基于地表覆盖物光谱特征的土地覆被分类系统——以鄱阳湖流域为例

Land Cover Classification System Based on Spectrum in Poyang Lake Basin

土地覆被主要表现为陆地表面的形态特性和动态变化特征,土地利用是其动态变化的外在驱动力,植被是土地覆被的“主要”而非唯一的组成部分,土地覆被也包含着它的功能特性。地表覆被物的光谱特征及其季节性变化规律是土地覆被的本质特征,也是土地覆被分类的基本依据之一;利用遥感资料所表达的地表覆被物特征值(如以NDVI为指标的绿度值、始绿期、终绿期、波形、峰值、谷值等)可以方便地对土地覆被分类标准进行量化处理;对于一些复杂的复合单元,应着重于它的组合(包括组成成分和结构)特性及其对地面过程的影响来赋予更确切的内涵。据此,构建了包含7个一级类别和20个二级类别的开放式(open-end)的鄱阳湖流域土地覆被分类系统,并且应用SPOT4-VEGETATION的时序NDVI数据,对各类别的面积比重进行了分析,结果显示鄱阳湖流域的现状覆被主要由常绿覆被和作物—树—草混合覆被以及人工栽培植被(作物)等构成。

In this paper, a new land cover classification system has been built according to the following ideas. Firstly, the formation of ＂Land Cover＂ has far-reaching geographic and biologic background, but global environmental change research and remote sensing give it new connotations. Land cover mainly represents morphological property and dynamic characteristics of the earth＇s surface. Vegetation is the primary component of land cover, but not only one. The functional properties of land cover, such as refiectivity, hydrothermal flux and net emissions of CO2 vary with different components and structures of covers. Secondly, spectral response and characteristics of seasonal variability are the fundamental characteristics of land cover. According to the difference in spectrum, it is easy to distinguish green land cover（mainly vegetation）, blue land cover （mainly water）, grey land cover （mainly urban or built-up land, and semi-desert）, white land cover （mainly snow and ice）, non-vegetated land cover（mainly desert）, etc. Similarly, characteristics of land-cover＇s seasonal variability derived from remote sensing images can make some typical land cover easy to be distinguished further, such as evergreen land cover has high NDVI index and no remarkable change in a year, deciduous forest single peak of NDVI sliding curve in a year, annual double crops double peaks, and annual triple crops triple peaks. Therefore, we can make the classification criteria quantified with the characteristic values indicated by NDVI including greenness value, green-up period, green-end period, crest value, trough value and waveform of NDVI sliding curve in a year, etc. Moreover, mixed cover should be differentiated by its structural characteristics and functional properties. Based on these ideas, a two-level land cover classification system has been proposed for the clustering of Poyang lake basin＇s land cover by using multi-temporal SPOT4-VEGETATION sensor data. The primary types include Evergreen Cover, Seasonal Green Cover Ⅰ （Woody）, Seasonal Green Cover Ⅱ （Herbaceous）, Seasonal Green Cover Ⅲ （Crops）, Seasonal Green Cover Ⅳ （Mixed）, Grey Cover and Blue Cover. The secondary level includes 20 types. According to area proportion, the main land-cover types in Poyang lake basin are evergreen cover （40.05%）, crop-forest-grassland mixed cover （38.29%）, and crops （16.63%） at present.