Data grid is powerful for distributed data sharing and collaborating, which is very important for the spatial information applications. But the difference among meta-data produced by space agencies around the world makes it difficult to re-organize data and share them. In this paper, a way named meta-data adapter is proposed, which can interpreter original meta-data with different syntax and semantic rules to required meta-data so that they can be organized as a distributed, heterogeneous, synergic, uniform interfaced, and inherent features based virtual remote sensing data catalog. Based on the meta-data adapter, SIG can build up a powerful and flexible data grid for spatial information applications. The experience of ldquoremote sensing data distribution system for WenChuan earthquake rescuingrdquo amazingly quick founding shows that meta-data adapter is a powerful and economical way to solve the problem of meta-data difference in data grid for spatial information applications. And it may be more powerful and extensible with the help of situation-based intelligent automatic adapter generating.