Data warehouses (DW) and OLAP systems are technologies allowing the on-line analysis of huge volume of data according to decision-makers' needs. Designing DW involves taking into account functional requirements and data sources (mixed design methodology). But, for complex applications, existing automatic design methodologies seem inecient. In some cases, decisionmakers need querying, as a dimension, data which have been defined as facts by actual automatic mixed approachs. Therefore, in this paper, we offer a new mixed renement methodology relevant to constellation multidimensional schema. The proposed methodolgy allows to decisionmakers to enrich a dimension with factual data. In order to validate our theoretical proposals, we have implemented an enrichment tool and we have tested it on a real case study from bird biodiversity