The synthetic aperture radar (SAR) is an airborne or spaceborne radar mapping technique for generating high-resolution maps of surface target areas and terrain. Its images usually have big size and contain a large amount of data. So it is a key problem in SAR image data processing that how to compress the image to reduce data amount effectively so that it can be saved or transmitted conveniently. In recent years, wavelet transform is widely used in the field of image compression. The Spatial-Orientation Tree (SOT) Structure plays a very important role in compression of SAR image based on Wavelet transform. Both the Embedded Zero-tree Wavelet (EZW) and the Set Partitioning in Hierarchical Trees (SPIHT) coding schemes utilize the parent-children relationship in SOT. EZW is a simple, yet remarkably effective, image compression algorithm, have the property that bits in the bit stream are generated in order of importance, yielding a fully embedded code. SAR image suffer from speckle noise that seriously degrades image quality and compressibility. Removal of speckle noise can enhance correlations of pixels and compressibility of SAR image. As a very efficient structure to investigate the spatial correlations among wavelet coefficients at different resolutions, SOT has not been well used in noise removal. \;In this paper we proposed a SOT structure based method, which integrated speckle noise removal and EZW algorithm. Results of compression of large numbers of Airborne SAR images validate the proposed method is efficient and better than JPEG and EZW algorithm.