A quality-control framework for sub-daily flow and level data for hydrological modelling in Great Britain

Abstract The absence of an accessible and quality-assured national flow dataset is a limiting factor in sub-daily hydrological modelling in Great Britain. The recent development of measuring authority APIs and projects such as the Floods and Droughts Research Infrastructure (FDRI) programme aim to facilitate access to such data. Basic quality-control (QC) of 15-minute data is performed by the data collection authorities and the National River Flow Archive (NRFA). Still, there is a need for a comprehensible and verifiable quality control methodology. This paper presents an initial assessment of the available data and examines what needs to be done for applicability of the data at national scale. The 15-minute flow series has many inconsistencies, and there are also inconsistencies with the NRFA Annual Maximum values. When producing a QCed dataset, decisions regarding the retention of data values need to be taken and recorded. Furthermore, QC should remove and rectify erroneous values, such as negative and above world record flows; and an assessment of homogeneity and truncated values in the stations could be beneficial to flag suspect data. The complex chain for production and changeability of flow and level data makes data curation and governance imperative to assure the longevity of the dataset.