A Screening Tool for REE Data Quality Assessment

The growing availability of advanced analytical techniques has led to a rapid increase in published geochemical data, much of which is now accessible through curated, domain-specific synthesis databases. These resources offer significant opportunities for innovative research in geochemistry and related fields through the analysis of compositional data from rocks, minerals, and other natural materials. However, integrating and evaluating data that were collected over several decades using widely different analytical methods is a challenge.

The rare earth elements (REEs) are widely used in geochemistry as tracers of chemical transport, differentiation, and broader Earth system processes. To address common issues in REE data usability, we developed an automated method to identify scattered, anomalous, or potentially erroneous REE patterns. Our approach draws on the expected geochemical behaviour of REEs, focusing on the “smoothness” of normalised REE patterns, the detection of single or multiple outliers, and various forms of scatter.

We validated our method using large datasets from the GEOROC (n=176,611) and PetDB (n=30,659) databases. Distinct types of anomalous REE patterns are related to potential artefacts and data quality issues. Additionally, we test the application of our REE screening method on individual published data sets from various sources and dare to give some recommendations.