人工衛星データ及び地理情報システムを用いた公共牧場の動態解明に関する研究

对岩手县46个公共牧场进行了调查。记录每个牧场的持续时间、总面积、人工草地面积、利用方式和放牧动物。每个牧场采集1：25000比例尺的地图图像和1996年5月29日观测的Landsat TM图像，并将这两幅图像叠加以选择草地区域。对部分牧场的芒花型植被进行了调查。结合实地调查结果，应用最大似然法进行土地覆盖分类。在Hinoto牧场、Kuzakai牧场、Yamatanikawame牧场、Sumida牧场、Natsum usiyama牧场、Niizato牧场、Kanegasaki牧场和Tonan牧场，芒草被引入人工草地。在某些情况下，直立死亡的草被错误地归类为芒属中华。秋季是芒花盛开的季节，利用遥感影像进行分类可以提高分类的准确性。1985年，随着畜牧业的减少和集约化经营的推广，草地的放牧利用逐渐停止，草地植被逐渐发生变化。2002年进行植被复垦，并与1981年进行比较。结缕草的SDR_2降低，蕨类植物的SDR_2降低。库恩、观赏花椒和西伯利亚海棠在这21年中有所增加。多时相遥感数据显示了归一化差异植被指数(NDVI)的年际变化，反映了该时期植被的变化。

46 public pastures in Iwate prefecture were investigated. Duration, total area, area of artificial grassland, way of use and grazing animals were recorded in each pasture. Map images on a scale of 1 to 25,000 and Landsat TM images which were observed May 29,1996 were collected by each pasture, and both images were overlayed to select grassland area. Miscanthus type vegetation was investigated in some pastures. And land cover classification by maximum likelihood method was applied with the result of filed investigation. In Hinoto pasture, kuzakai pasture, Yamatanikawame pasture, Sumida pasture, Natsumusiyama pasture, Niizato pasture, Kanegasaki pasture and Tonan pasture, Miscanthus sinebsis was invated into artificial grasslands. In some cases, standing dead of grass was misclassified into Miscanthus sinensis. The use of remote sensing images observed in autumn, which is the blooming season of Miscanthus sinensis, will improve the accuracy of classification.Zoysia type grassland was maintained for long period in Appi pasture. But grazing use was ended in 1985 by decrease of livestock farmer and extension of intensive management, and vegetation of garssland was changing gradually. Vegetation servay was practiced in 2002,and the result was compared with that in 1981.SDR(Summed Dominance Ratio) 2 of Zoysia japonica decreased and SDR2 of Pteridium aquilinum (L.) Kuhn, Anthoxanthum odoratum L.and Malus sieboldii increased in these 21 years. Multi-temporal remote sensing data showed yearly change of normalized difference vegetation index(NDVI), which reflect the vegetation change of this period.