Geothermal Heritage Data into Digital Repositories: The IGME-CSIC Case Study

The effective use of Artificial Intelligence (AI) in geosciences depends on the availability of high-quality, consistent, and traceable data. In geothermal research, the acquisition of primary data (such as boreholes, geochemical sampling, and geophysical surveys) requires substantial technical and financial resources, making the recovery and structuring of historical archives a priority. This work presents the first phase of the digital transformation of the geothermal heritage of IGME-CSIC, carried out within the framework of the Sustainable Management of Mineral Raw Materials Project (MPMIN). The process focused on the systematic search, validation, and digitisation of more than 70 years of documentary references, combining consultation of physical and digital repositories with the design of bibliographic records based on metadata inspired by Dublin Core and integrated into a relational SQL database with a management interface in Microsoft Access. The selected documents were digitised in preservation formats (TIFF, PDF) with Optical Character Recognition (OCR), and the metadata quality was verified through manual review. The results include the identification and cataloguing of more than 400 references, of which 55% were digitised with OCR, 20% were only images, and 25% remained pending processing. The resulting repository, linked to the IGME-CSIC Document Information System, provides advanced search and export functionalities, thus becoming an active and scalable data infrastructure. This experience demonstrates that transforming dispersed archives into structured and interoperable repositories preserves scientific heritage and enables future comparative analyses, multitemporal studies, and robust predictive models through data extraction.