Development of new river channels and river maturation in the Rhine-Meuse delta, during the Late Roman Period and Early Middle Ages, AD 300-1000.

This study focuses on the spatial trends in new river course development of three major rivers in the Netherlands. The alluvial architecture and paleogeography of the Dutch river plain area have been studied over years. There are over 200,000 corings and 1700 relevant 14C-dates, which made identification of different river generations possible. However, temporal resolution of most dates is too low to answer questions of longitudinal and lateral river channel development on a centennial time step resolution for a single river channel. Higher time step resolution of river channel maturation can improve conceptual models of alluvial architecture in general and can be used to validate computer models on river channel development and maturation. This study aims to (1), avulsion history reconstruction on a centennial time step resolution, (2) evaluation of assembly of multidisciplinary datasets and (3) to give more insight in river channel maturation in longitudinal and lateral direction. A conceptual river phase model (from non-existence to buried residual channel) is introduced to enable integration of both existing and new 14C-dates to distinguish local sedimentary and hydrological footprints of individual dates from regional trends. For each phase, from non-existence to buried residual channel, related dating methods are combined originating from historical sources, archeology, lithology, dendrology and type of 14C-dates. New sampling sites have been selected based on existing cross-sections and new cross-sections, that were drawn using existing data. Loss-on-Ignition (LOI) of the sampled cores is used as 14C sub-sample selection strategy as clay-on-peat transitions could be identified in more detail. For sequenced dates AMS 14C-dates have been calibrated using Oxcal Sequence Dating. Begin of sedimentation ages for the river channels were tightened with new 14C-dates. Sedimentation started almost simultaneously over the river course, as no clear longitudinal and lateral trends in river channel development during maturation could be deduced from the data. Classification of 14C-dates to different phases of the river phase model was useful, although still a large within phase heterogeneity of 14C-dates was visible. As no large actual time hiatus in lateral direction between levee and floodplain deposits is observed in this study, future clay-on-peat samples (for type 1 dates, cf. Berendsen, 1982) should and could be taken not too close to the natural levee. 14C-samples thereby always represent a local sedimentary and hydrological footprint and cannot be simply combined in analysis only based on their bare ages. And although lithological information is present in 14C databases, lithological sequence information, as well as LOI-curves when used for 14C-sampling, should be inserted in the 14C databases, as this information is conditional for meaningful further analysis of 14C-dates. Moreover LOI-curves can be used to identify possible oxidation of peat, and to visualize local river maturation (sudden or gradual) in more detail. Further improvement of timing of river maturation also contributes to studies which attempt to combine environmental, cultural or archaeological data in a high time step resolution