在遥感时间序列数据分析中马尔柯夫链方法与空间信息最佳结合的探讨

A STUDY ON THE BEST COMBINATION OF MARKOV METHOD AND SPATIAL ANALYSIS WITH TIME-SERIES DATA OF REMOTE SENSING

以北京顺义区为例 ,在 TM图像资料分析基础上 ,建立马尔柯夫过程与直接距离法及距离加权法 2种不同空间相关分析法的时空模型 .分别以 8,4 8,80 ,12 0和 2 2 4邻域的不同邻域数对顺义县 1996 ,1997年的土地利用变化进行模拟检验 .结果表明马尔柯夫过程的时空模型中 ,采用 4 8邻域或

Based on TM data of Shunyi County, Beijing,China,two different spatial temporal models are developed, including weighted distance method and direct neighborhood function. And the land use transition of Shunyi is simulated on 8,48, 80,120 and 224 neighbors respectively in 1996 and 1997. The results show that the spatial temporal model with 48 or 80 neighbors of weighted distance neighborhood analysis is the best.