CSIRO Mafic-Ultramafic Geochemical Dataset

This dataset includes whole-rock geochemical (public) processed data collected through CSIRO research projects from the mid-1980s to 2021. It represents information over 6000 samples, primarily from the Archean of Western Australia, with a heavy bias towards komatiites and immediately associated rocks and nickel sulphide ores. Most of the data were collected as part of the research projects exploring magmatic sulphide nickel-copper ore deposits. The dataset also includes samples from mafic-ultramafic intrusive rocks in the Halls Creek and Albany-Fraser orogens. \n\nThe dataset has been used in several review publications, as well as in a large number of publications on specific localities. The prominent review publications also draw extensively on other public datasets such as GeoROC and PetDB; for further information, see Related Links.\n\nWe envisage the collection as an interrogation tool for all subsequent Nickel-related projects in Australia.\n\nThe whole dataset can be downloaded as a CSV file (see 'Files' section). Alternatively, you can download the same dataset by a specific area using the Exploration Toolkit data portal at https://data.exploration.tools/\nLineage: This dataset represents an accumulation of whole rock geochemistry and associated location metadata accumulated over approximately 30 years of CSIRO research on magmatic ore deposits and Archean and Proterozoic greenstone belts and mobile belts. \n\nThe dataset is a subset of the migrated legacy whole rock geochemistry and mineral database, from Microsoft Access to PostgreSQL (an open-source object-relational database system). The migration includes several improvements over the original database. For example, we normalized the records to avoid adding redundant data to the database. We specified the primary key and the foreign key constraints for each table to enforce data integrity in the tables. Tables and column names follow a standardized naming convention (i.e., snake-case). All identical fields have similar names throughout the tables for ease of use. We updated column types as required, e.g. from integer to float. We used external data sources (e.g., GSWA Mineral Drillholes Database) to clean existing information on drill holes. We apply controlled vocabularies to allow a consistent representation of data across records. For example, spatial projection now uses the EPSG codes, and rock types are specified based on the rock names classification from Mindat.org. Variables measured (analytes) are accompanied by appropriate units.\nWe only included data with geographic location in the data file. If you need all data including raw and processed, please get in touch with the authors.\n\n