The Geography Markup Language (GML) has been introduced to standardize the representation of geographical data in Extensible Markup Language (XML), which makes the exchanging and sharing of geographical information easier. As GML becomes the geospatial language of the Web, the compression of GML files becomes a new challenge to the compression research field. That is because, GML documents tend to be much larger in size than other documents containing the same information, which raising concerns about processing and transportation. In this paper, we first introduce the architecture of the explored GML compressor. Then, analyze the main redundancy characteristic of GML. After that, we explain the precision decrease, delta compression and extrapolation techniques for GML spatial data compression. Experimental results show the efficiency of our proposed methods.