中文摘要：

在遥感解译和景观类型制图的基础上，运用滑箱扫描的方法，基于组分边界特征计算景观整体结构指数，同时综合考虑景观要素的空间分布，采用信息熵法和孔隙度指数法对黄河三角洲农田和柽柳-芦苇2种景观组分进行异质性特征分析。结果表明：黄河三角洲景观组分之间的边界趋于复杂化，城镇扩展、石油开采及黄河断流等使耕地大量退化为柽柳-芦苇，潮滩植被侵满滩涂；与基于面积的景观指数相比，基于边界特征的景观指数更能反映景观结构的细节变化；农田和柽柳-芦苇均表现为宏观异质性特征，部分样线中小尺度上表现为微观异质性特征，但柽柳-芦苇的异质性表现尺度较农田小50个像元左右，且这种差异随时间有加大的趋势；2种异质性分析方法从不同的侧面揭 示了组分类型的异质性特征，形成有益的补充。

英文摘要：

Taking Yellow River Delta as study area, the landscape element information of August 24, 1992 and August 9, 2001 were extracted by supervised classification, and the boundary characteristics of various elements were acquired by using 2 × 2-pixel boxes skimming over the landscape map. The landscape indices such as diversity, evenness, heterogeneity, and spatial complexity were calculated based on boundary number and cumulative length, and, considering of the spatial distribution and gradient change of various landscape elements, the landscape components heterogeneity of farmland and Tamarix chinensis - Phragmites communis were analyzed by using information entropy and lacunarity index methods. The results indicated that in study period, the boundary characteristics were more complicated, reflecting the transformation from farmland to T. chinensis -P. communis as a result of town expansion, oil exploitation, and water break off in the Yellow River. Compared with those based on element area proportions, the landscape indices based on boundary characteristics could give more detailed information on landscape structure. In 1992-2001, the spatial distribution of farmland and T. chinensis - P. communis in study area showed macroscopic heterogeneity, while scales. The heterogeneity scale of T. chinensis -P. partly microscopic heterogeneity at smaller communis was 50-pixel smaller than that of farmland, and the scale gap between them enlarged with time. Information entropy and lacunarity index methods, especially used together, could well indicate the landscape heterogeneity.