Abstract: An improved database data structure and datatype is disclosed for storing, manipulating and accessing multidimensional spatial data in a database. Binary helical hyperspatial code (HH CODE) is used to represent data of N dimensions. The binary HH CODE data structure maintains the dimensional organization of multidimensional data within the data itself. Spatial data is stored using BH code which is modeled as a N-tree structure derived using recursive decomposition. A high water mark is set as the upper limit for data volume which may be stored in any one partition. As data stored in a partition exceeds the high water mark, the data is decomposed into child partitions such that no partition data stores exceed the high water mark. If the high water mark is exceeded, additional child partitions are automatically created and the parent table is not retained. A data structure is defined which represents the partitioned tables and BH code values.