空间粒度变化在时间序列上对景观指数的影响——以环洱海区域为例

Effects of Grain Size Change on Landscape Index in Time Series: A Case Study on Erhai Lake Area

以环洱海区域为例,对1990,2000,2009,2014年不同传感器的Landsat遥感数据进行监督分类,在此基础上计算1990—2014年尺度在35~200 m范围内的空间粒度上25种景观指数的变化情况,并结合各指数在此时间序列上的变异系数、粒度效应曲线和函数拟合模型进行空间粒度效应的探讨。结果表明:(1)景观指数在空间粒度序列上的响应类型分为无规律型、上升型、下降型和稳定型4类;(2)在5类景观指数中,多样性类指数对空间粒度变化不敏感;(3)在时间序列上,大部分指数的年度变化规律因粒度变化而明显改变,直接影响景观格局变化研究的结果;(4)多数面积/边缘类指数、形状类指数和聚集类指数的线性拟合度良好,适合进行粒度上推绎预测。

Taking Erhai Lake Area as the case study,based on images processing and supervised classification of landsat remote sensing data of 1990,2009,2000 and 2014,the change of 25 landscape indices in 35 ~ 200 m range of spatial granularity from 1990 to 2014 is calculated to investigate the influence of combined index,variation coefficient,particle size effect curve and linear function fitting model of each index in the time series of grain size effect. Results show that:( 1) Landscape indices are divided into four categories in the space size sequence: no rules,rising type,descend type and stable type;( 2) The diversity index is not sensitive to the change of grain size in the five landscape index.( 3) In time series studies,most of the annual variation of the index changes with the change of grain size. This directly affects the results of the study of landscape pattern change;( 4) The majority of the area/edge index,shape index and aggregation index of the degree of linear fitting is good,suitable for particle size prediction.