Extracting palaeoflood data from coarse-grained Pleistocene river terrace archives: an example from SE Spain

Field based palaeoflood event reconstruction has the potential to contribute to the development of our understanding of longterm landscape evolution. However the reconstruction of past flow event histories (magnitude and frequency) over long-term (Quaternary) timescales is fraught with difficulties. Here we make a preliminary exploration of some of the practicalities of flood reconstruction from fluvial terrace archives using commonly available sedimentological and geomorphological observations from a field perspective. We utilize Manning and palaeostage indicators to reconstruct historic events that can be used as benchmarks for a lesser used competence based approach (Clarke 1996), which is applied to coarse-grained strath terrace deposits. We evaluate the results against gauged records for extreme and catastrophic events that affected the same region in 1973 and 2012. The findings suggest that the competence approach is most effectively applied to terrace deposits if the channel geometry is taken into account when sampling both in cross section and in longitudinal section and calibrated against the sedimentology for palaeo-flow depth. Problems can arise where constrictive channel geometries allow boulder jams to develop, acting as sediment traps for the coarsest material and leading to downstream ‘boulder starvation’. Useful sites to target for palaeoflood reconstruction, therefore, would be upstream of such constrictive reaches where the coarsest transportable bedload has been effectively trapped. Sites to avoid would be downflow, where the deposited material would poorly represent palaeoflood competence. Underestimation from maximum boulder preservation and limited section exposure issues would appear to outweigh possible overestimation concerns related to fluid density and unsteady flow characteristics such as instantaneous acceleration forces. Flood data derived from river terrace deposits suggests that basal terrace geometries and coarse boulder lags common to many terrace sequences are likely the result of extreme flow events which are subsequently filled by lesser magnitude flood events, in this environmental setting.