In this paper, the general statistical structure and characteristics of meteorology grid data were analyzed. By analyzing the correlation between neighbor grid data, calculating its relevant symbol entropy and information redundancy, it was found that there is more information redundancy in meteorology grid data, so the grid data set can be better potentially compressed. The higher the correlation is, the more the condensability is. Based on above analyses, a 2-dimensions liner predictive statistical model for meteorology grid data was established to reduce information redundancy, and by combining of Huffman encoding which is efficient information source encoding scheme, a new scatheless compression scheme was designed to deal with meteorological grid data. By using the new scheme, the compression ratio for general meteorology grid data can be effectively promoted, and the compressing and uncompressing process of data is complete scatheless within available precision. Finally, a set of contrast experiment based on the usual meteorology grid data were presented and carried out. It is proved that the compression efficiency of the new scheme is evidently superior to that of GRIB and netCDF. The new compression encoding scheme may be widely used to treat with the vast data in meteorology and other earth sciences to improve the efficiency of data collecting, storing, translating and exchanging.