Framework development and flag-based quality control for a national scale dataset using UK's Historical 15-Minute Flow Data

The United Kingdom has an extensive repository of 15-minute flow data dating back to the 1930s, yet this wealth of information has remained decentralized within respective measuring authorities responsible for localized quality control. Consequently, the absence of standardization has resulted in heterogeneous data records. Several discrepancies can be observed, ranging from minor issues such as having different decimal places, to bigger issues such as having duplicate records with different values or having different quality codes in the data.

In the aim of producing a quality assured and consistent 15-min flow dataset for the whole UK, data has been requested from all UK measuring authorities. The data collected laid the groundwork for the development of a quality control framework, featuring both traditional, amply academically used and UK specific quality control flags. These flags have been used to standardise the data and produce a quality assured 15-min flow dataset for the UK.

More than 1000 stations and tens of thousands of years of data have been passed through different flags aiming to identify data and stations that have suspicious data. 14 flags have been generated in the framework. The flags vary in complexity and aim to provide better understanding of the data.  Even simple flags, such as detecting negative values serve multiple purposes: from identifying tide-influenced stations characterized by negative flows, to using the flag to remove/replace the negative values for hydrological analysis.  Conversely, complex hydrology flags such as identifying large flow events preceded by large rainfall events or identifying the relationship between the high flow of stations in the same river can be used for an enhanced comprehension of hydrological systems at a national scale.

This presentation aims to elucidate the flags that have been applied to the data; spotlight interesting case studies discovered in the quality control process; and showcase the versatile applications of the flags in data selections for specific hydrological analysis. In this PICO we want to emphasize the pivotal role that appropriate data selection has in shaping robust conclusions in the field of large sample hydrology.