利用Terra/MODIS数据提取冬小麦面积及精度分析

利用遥感技术提取作物播种面积是农情监测研究中一种实用而可行的方法。以河北省藁城市为研究区域,研究利用多时相Terra/MODIS数据提取2004年冬小麦播种面积的技术方法。在分析MODIS波谱特性与冬小麦生物学特征,并考虑了有关植被指数图像对面积提取精度的影响基础上,选择将MODIS数据的Red、Blue、NIR和ESWIR波段作为基础工作波段。利用4种方法提取了2004年藁城市各乡镇的冬小麦播种面积,并利用2004年各乡镇统计数据及土地利用数据进行了精度评价。结果表明：4种方法提取的冬小麦播种面积总体上都与参考值比较吻合,总体误差和平均误差均小于5%。可见,利用Terra/MODIS数据提取冬小麦播种面积是完全可行的。研究还发现,最佳的MODIS数据是冬小麦抽穗期的Red、Blue、NIR、ESWIR波段图像和冬小麦播种期与抽穗期EVI差值图像的组合。     

多时相卫星影像在2008年新疆特大干旱监测中的应用

针对新疆2008年的特大旱情,利用多时相SPOT遥感数据计算近10年新疆距平植被指数(DVI),揭示新疆2008年7月下旬旱情发生的范围和强度,并结合新疆气象站点数据检验其监测结果。结果表明:利用SPOT VGT数据能够较准确地监测出新疆干旱的发生范围和程度,其结果与气象部门旱情统计相吻合;同传统地面监测比较,该方法克服了单点的空间局限,能快速、便捷提取干旱的详细空间分布,在干旱动态监测业务中具有较强的实用性。     

基于多时相 Sentinel-2 卫星数据的农作物分类研究

【目的】探究研究区农作物分类最佳时相;结合遥感指数探究一种有效的多时相分类方法,提取主要农作物种植分布情况。【方法】基于多时相Sentinel-2卫星数据,采用支持向量机的分类方法对不同时相进行分类,对比分类精度;融合时间序列的NDVI、MNDWI指数之后采用支持向量机的方法进行分类,之后分别利用MNDWI和CI指数结合决策树的分类方法提取水域和田埂。【结果】7月份的分类效果最好,总体精度达到91.05%,Kappa系数达到0.8518;通过时相数据不同组合的分类精度比较,采用3—10月NDVI数据叠加后分类的效果较好,总体精度达到92.25%,Kappa系数达到0.8736;对比3种不同分类方法,以支持向量机的分类结果精度最高,总体精度达到94.19%,Kappa系数达到0.9024。【结论】7月份是研究区农作物分类的最佳时相;多时相分类精度明显高于单景数据分类;结合多时相NDVI、MNDWI、CI 3种遥感指数进行分类可以有效提取研究区主要农作物的种植分布情况。     

Analysis of land-use change in a sector of Upper Franconia (Bavaria,Germany) since 1850 using land register records

This study analyses changes in the landscape of a sector of Upper Franconia (Bavaria, Germany) by comparing land use changes over four time periods (1850, 1900, 1960, 2000). Geodetic and other data derived from the Bavarian real estate tax and land register were entered into various temporal layers of a land register-based vector GIS. This multitemporal GIS permits a precise analysis of the historical structure and development of landscapes on the basis of land plots.In 1850, the study area was almost exclusively agricultural in structure. Woodlands made up only 18% of the total surface. Rough pastures and wastelands, which covered about 9% of the total surface, were used for grazing. During the first half of the 20th century, the proportion of wooded areas increased considerably. The rough pastures that had formerly been a typical feature of the region nearly disappeared during this period. Agricultural use declined to less than 50% of the total area. In the course of the period between 1960 and 2000, the livestock industry has become an almost exclusively indoor activity. Village development has started spilling over into the adjacent fields. The causes and background of these changes are discussed in detail.From an ecological standpoint, the land use categories surveyed in this analysis of landscape change can be regarded as vegetation types, thereby constituting habitats for specialized biota. The intensity and frequency of any type of land use creates a certain disturbance regime, which disrupts and controls the succession in a certain way. The concept of categories of change incorporated into the GIS helps to evaluate these habitat types and the rate of change more accurately, e.g. for nature conservation purposes.