Permafrost is a direct indicator of climate change and permafrost temperature and active-layer thickness have been identified as Essential Climate Variables (ECV) by the global observing community. The existing data, however, were far from being homogeneous and were not yet optimized for databases, without framework for data reporting or archival and the data documentation was incomplete. Within the EU FP7 project PAGE 21, Arctic Portal has developed a central Data Management System (DMS) for permafrost monitoring parameters of the Global Terrestrial Network for Permafrost (GTN-P) and others. Each component of the DMS, including parameters, data levels and metadata formats were developed in cooperation with the GTN-P, the International Permafrost Association (IPA) and Arctic Portal. The researcher can now edit, visualize and download standardized datasets, metadata, charts and statistics of all relevant parameters for a specific site in all partner countries. The GTN-P DMS is based on an object oriented model (OOM) following the framework Model/View/Controller (MVC) of Cakephp. It is implemented with open source technologies with the PostGIS database and Geoserver. To ensure interoperability and enable potential inter-database search, the system follows the evolution of the Semantic Web (Linking Geospatial Data); the database structure and content are mapped towards xml, xslt, rdf, and owl. Moreover, metadata comply with the ISO 19115/2 and ISO TC/211 standards for geospatial information. Datasets are then normalized based on a control vocabulary registry. Tools are further developed to provide data processing, analysis capability and quality control. The end of the distribution chain deliver highly structured datasets towards modelers in NetCDF files, format developed by UNIDATA. The elaboration of this project highlights the absence of standardized data model for scientific relational databases as well as a lack of ontology definition and mapping within and between scientific communities.