Landscape pattern analysis is an important topic of landscape ecology. The ultimate goal of landscape pattern analysis is to link spatial patterns of landscape with ecological processes, and detect status of processes using landscape pattern information. Landscape metrics can represent the spatial distribution of landscape and it has been used as a common tool in landscape pattern analysis. There are some inherent limitations of landscape metrics. Multi-distance spatial cluster analysis based on the Ripley's k-function can fetch up the landscape metrics faults. Based on the image data of Land Resources Satellite TM (2000), China-Brazil Earth Resources Satellite (2004), and Environment Disaster Monitoring and Forecasting Small Satellite (2011), future landscape change were forecasted with bringing forward simplified process on data transform of landscape pattern change prediction and multi-distance spatial cluster analysis, via analyzing landscape characteristics and its spatial pattern in Xi'an city, using ENVI 4.7, ARCGIS 9.2, and IDRISI 15 in this study..Results showed that, tremendous landscape patterns changes have taken place in Xi'an during the past two decades. Complex landscape matrix consisted of forest land and cropland accounted for 67.98% of the total landscape in 2000 and 54.46% in 2011 in the study area. Built-up land area increased from 33213.91 hm² to 68380.79hm², and amount of increased area between 2000 and 2004 was more than that between 2004 and 2011. The area of forest land decreased insignificantly. The shift of water and unused land area was insignificant. Patch number in the class and landscape level decreased obviously, which showed that decreasing landscape fragmentation, enhancing connectivity of forest landscape, and reducing connectivity of farmland was taking place during the study interval. Landscape spatial pattern had a significant aggregation in the expected maximum distance (40 km). The critical thresholds discrepancy of clustered, random, and dispersed distribution between different landscape types in different years were relatively evident; the spatial aggregation levels of water and unused land were significantly higher than those of farmland, woodland, grassland, and built-up land; there were the largest characteristic scale of heterogeneous spatial distribution of aggregated distribution, random distribution, and dispersed distribution for cropland and grassland. The effect of landscape pattern characteristic study via coupling landscape index analysis and multi-distance spatial cluster analysis is better than that of using single landscape index analysis. The simulation of CA-Markov model could basically reflect the future landscape changes. The method and the simplified step of multi-distance spatial cluster analysis by ARCGIS and data conversion for landscape pattern forecasting by IDRISI were proposed in this paper.