Aiming　at　the　unstructured　brain　data　and　data-driven　research　process,　provenances　have　become　an　important　component　of　brain　and　health　big　data　rather　than　the　accessory　of　raw　experimental　data　in　the　systematic　Brain　Informatics　(BI)　study.　However,　the　existing　file-based　or　transaction-database-based　provenance　queries　cannot　effectively　support　quickly　understanding　data　and　generating　decisions　or　suppositions　in　the　systematic　BI　study,　which　need　multi-aspect　and　multi-granularity　provenance　information　and　a　process　of　incremental　modification.　Inspired　by　studies　on　the　data　warehouse　and　online　analytical　processing　(OLAP)　technology,　this　paper　proposes　a　BI　provenance　warehouse.　The　provenance　cube　and　basic　OLAP　operations　are　defined.　A　complete　Data-Brain-based　development　approach　is　also　designed.　Such　a　BI　provenance　warehouse　represents　a　radically　new　way　for　developing　the　brain　big　data　center,　which　regards　raw　experimental　data,　provenances　and　domain　ontologies　as　different　levels　of　brain　big　data　for　data　sharing　and　mining.