Based on the singular value decomposition (SVD) of a data matrix collected by the array sensors
a new approach to space-time two dimensional spectrum estimation is presented in this paper.The noise singular vectors are got under the singular value decomposition of the formed matrix.By using the orthogonal properties of the noise space and the signal space
we can estimate a high resolution space-time two dimensional spectrum.The presented method applies to the fields such as taking bearings and capturing information of the velocities.The results of computer simulation illustrate that the new method is quite effective.