A critical re-evaluation of the prediction of alkalinity and base cation chemistry from BGS sediment composition data

The model of Begum et al. (2010) that predicts alkalinity and Ca and Mg concentrations in river water from available sediment composition data has been critically re-evaluated using an independent validation data set. The results support the hypothesis that readily available stream water sediment elemental composition data are useful for prediction of mean and minimum concentrations of alkalinity and Ca and Mg in river water throughout the River Derwent catchment in North Yorkshire without requiring land-use data inputs as stream water sediment composition reflects all aspects of the riparian zone soil system, including land-use. However, it was shown for alkalinity prediction that rainfall exerts a significant dilution effect and should be incorporated into the model in addition to flow path-weighted sediments Ca% and Mg%. The results also strongly suggest that in catchments with substantial rough moorland land-use neutralization of organic acids consumes alkalinity and this fact should be considered in any future development of the model. •Published river sediment composition data are valuable for predicting water chemistry.•River water Ca and Mg and alkalinity may be predicted from sediment chemistry.•Organic acids reduce alkalinity, especially in upland streams.•Rainfall regulates river alkalinity distribution across a large calcareous catchment.