Due to the diversified behavioral characteristics and regular pattern of moving objects， the trajectory data generated by these objects shows obvious uneven distribution feature in time and space， which may lead to worse performance for trajectory data indexing and querying. However， the existing trajectory data indexing methods rarely consider this problem. In this paper， a temporal-spatial distribution based indexing and querying method is proposed. In the method， Geohash code is introduced and pre-partitioned spatially by utilizing the temporal-spatial similarity of the trajectory data distribution. Then， we use the pre-partitioned Geohash code， partition number and the trajectory data timestamp to compose the index structure. With this index structure， a storage model based on HBase and a query algorithm based on Geohash partition are designed respectively. The empirical study using real trajectory dataset shows that the method improves the spatiotemporal query performance of trajectory data by comparing with the Extend_HGrid and ST-hash methods， and effectively reduces the number of sub-queries during query.