一种应用遥感技术快速确定鄱阳湖区有螺洲滩的方法(英文)

A METHOD OF RAPID IDENTIFICATION SNAIL HABITAT IN MARSHLAND OF POYANG LAKE REGION BY REMOTE SENSING

目的 应用地理信息系统 /遥感技术 (GIS/ RS)快速确定鄱阳湖区钉螺孳生地带及血吸虫病高危区域。 方法 选择鄱阳湖区内 3个洲岛型血吸病重度流行村周围的洲滩地图进行数字化 ,在数字化地图的基础上分别对两张 L andsat5 (洪水期和沽水期 ) TM卫星遥感片进行校正 ,并提取出冬陆夏水洲滩部分 ,运用 ERDAS IMAGINE软件对此洲滩进行计算机非监督分类 (U nsu-pervised Classification)、校正植被指数 (Normalized Difference Vegetation Index,NDVI)和穗帽湿度变换指数 (Tasseled Cap,TC)模型计算 ;用传统方法进行螺情调查 ,卫星定位仪 (GPS)记录有钉螺点的经纬度 ,同时选择部分堤内的农田和水溏 (无螺区 )作为对照点。 结果 用计算机非监督分类共分为 10类 ,根据现场调查的钉螺分布图 ,有螺点主要分布在计算机非监督分类的 6、7和 8类中 ,进一步模型分析表明 ,有螺点主要分布在校正植被指数 NDVI>115和穗帽湿度变换指数 TC在 - 10到 3之间 ,而堤内农田和水溏等对照点中的值则主要分布此数值以外。 结论 鄱阳湖区钉螺分布和孳生与洲滩湖草的生长状况与及湿度有着密切的关系 ,本项研究提示 ,运用计算机非监督分类法可大致确定 6、7和 8类为有螺植被环境 ,然后通过校正植被指数 (NDVI>110 )

Objective To identify Onocmelania snail habitats and areas with high transmission potential by using Geographic Information System (GIS) and Remote Sensing (RS). Methods Marshland areas near high endemic villages of schistosomiasis in the Poyang Lake region were selected. Corresponding map was digitized and (Landsat 5 TM) the image was corrected according to the digital map. The image where water and land shifted during floody and dry seasons was extracted and classified by unsupervised method. The image in dry seasons was calculated by both Normalized Difference Vegetation Index (NDVI) and Tasseled Cap model. Traditional snail survey was done at the same areas and recorded the geographic position by Geographic Position System (GPS). Some cultured lands inside the embankment and water pound in marshland areas were selected as control group. Results The image of pilot areas was classified into 10 classes. The value of image after classification was extracted according to snail habitat spots as well as cultured land and water spots. The result shows snails spots distributed in class 6,7 and 8. Farther analysis of both NDVI and Tasseled Cap model shows that the snail habitats were mainly distributed in the NDVI value of more than 115, and in Tasseled Cap wetness value between 10 to 3, while the most cultured lands, inside the embankment and water pound in marshland distributed out of these values. Conclusion The marshland shifting water to land and with healthy vegetation is correlated to snail habitats in the Poyang Lake. TM data analysis shows 94.93% correction with three steps to identify the snail habitats: first step, to highlight the class 6, 7 and 8 by unsupervised classification to define the snail habitat environment; second step, to extract the value by NDVI model, where NDVI value is more than 110 so as to define a healthy vegetation as snail suspicious habitat, and third step, to use Tasseled Cap wetness model to define the areas which value is between 10 to 3 as snail habitats. With the three step of computer analysis, it is easy and rapid for identification of the snail habitats in the Poyang Lake region with 94.93% correctness.