To address the application scenarios of cluster based or multiple sink based wireless sensor networks, this paper proposes a Top-|K| query based distributed data reconstruction method which includes the distributed iterative hard thresholding algorithm and the dual thresholds based distributed Top-|K| query algorithm. The administrative node and member nodes run the distributed iterative hard thresholding algorithm, realizing the iterative hard thresholding algorithm in a distributed manner at the same time. Meanwhile, the administrative node and member nodes run the dual thresholds based distributed Top-|K| query algorithm which can query the Top-|K| sum of the elements in the result of the previous algorithm in a distributed manner as well. Experimental result show that the proposed method can achieve the same data reconstruction performance as the existing methods. Furthermore, it can also decrease the transmitted data number and interactive time between the administrative node and member nodes.