slackwater) ideally in the bedrock canyons (Benito et al., 2003a,b).

    • The rivers of civilization

      2015, Quaternary Science Reviews
    View all citing articles on Scopus
    View full text