EN中文

Journal of Geo-information Science >

2016 , Vol. 18 >Issue 6: 824 - 832

DOI: https://doi.org/10.3724/SP.J.1047.2016.00824

Orginal Article

The Effects of Changing Grain on Landscape Metrics with High-resolution Image in the South-eastern Coastal Region of Fujian

  • RAN Jianbo , 1 ,
  • CHEN Xingwei , 1, 2, 3, *
Expand
  • 1. College of Geographic Sciences, Fujian Normal University, Fuzhou 350007, China
  • 2. Cultivation Base of State Key Laboratory of Humid Subtropical Mountain Ecology, Fuzhou 350007, China
  • 3. Fujian Provincial Engineering Research Center for Monitoring and Assessing Terrestrial Disasters, Fuzhou 350007, China
*Corresponding author: CHEN Xingwei, E-mail:

Received date: 2014-11-25

  Request revised date: 2015-09-09

  Online published: 2016-06-10

Copyright

《地球信息科学学报》编辑部 所有
Fold

Abstract

The effects of changing grain on landscape metrics is an important topic in landscape ecology studies. The original data in the most previous studies were derived from moderate/low resolution data and how the high-resolution data could affect the grain effect has not yet been well investigated. Therefore, we selected the land cover datasets produced from SPOT 5 imagery (with a spatial resolution of 2.5 m) in three watersheds located in the south-eastern coastal region of Fujian Province, China. We examined the behaviors of 28 landscape metrics in varying the range of grain size varing from 2.5 m to 150 m, where the grains coarser than 2.5 m were aggregated through majority filters. We then compared the differences between scaling functions fitted with the data from 2.5 m to 150 m and 30 m to 150 m. The results show that the effects of changing grain on 28 landscape metrics are obvious in the three watersheds examined. The responses of the metrics to changing grain size can be divided into four categories, while previous studies reported only three categories. The newly category is named as TypeⅡ including eight metrics that behaved split-up predictable responses with scale inflexion at 5 m, 7.5 m or 10 m respectively. This indicates that the high-resolution data can reveal more detailed effects of changing grain on landscape metrics. The results also show that the scaling functions for ED, SHAPE_MN, CONTIG_MN and AI, are sensitive to the spatial resolution of the raw data. Those scaling functions obtained from the moderate-resolution data may not be applicable to estimating the landscape metrics for grains finer than 10 m.

Key words： landscape metrics; grain effect; scaling functions; SPOT5; south-eastern coast region of Fujian

Cite this article

RAN Jianbo , CHEN Xingwei . The Effects of Changing Grain on Landscape Metrics with High-resolution Image in the South-eastern Coastal Region of Fujian[J]. Journal of Geo-information Science, 2016 , 18(6) : 824 -832 . DOI: 10.3724/SP.J.1047.2016.00824

1 引言

景观指数（Landscape Metrics/Indices）可高度概括复杂的景观格局信息,定量描述景观的结构组成与空间配置等特征[1]。尽管景观指数的生态学意义有待进一步探讨,但它们仍是景观格局分析的重要工具[2-3]。随着空间信息技术的发展,景观指数已广泛应用于土地利用/覆被变化分析[4]、森林管理[5]、生物多样性评价[6]、水质监测[7]等方面,因而景观指数的相关研究也倍受关注[8]。
景观格局与生态过程的相互关系往往表现为多尺度特性[1,9]。尺度一般包括幅度与粒度2个方面[9],空间粒度（空间分辨率）指景观中最小可辨识单元大小[1]。众多研究表明,空间粒度变化对景观指数具有显著影响[10-14]。获取不同粒度下景观格局的常用方法是通过优势规则把矢量（或者栅格）类型图重建为一系列更粗粒度大小的栅格形式,进而反映其相应粒度下实际生态系统中的景观格局特征[15-16]。采用该方法,Wu等[12]以4种真实景观为对象,通过尺度检测图分析,将19种景观指数的粒度变化响应规律分为预测响应型、阶梯变化型、波动变化型3类,并指出预测响应型指数可使用线性、对数、幂函数等尺度函数下推至更细粒度并对其格局特征进行预测。同样,在爱沙尼亚[17]、美国凤凰城区[18]、地中海沿岸干旱区[19]的相关研究中也使用了预测响应型指数。值得注意的是,高分辨率数据的景观指数粒度变化响应规律研究集中于城市景观[20-22],其所得规律与中、低分辨率数据也不尽相同。
尺度下推是重要的尺度问题之一[23-24],通过尺度变化响应规律,建立尺度函数关系,推测细尺度下的格局特征,以便进一步研究不同尺度下格局与过程之间的相互关系[25]。然而,以往有关景观指数的粒度效应研究多局限于定性描述粒度变化响应规律。Saura等[23]以中、低分辨率数据为基础,测试了Wu等[12]报道的预测型指数,表明平均斑块面积、边界总长等景观指数的尺度下推误差较小。高燕[26]基于空间分辨率为30 m的TM影像,选取4种预测响应型指数,发现2个粒度下的真实值与其在30~300 m以内所建立尺度函数拟合值的相对误差较小,但并未进一步分析这些预测响应型指数的尺度函数下推误差。这些研究均以中、低分辨率遥感影像（或中小比例尺类型图）为基础数据,高分辨率景观指数的粒度变化响应规律及其尺度下推精度尚需进一步的研究。
闽东南沿海地区是福建省经济最为活跃的地区之一。随着经济社会的快速发展,园地大面积侵占林草地,农田城镇化明显,道路交通纵横交错,景观格局破碎化严重;不合理的开发,加上复杂的地形地貌等自然因素的影响,使本区生态环境脆弱,水土流失严重。因此,本文基于SPOT5遥感影像（空间分辨率为2.5 m）解译所得的土地覆被矢量图,通过优势规则聚合为一系列粒度大小的栅格类型图,并统计各类型图在景观水平上28种常用景观指数的数值;系统地分析了高分辨率影像下景观指数的粒度变化响应规律与尺度函数下推精度,为闽东南沿海地区的生态修复与景观规划提供参考依据。

2 研究数据与方法

2.1 研究区概况与数据

闽东南沿海地区属于亚热带海洋性季风气候,年均降水量为1000 mm,年均气温在20 ℃左右。本区代表性土壤为红壤和黄壤;地形地貌复杂,以山地丘陵为主,水系密布,河流众多;森林植被在1949年以前破坏殆尽,原生特征已不明显[27]。
Fig. 1 The location and elevation distribution of the three watersheds in the south-eastern coast region of Fujian Province

图1 闽东南沿海地区3个流域的地理位置与高程分布

本文从闽东南沿海地区的西部中低山区、中部低山丘陵区、东部台地平原区中,分别选出坑仔口溪（290 km2）、诗溪（250 km2）、九十九溪（360 km2）3个流域作为研究区（图1）;以2010年3个流域SPOT 5遥感影像为观测数据。首先,经过几何校正、图像融合（将多光谱影像与全色波段融合为2.5 m空间分辨率的遥感影像）等预处理;然后,根据已建立解译标志,采用面向对象分类软件eCongnition Developer 8.9人机交互解译所得29类土地覆被景观（矢量数据,总体精度为89.12%）,并对解译数据进行拓扑检查。考虑与福州市区域特征的相似性,将29类土地覆被景观归并为耕地、园地、林地、灌丛、草地、建设用地、道路、水域、其他用地9类[28]。从图2可知,坑仔口溪与诗溪均以林地为主,九十九溪以建设用地、林地、耕地为主。

2.2 研究方法

从景观水平上选取常用的28种指数（表1）,具体计算公式与生态学意义可参见文献[1]、[29]。考虑部分景观指数之间度量格局信息存在重复与冗余[30],如斑块数量（NP）与斑块密度（PD）,边界总长（TE）、景观形状指数（LSI）与边界密度（ED）,以及有效网格大小（MESH）与加权平均斑块面积（AREA_MN）,本文未予重复选取。
Tab. 1 List of the 28 landscape metrics used in the study

表1 景观水平上所用到的28种指数

景观指数 英文名称 缩写 单位
组成特征 斑块密度 Patch density PD 个/100 hm²
边界密度 Edge density ED m/hm²
最大斑块指数 Largest patch index LPI %
平均斑块面积 Patch area distribution_Mean AREA_MN hm²
加权平均斑块面积 Patch area distribution_Area-weighted Mean AREA_AM hm²
斑块面积标准差 Patch area distribution_Standard Deviation AREA_SD hm²
斑块面积变异系数 Patch area distribution_Coeffient of Variation AREA_CV %
平均斑块回旋半径 Radius of gyration GYRATE_MN m
加权平均斑块回旋半径 GYRATE_AM m
形状特征 平均斑块形状指数 Shape index distribution_Mean SHAPE_MN
加权平均斑块形状指数 Shape index distribution_Area-weighted Mean SHAPE_AM
平均斑块邻近指数 Contiguity index distribution_Mean CONTIG_MN
加权平均斑块邻近指数 Contiguity index distribution_Area-weighted Mean CONTIG_AM
周长面积分维数 Perimeter-area fractal dimension index PAFRAC
平均周长面积比值 Perimeter-area ratio distribution_Mean PARA_MN
加权平均周长面积比值 Perimeter-area ratio distribution_Area-weighted Mean PARA_AM
平均斑块分维数 Patch fractal dimension distribution_Mean FRAC_MN
加权平均斑块分维数 Patch fractal dimension distribution_Area-weighted Mean FRAC_AM
斑块分维数标准差 Patch fractal dimension distribution_Standard Deviation FRAC_SD
斑块分维数数变异系数 Patch fractal dimension distribution_Coeffient of Variation FRAC_CV %
聚散性特征 蔓延度 Contagion index CONTAG %
聚合度 Aggregation index AI %
散布与并列指数 Interspersion and juxtaposition index IJI %
分离度 Landscape division index DIVISION %
破碎度 Splitting index SPLIT
多样性特征 香农多样性指数 Shannon's diversity index SHDI
辛普森多样性指数 Simpson's diversity index SIDI
香农均匀度指数 Shannon's evenness index SHEI

注：本文中各种景观指数按照文献[29]进行英文缩写,一些指数的缩写可能与其他文献有所差别

在ArcGIS 9.3中,采用优势规则,将2010年土地覆被类型矢量图聚合为一系列不同粒度大小的栅格图。景观指数的粒度效应需重点关注第一尺度域的特征[13],因此聚合采用分段间隔,粒度以2.5 m为起点,150 m为终点,2.5~15 m间隔为2.5 m,15~40 m间隔为5 m,40~150 m间隔为10 m,并以原始矢量图为聚合对象[20],从而得到3个流域各22幅栅格类型图。
Fig. 2 The pattem of Land cover in the three watersheds

图2 3个流域的土地覆盖类型图

应用景观格局分析软件Fragstats 4.2,采用八邻法（8-neighborhood rule）定义斑块[25],并统计3个流域22种粒度下的28种常用景观指数值。以粒度大小为横轴,以景观指数值为纵轴制作尺度检测图,表示景观指数的粒度变化响应规律。选择Wu[12,16]所采用的线性、对数和幂函数3种简单函数,分别拟合2.5~150 m与30~150 m粒度范围内的景观指数粒度变化响应规律,得到各种景观指数粒度效应的函数关系（即尺度函数）,拟合效果用相关系数（R2）加以表示,如图3、4所示。
Fig. 3 Split-up predictable responses with scaling relations

图3 分段预测型指数

Fig. 4 Predictable responses with scaling relations

图4 预测响应型指数

3 景观指数粒度效应的高分辨率遥感分析

3.1 景观指数的粒度变化响应规律

综合分析3个流域中28种景观指数在2.5~150 m范围内的粒度变化响应规律,将其分为4类。其中3类（第1、3、4类）与Wu等[12,16]的响应规律相同：第1类（TypeⅠ）为预测响应型,可用简单函数较好地拟合,并且大体上呈单调性变化;第3类（TypeⅢ）为阶梯变化型;第4类（TypeⅣ）为波动变化型。而与Wu等[12,16]所不同的是新增了第2类（TypeⅡ）,可称为分段预测型,这类指数是由原有的第1类中分化出来,随着粒度变粗呈现出2段有规律的响应变化。
第1类指数（TypeⅠ）有11种,可用线性、对数、幂函数3种简单函数进行拟合（R2>0.9）,并且这些指数随着粒度变粗而单调下降,图4表示其中具有代表性的7种指数粒度变化响应曲线;而SIDI、SHEI与SHDI,CONTIG_AM与AI,以及PARA_AM与ED的粒度变化规律相似,因此未在图中一一展示。在不同区域中,同一指数的变化趋势具有相似性,选用相同函数拟合[12,16]。其中,SHDI、SIDI、SHEI随着粒度变粗呈线性下降,而ED、CONTIG_AM、PARA_AM、AI呈对数下降;SHAPE_MN、CONTIG_MN、FRAC_MN、CONTAG呈幂函数下降,约在90 m后则趋于稳定。
LPI、AREA_AM、GYRATE_AM、DIVISION、SPLIT 5种景观指数呈现阶梯状变化（TypeⅢ）且具有明显的尺度域特征。此外,波动变化型指数（TypeⅣ）包括AREA_SD、AREA_CV、SHAPE_AM、FRAC_AM共4种。
与Wu等[12]基于30 m中分辨率遥感影像所揭示的结果不同,图3表明有8种指数（PD、AREA_MN、GYRATE_MN、PAFRAC、PARA_MN、FRAC_SD、FRAC_CV、IJI）随着粒度变粗,指数值呈先上升后下降,或先下降后上升,其尺度转折点分别在5、7.5或10 m粒度处,即由于高分辨率影像的应用,从原有的属于中分辨率数据确定的第1类指数中分化出来一种新的粒度变化响应类别。本研究将其列为第2类（TypeⅡ）,其规律为分段预测型,尤其是尺度转折点之后的粒度变化响应规律与第1类指数一样,呈简单函数关系变化。孟陈等[20]基于2.5 m的高分辨率数据在上海城市土地利用景观研究中发现了PD与AREA_MN呈分段响应规律。不同的是,其2种指数粒度效应的尺度转折点出现在20 m左右,而本文却出现在7.5 m或10 m粒度处,这也是目前常用指数中仅有的分段预测型报道。
为了进一步探究其分段响应规律的原因,本文结合不同粒度大小的栅格类型图发现,3个流域的道路、河流、狭长型耕地、带状植被等线状廊道较窄;在5~15 m范围内随着粒度变粗,线状景观被快速地分割成多个小斑块,而小斑块的融合处于劣势,因而第2类指数可能受线状特征的影响而表现出明显的分段响应特征;朱明等[21]指出,道路对粒度变化敏感是造成城市景观呈现出明显粒度效应的主要原因。此外,Alhamad等[19]基于ETM+影像在地中海沿岸干旱区景观研究中发现,描述核心斑块这一组指数（如核心斑块总面积）的粒度变化响应规律,具有显著的尺度域特征,在尺度转折点的两侧呈可预测性的变化响应规律;但这类指数在其转折点的左侧或右侧随着粒度变粗并不改变数值大小,并且与阶梯变化型指数（TypeⅢ）一样出现明显的尺度断点。然而,本文第2类指数随粒度变粗呈连续变化,且在尺度转折点的左右两侧表现出相反的变化趋势。

3.2 空间分辨率对尺度函数及其下推的影响

第1类可预测响应型的指数,通常可以通过建立其尺度函数进行尺度下推[12]。为了进一步分析空间分辨率对尺度函数的影响,本文选用ED、SHAPE_MN、CONTIG_MN、FRAC_MN、CONTAG、AI、SHDI这7种第1类可预测响应型指数（TypeⅠ）进行分析,并选取线性、对数、幂函数3种简单函数,分别在30~150 m与2.5~150 m范围内进行拟合。结果表明,不同粒度分析范围内景观指数粒度效应关系的拟合参数不尽相同,其中CONTIG_MN、SHAPE_MN、ED和AI的拟合参数差异性较大（表2）。这也说明景观指数基于高分辨率数据所建立的尺度函数与中分辨率所建立的函数,存在较大差别。
Tab. 2 Constants that characterize the TypeⅠ metrics scaling functions between 30~150 m and 2.5~150 m

表2 第1类指数（TypeⅠ）在30~150 m与2.5~150 m内的拟合参数

景观指数 拟合
函数		 粒度分析
范围/m		 坑仔口溪 诗溪 九十九溪
a b R² a b R² a b R²
ED 对数 30~150 -31.889 200.463 1.000 -43.729 260.818 0.996 -44.987 264.643 0.996
2.5~150 -26.682 178.560 0.972 -40.624 248.552 0.982 -42.196 253.778 0.983
SHAPE_MN 幂函数 30~150 2.339 -0.136 0.987 1.577 -0.046 0.960 1.517 -0.045 0.936
2.5~150 2.517 -0.153 0.993 2.129 -0.115 0.909 2.024 -0.111 0.907
CONTIG_MN 幂函数 30~150 2.907 -0.575 0.990 0.750 -0.301 0.994 0.799 -0.326 0.982
2.5~150 1.436 -0.414 0.972 1.165 -0.402 0.985 1.151 -0.409 0.990
FRAC_MN 幂函数 30~150 1.179 -0.028 0.992 1.104 -0.013 0.968 1.102 -0.014 0.969
2.5~150 1.197 -0.031 0.993 1.181 -0.029 0.936 1.176 -0.029 0.938
CONTAG 幂函数 30~150 75.679 -0.090 0.972 69.552 -0.085 0.986 59.985 -0.098 0.965
2.5~150 75.487 -0.088 0.985 71.891 -0.092 0.993 64.261 -0.113 0.989
AI 对数 30~150 -11.019 123.938 0.999 -10.178 117.515 1.000 -8.946 113.148 0.998
2.5~150 -8.147 111.623 0.956 -8.452 110.197 0.977 -7.925 108.905 0.983
SHDI 线性 30~150 -0.001 1.352 0.986 0.000 1.424 0.973 -0.001 1.778 0.997
2.5~150 -0.001 1.345 0.977 0.000 1.422 0.985 -0.001 1.775 0.995
为了进一步探讨中、高分辨率所建立的尺度函数的差异,本文采用表2中分辨率范围（30~150 m）内的拟合函数,用于估计在高分辨率（2.5 m、5 m、10 m）下的指数值,并与实测值进行比较,结果见表3。分析这些结果,可得到如下结论：（1）部分指数（如FRAC_MN、CONTAG、SHDI）对空间分辨率不敏感,由中分辨率范围内所建立的尺度函数估计高分辨率下的指数值,与实测值相差无几,其相对误差最大不超过7.5%;而其他4种指数（ED、SHAPE_MN、CONTIG_MN、AI）的尺度函数与数据的空间分辨率密切相关,由中分辨率所建立的尺度函数预测高分辨率下指数值的误差较大,其相对误差最大可达101.7%。（2）不同流域之间（即不同的格局特征）也存在一定差异,如CONTIG_MN在诗溪与九十九溪的相对误差分别为-32.8%与-28.0%,在坑仔口溪却高达101.7%。
Tab. 3 Comparison of the TypeⅠmetrics between the estimate and actual value in high resolution

表3 第1类指数（TypeⅠ）中尺度下推的估计值与其实测值之间的比较

景观指数 研究区 估计值 实测值 相对误差/（%）
2.5 m 5 m 10 m 2.5 m 5 m 10 m 2.5 m 5 m 10 m
ED 坑仔口溪 171.243 149.140 127.036 136.167 132.554 122.277 25.8 12.5 3.9
诗溪 220.756 190.451 160.147 189.370 180.905 162.781 16.6 5.3 -1.6
九十九溪 223.422 192.239 161.057 192.410 184.338 164.612 16.1 4.3 -2.2
SHAPE_MN 坑仔口溪 2.065 1.879 1.710 2.153 2.041 1.778 -4.1 -7.9 -3.8
诗溪 1.512 1.464 1.419 2.071 1.893 1.632 -27.0 -22.7 -13.1
九十九溪 1.456 1.411 1.368 1.976 1.811 1.559 -26.3 -22.1 -12.3
CONTIG_MN 坑仔口溪 1.716 1.152 0.773 0.851 0.713 0.537 101.7 61.6 44.2
诗溪 0.569 0.462 0.375 0.847 0.679 0.470 -32.8 -32.0 -20.2
九十九溪 0.593 0.473 0.377 0.823 0.650 0.452 -28.0 -27.3 -16.5
FRAC_MN 坑仔口溪 1.149 1.127 1.105 1.161 1.146 1.116 -1.0 -1.6 -1.0
诗溪 1.091 1.081 1.071 1.164 1.145 1.108 -6.3 -5.6 -3.3
九十九溪 1.088 1.077 1.067 1.160 1.139 1.102 -6.2 -5.4 -3.1
CONTAG 坑仔口溪 69.688 65.474 61.514 67.065 65.209 62.410 3.9 0.4 -1.4
诗溪 64.341 60.659 57.189 64.493 62.219 58.945 -0.2 -2.5 -3.0
九十九溪 54.833 51.232 47.868 56.338 53.893 50.406 -2.7 -4.9 -5.0
AI 坑仔口溪 113.841 106.204 98.566 98.305 96.698 93.905 15.8 9.8 5.0
诗溪 108.189 101.134 94.079 97.642 95.492 91.883 10.8 5.9 2.4
九十九溪 104.951 98.750 92.549 97.605 95.409 91.799 7.5 3.5 0.8
SHDI 坑仔口溪 1.350 1.347 1.342 1.337 1.337 1.337 0.9 0.7 0.4
诗溪 1.424 1.424 1.424 1.418 1.418 1.418 0.4 0.4 0.5
九十九溪 1.776 1.773 1.768 1.768 1.768 1.767 0.4 0.3 0.0
此外,当坑仔口溪在30~150 m内采用对数函数拟合,并利用其函数尺度下推至2.5 m时的相对误差只有11.2%,但诗溪、九十九溪的相对误差却分别达到了-49.8%、-58.3%。对于CONTIG_MN而言,坑仔口溪适合于选用对数函数进行尺度下推,而诗溪、九十九溪则宜选用幂函数。
图3结果表明,PD和AREA_MN 2种景观指数表现为分段可预测型,分别在7.5 m和10 m出现尺度转折点。利用中分辨率所建立的尺度函数下推至2.5 m时,ED在3个流域的相对误差可达10%以上。然而,Saura等[23]与Arganaraz等[26]以中低分辨率数据为基础,发现PD、AREA_MN、TE（ED）等指数使用不同粒度范围内拟合的尺度函数来进行尺度下推的误差均较小。因此,景观指数的尺度下推精度可能与原始矢量数据的空间分辨率有密切关系。此外,在景观格局分析中,主题分辨率[18]、遥感分类方法[23]、最小制图单元大小[31]、分类体系[32]等因素也会影响尺度函数的建立及其下推精度,从而使景观格局特征的尺度下推表现出不确定性。

4 结论

本文基于SPOT 5影像系统地分析了闽东南沿海地区3个流域下28种常用景观指数的粒度效应,得出如下结论：
（1）闽东南沿海地区各种景观指数的粒度效应明显,其变化响应规律可以分为预测响应型、分段预测型、阶梯变化型与波动变化型4类。其中,新增的分段预测型,由原来属于中分辨率数据确定的预测响应型分化而来;这类指数的尺度转折点位于5 m、7.5 m或10 m粒度下,说明高分辨率影像对景观指数粒度效应的分类有重要影响。
（2）在预测响应型指数中,香农多样性指数（SHDI）、蔓延度（CONTAG）和平均斑块分维数（FRAC_MN）的尺度函数对空间分辨率大小不敏感;而高分辨率影像对边界密度（ED）、平均斑块形状指数（SHAPE_MN）、平均斑块邻近指数（CONTIG_MN）和聚合度（AI） 4种指数的尺度函数建立影响较大,并且区域差异性也较大,因而由中分辨率数据所建立的尺度函数进行下推,会带来较大 误差。
（3）对于同一指数而言,不同区域可能需要选用不同的尺度函数进行尺度下推。

The authors have declared that no competing interests exist.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
邬建国. 景观生态学:格局、过程、尺度与等级(第二版)[M].北京:高等教育出版社,2007.

[ Wu J G.Landscape ecology:pattern, process, scale and hierarchy(2nd ed.)[M]. Beijing: Higher Education Press, 2007. ]

[2]
刘宇,吕一河,傅伯杰.景观格局——土壤侵蚀研究中景观指数的意义解释及局限性[J].生态学报,2011,31(1):267-275.景观格局分析是景观生态学研究的重要组成部分。景观指数是景观格局分析的有力工具。近年来,景观格局与土壤侵蚀关系的相关研究增多,常规景观格局指数得到应用。但针对土壤侵蚀过程的景观指数意义解释不足,景观指数在刻画景观格局-土壤侵蚀过程关系存在局限。选择了连接性、多样性、边界/斑块密度、形状4个方面的12个常用景观指数,对这些指数在景观格局-土壤侵蚀过程关系研究中的意义进行阐述,对指数应用的局限性及其原因进行了分析。景观数据属性、景观指数本身性质和土壤侵蚀过程的复杂性使得常规景观格局指数在景观格局-土壤侵蚀关系研究中存在不足。这3方面的影响使得常规景观格局指数与土壤侵蚀表征变量之间不存在确定的关系,从而难以通过景观指数来表征景观土壤侵蚀特征。缺乏土壤侵蚀过程基础是常规景观指数在土壤侵蚀研究应用中存在局限的主要原因。因此,构建基于土壤侵蚀过程的景观指数是景观格局-土壤侵蚀关系研究的需要和新的发展方向。

[ Liu Y, Lv Y H, Fu B J.Implication and limitation of landscape metrics in delineating relationship between landscape pattern and soil erosion[J]. Acta Ecologica Sinca ,2011,31(1):267-275. ]

[3]
林炳青,陈兴伟,陈莹,等.流域景观格局变化对洪枯径流影响的SWAT模型模拟分析[J].生态学报,2014,34(7):1772-1780.为了进一步揭示流域景观格局变化的水文效应，以晋江流域为研究区，基于SWAT模型分别模拟1985年和2006年两种景观格局下，2002-2010年气象条件时的日径流过程；应用景观格局分析法和Pearson相关分析法，定量分析了景观格局变化对洪枯径流的影响，探讨了影响机制。研究表明：（1）SWAT模型能够较精确模拟晋江流域的日径流过程；（2）2002-2010年气象条件时，与1985年景观格局相比，在2006年景观格局下晋江流域年平均最大1 d和连续最大5 d径流深分别增加5.46%和4.97%，年平均最小1 d和连续最小7 d径流深分别减少3.79%和2.55%；（3）洪水径流与景观格局相关性最显著，最大1 d和连续最大5 d与林地景观面积呈显著负相关，相关系数分别为-0.764和-0.721；最大1 d与Shannon多样性指数和Shannon均匀度指数呈显著正相关，相关系数分别为0.721和0.736；与最大斑块指数和蔓延度呈显著负相关，相关系数分别为-0.61和-0.596，说明林地景观面积的减少降低了流域对强降水的截留能力；景观类型均衡化分布，连通性降低，降低了水分在流域的内部循环能力，导致洪峰流量的增加；（4）景观格局指数与最小1 d和连续最小7 d的相关性不显著，说明景观格局变化对枯水径流的直接影响较小。

DOI

[ Lin B Q, Chen X W, Chen Y, et al.Simulations and analysis on the effects of landscape pattern change on flood and low flow based on SWAT model. Acta Ecologica Sinica, 2014,34(7):1772-1780. ]

[4]
李娜,闫冬梅,张丽.天津滨海新区——曹妃甸典型地区土地利用演变分析[J].遥感技术与应用,2012,27(2):289-295.<p>根据1984、2000、2006和2009年4期Landsat影像,运用分类后比较法以及土地利用演变的时空分析方法,探讨和分析天津滨海新区和河北曹妃甸地区近30 a的土地利用变化特征和规律,阐明了研究区土地利用变化的区域特点,为土地可持续利用提供有效的决策支持。结果表明：坑塘水面是两个区域最主要的土地利用类型,尤其在曹妃甸地区,坑塘水面占用比例达60%~70%。1984~2009年两地区围海造地面积呈增加趋势,有&ldquo;向海洋要土地&rdquo;的发展趋势。土地利用变化的主要特征为耕地减少和建设用地、坑塘水面增加,2000年以后土地利用变化趋势尤为显著。滨海新区建设用地重心有西南方向移动的趋势,其他土地利用类型重心均有向东向北迁移的趋势,曹妃甸地区各土地类型均有明显的南移趋势。</p>

[ Li N, Yan D M, Zhang L.Analysis of the evolution of land use in Tianjin Binhai new area and Caofeidian region[J]. Remote Sensing Technology and Application, 2012,27(2):289-295. ]

[5]
Diaz-Varela E R, Marey-Perez M F, Rigueiro-Rodriguez A, et al. Landscape metrics for characterization of forest landscapes in a sustainable management framework: potential application and prevention of misuse[J]. Annals of Forest Science, 2009,66(3):1-10.

[6]
Schindler S, Poirazidis K, Wrbka T.Towards a core set of landscape metrics for biodiversity assessments: a case study from Dadia National Park, Greece[J]. Ecological Indicators, 2008,8(5):502-514.<h2 class="secHeading" id="section_abstract">Abstract</h2><p id="">Spatial heterogeneity has an important influence on a wide range of ecological patterns and processes, and many landscape metrics in GIS environment are used to facilitate the investigation of the relation between landscape structure and biodiversity. Data reduction analyses have been applied to tackle the problem of highly correlated indices, but valid landscape predictors for fine scale Mediterranean forest-mosaics are still missing. Therefore, we analyzed the landscape structure of Dadia National Park, Greece, a Mediterranean forest landscape of high biodiversity, characterized by pine, oak and mixed woodland. By distinguishing nine land cover classes, 119 variables were computed and factor analysis was applied to detect the statistical dimensions of landscape structure and to define a core set of representative metrics. At landscape level, diversity of habitats, fragmentation and patch shape and at class level dominance of mixed forest and the gradient from one pure forest type to another turned out to be the crucial factors across three different scales. Mapping the encountered dimensions and the representative metrics, we detected that the pattern of landscape structure in Dadia National Park was related to dominating habitat types, land use, and level of protection. The evaluated set of metrics will be useful in establishing a landscape monitoring program, to detect the local drivers of biodiversity, and to improve management decisions in Dadia NP and similar mosaic-landscapes.</p>

DOI

[7]
Uuemaa E, Roosaare J, Mander U.Landscape metrics as indicators of river water quality at catchment scale[J]. Nordic Hydrology, 2007,38(2):125-138.

[8]
Uuemaa E, Antrop M, Roosaare J, Marja R, et al.Landscape metrics and indices: an overview of their use in landscape research[J]. Living Reviews in Landscape Research, 2009,3(1):1-28.

[9]
Turner M G, O'Neill R V, Gardner R H, et al. Effects of changing spatial scale on the analysis of landscape pattern[J]. Landscape Ecology, 1989,3(3-4):153-162.

[10]
Simova P, Gdulova K.Landscape indices behavior: a review of scale effects[J]. Applied Geography, 2012,34: 385-394.

[11]
Baldwin D J B, Weaver K, Schnekenburger F, et al. Sensitivity of landscape pattern indices to input data characteristics on real landscapes: Implications for their use in natural disturbance emulation[J]. Landscape Ecology, 2004,19(3):255-271.<a name="Abs1"></a>Resource management strategies have begun to adopt natural landscape disturbance emulation as a means of minimizing risk to ecosystem integrity. Detailed understanding of the disturbance regime and the associated spatial landscape patterns are required to provide a <img src="/content/K408P35N9V63M545/xxlarge8220.gif" alt="ldquo" align="MIDDLE" border="0">natural<img src="/content/K408P35N9V63M545/xxlarge8221.gif" alt="rdquo" align="MIDDLE" border="0"> baseline for comparison with the results of emulation strategies. Landscape pattern indices provide a useful tool to quantify spatial pattern for developing these strategies and evaluating their success. Despite an abundance of indices and tools to calculate these, practical knowledge of interpretation is rare. Quantifying changes in landscape pattern indices and the meaning of these changes is confounded by index sensitivity to input data characteristics such as spatial extent, spatial resolution, and thematic resolution. Sensitivity has been examined for simulated landscapes but rarely using real data for large areas as real landscapes are more difficult to manipulate systematically than simulated data. While simulated data offer a control, they do not provide an accurate portrayal of reality for practical applications. Our goal was to test the sensitivity of a suite of landscape pattern indices useful for disturbance emulation strategy development and evaluation to spatial extent, spatial resolution, and thematic resolution using current land cover data for a case study of the managed forest of Ontario, Canada. We also examined how sensitivity varies spatially across the study area. We used Landsat TM-based land cover data (> 45.5 million ha), controlling spatial extent (2,500 to 2,560,000 ha), spatial resolution (1 to 16 ha), and thematic resolution (2 to 26 classes). For each index we tested a hypothesis of insensitivity to changes in each input data characteristic using a combination of ANOVA and regression and compared our results with previous studies. Of the 18 indices studied, significant (p< 0.01)="" effects="" were="" found="" for="" 17="" indices="" with="" changes="" in="" spatial="" extent,="" 13="" indices="" with="" changes="" in="" spatial="" resolution="" and="" 18="" indices="" with="" changes="" in="" thematic="" resolution.="" a="" significant="" (p="">< 0.01)="" linear="" trend="" accounted="" for="" the="" majority="" of="" the="" variance="" for="" all="" of="" the="" significant="" relationships="" identified.="" most="" of="" the="" mean="" index="" responses="" were="" consistent="" with="" those="" interpreted="" from="" previous="" studies="" of="" simulated="" and="" real="" landscapes;="" however,="" sensitivity="" varied="" greatly="" among="" indices="" and="" over="" space.="" we="" suggest="" that="" variation="" in="" sensitivity="" to="" input="" data="" characteristics="" among="" indices="" and="" over="" space="" must="" be="" explicitly="" incorporated="" in="" the="" design="" of="" future="" natural="" disturbance="" emulation="">

DOI

[12]
Wu J G, Shen W J, Sun W Z, et al.Empirical patterns of the effects of changing scale on landscape metrics[J]. Landscape Ecology, 2002,17:761-782.lt;a name="Abs1"></a>While ecologists are well aware that spatial heterogeneity is scale-dependent, a general understanding of scaling relationships of spatial pattern is still lacking. One way to improve this understanding is to systematically examine how pattern indices change with scale in real landscapes of different kinds. This study, therefore, was designed to investigate how a suite of commonly used landscape metrics respond to changing grain size, extent, and the direction of analysis (or sampling) using several different landscapes in North America. Our results showed that the responses of the 19 landscape metrics fell into three general categories: Type I metrics showed predictable responses with changing scale, and their scaling relations could be represented by simple scaling equations (linear, power-law, or logarithmic functions); Type II metrics exhibited staircase-like responses that were less predictable; and Type III metrics behaved erratically in response to changing scale, suggesting no consistent scaling relations. In general, the effect of changing grain size was more predictable than that of changing extent. Type I metrics represent those landscape features that can be readily and accurately extrapolated or interpolated across spatial scales, whereas Type II and III metrics represent those that require more explicit consideration of idiosyncratic details for successful scaling. To adequately quantify spatial heterogeneity, the metric-scalograms (the response curves of metrics to changing scale), instead of single-scale measures, seem necessary.

DOI

[13]
赵文武,傅伯杰,陈利顶.景观指数的粒度变化效应[J].第四纪研究,2003,23(3):326-333. [ Zhao W W, Fu B J, Chen L D. The effects of grain change on landscape indices. Quaternary Sciences, 2003,23(3):326-333. ]本文以延河流域的1:250000和1:500000土地利用图为对象,以景观格局分析程序Fragstats3.3为分析工具,探讨了不同比例尺条件下景观指数随粒度增加的变化特征.研究结果表明,随着粒度值由25m到400m的逐渐增加,除斑块丰富度外的景观指数均具有明显的尺度效应,其中聚集度和集合度没有尺度转折点,其他指数具有明显或不明显的尺度转折点.对比分析1:250000和1:500000土地利用格局指数的计算结果可以发现:1)尺度转折点不是一个值,而是一个相对较小的区间;2)尺度转折点与研究图件的比例尺有关,比例尺越大,所发生的第一次尺度转折点的粒度就越小;3)第一尺度域是选择适宜粒度的较好取值范围.对延河流域1:250000土地利用图进行景观指数计算的适宜粒度范围是70～90m,1:500000土地利用图的适宜粒度范围是90～120m.

DOI

[14]
Li X Z, He H C, Bu R C, et al.The adequacy of different landscape metrics for various landscape patterns[J]. Pattern Recognition, 2005,38:2626-2638.The behavior of several landscape pattern metrics were tested against various pattern scenarios generated by neutral landscape models, including number of classes, scale-map extent, scale-resolution, class proportion, aggregation level-RULE, and aggregation level-SimMap. The results demonstrate that most of the metrics are sensitive to certain pattern scenarios, yet are not sensitive to others; therefore, none of them is appropriate for all aspects of a landscape pattern. Despite these limitations, some of these metrics are recommended for future use, which include total number of patches, average patch size, total edge density, double-logged fractal, contagion, and aggregation index. Special attention should be paid to the relationships between metric values and ecological processes rather than the numbers themselves. (c) 2005 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.

DOI

[15]
Benson B J, MacKenzie M D. Effects of sensor spatial resolution on landscape structure parameters[J]. Landscape Ecology, 1995,10(2):113-120.

[16]
Wu J G.Effects of changing scale on landscape pattern analysis: scaling relations[J]. Landscape Ecology, 2004,19(2):125-138.lt;a name="Abs1"></a>Landscape pattern is spatially correlated and scale-dependent. Thus, understanding landscape structure and functioning requires multiscale information, and scaling functions are the most precise and concise way of quantifying multiscale characteristics explicitly. The major objective of this study was to explore if there are any scaling relations for landscape pattern when it is measured over a range of scales (grain size and extent). The results showed that the responses of landscape metrics to changing scale fell into two categories when computed at the class level (i.e., for individual land cover types): simple scaling functions and unpredictable behavior. Similarly, three categories were found at the landscape level, with the third being staircase pattern, in a previous study when all land cover types were combined together. In general, scaling relations were more variable at the class level than at the landscape level, and more consistent and predictable with changing grain size than with changing extent at both levels. Considering that the landscapes under study were quite diverse in terms of both composition and configuration, these results seem robust. This study highlights the need for multiscale analysis in order to adequately characterize and monitor landscape heterogeneity, and provides insights into the scaling of landscape patterns.

DOI

[17]
Uuemaa E, Roosaare J, Mander U.Scale Dependence of landscape metrics and their indicatory value for nutrient and organic matter losses from catchments[J]. Ecological Inicators, 2005,5:350-369.We investigated scale dependence of landscape metrics and the relationship between land use parameters and FRAGSTATS-based landscape metrics (edge density (ED), patch density (PD), mean shape index (SHAPE_MN), mean euclidean nearest neighbor index (ENN_MN), contagion (CONTAG), patch richness density (PRD), and Shannon's diversity index (SHDI)) and nutrient/organic-matter-based water quality indicators (BODandvalues, total-N and total-P concentrations in water) in 24 catchments with various land use patterns in Estonia. We used the Basic Map of Estonia (1:10,000), the Base Map of Estonia (1:50,000) and the CORINE Land Cover Map (1:100,000). In scale analysis, we calculated landscape metrics on artificial and real landscapes. Scale analysis showed that responses of landscape metrics to changing grain size vary among landscapes and metrics. Analysis of artificial landscapes showed that mean euclidean nearest neighbor distance and contagion are directly dependent on grain size and should therefore be used carefully. When finding relationships between landscape metrics and water quality indicators, significant differences between the relationships derived from the Base Map and the CORINE Land Cover Map were found. In the case of the Base Map, landscape metrics correlated strongly with land use and showed no expected relationships with water quality data. This underlines the importance of land use classification in such kind of analysis. Correlation between the landscape metrics calculated on the basis of the CORINE Land Cover Map and water quality data was stronger than in the case of the Base Map. Thevalue significantly correlated with all land use types. For instance, thevalues are higher when fens and natural areas form a higher proportion of the catchments鈥 land use. Except for the BODvalue, all the water quality indicators showed significant correlation with urban land use proportions. Strong relationship between the patch density and thevalue is most likely caused by the fact that both parameters were significantly correlated with the proportion of natural areas. As the landscape metrics depend on pixel size, topographic scale, and land use classification, and as the effect of land use on water quality in catchments is the most significant of the factors, it was impossible to separate the influence of land use pattern from the influence of FRAGSTATS-based landscape metrics.

DOI

[18]
Buyantuyev A, Wu J G, Gries C.Multiscale analysis of the urbanization pattern of Phoenix Metropolitan landscape of USA: Time, space, and thematic resolution[J]. Landscape and Urban Planning, 2010,94:206-217.

[19]
Alhamad M N, Alrababah M A, Feagin R A, et al.Mediterranean drylands: The effect of grain size and domain of scale on landscape metrics[J]. Ecological Indicators, 2011,11:611-621.

[20]
孟陈,李俊祥,朱颖,等.粒度变化对上海市景观格局分析的影响[J].生态学杂志,2007,26(7):1138-1142.采用1989年上海市1∶50 000彩红外航空遥感影像解译的土地利用数据，系统分析了粒度变化对中心城区、郊区和城郊结合部3类景观格局分析的影响。结果表明：景观格局指数对粒度具有尺度依赖性，随粒度的变粗，格局指数的变化可归为单调下降、阶梯上升或单峰变化以及不规则变化3类;景观格局指数随粒度变化的曲线存在多个拐点，并对应于合适的景观格局分析尺度，上海城市景观格局分析的合适尺度为10～20 m;不同城市景观类型对景观格局指数的粒度效应具有较大影响，景观格局指数在城郊结合部表现出与城区和郊区不同的变化趋势。

[ Meng C, Li J X, Zhu Y, et al.Effects of changing grain size on urban landscape pattern analysis of Shanghai[J]. Chinese Journal of Ecology, 2007,26(7):1138-1142. ]

[21]
朱明,濮励杰,李建龙.遥感影像空间分辨率及粒度变化对城市景观格局分析的影响[J].生态学报,2008,28(6):2753-2763.尺度是景观生态学和遥感领域的重要问题,已有研究主要考虑景观的粒度效应,很少涉及遥感影像空间分辨率对景观格局,尤其是对城市景观格局的影响。基于2000年和2002年分别由ETM+和IRS-PAN影像解译得到的土地利用图,从景观和类型两个层次分析了不同粒度下上海市一城市化样带的景观格局。结果表明,城市景观有明显的尺度效应,空间分辨率和粒度变化都会影响城市景观格局,而道路等线性廊道对尺度变化的敏感则是造成这一影响的主要原因。斑块密度、边界密度、平均斑块分维度、景观聚集度和最大斑块指数的粒度效应比较明显,而景观比例、Shannon多样性指数对粒度变化并不敏感。高分辨率影像适用于城市景观格局,尤其是破碎度的分析,其用于分析的合适粒度是5~10m。

[ Zhu M, Pu L J, Li J L.Effects of varied remote sensor spatial resolution and grain size on urban landscape pattern analysis[J]. Acta Ecologica Sinica, 2008,28(6):2753-2763. ]

[22]
游丽平,林广发,杨陈照,等.景观指数的空间尺度效应分析——以厦门岛土地利用格局为例[J].地球信息科学,2008,10(1):74-79.

[ You L P, Lin G F, Yang C Z, et al.The effects of spatial scales on landscape indices: a case study of landuse pattern of Xiamen Island[J]. Geo-information Science, 2008,10(1):74-79. ]

[23]
Saura S, Castro S.Scaling functions for landscape pattern metrics derived from remotely sensed data: Are their subpixel estimates really accurate?[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2007,62(3):201-216.<h2 class="secHeading" id="section_abstract">Abstract</h2><p id="">One of the most rapidly growing applications of remotely sensed data is the derivation of landscape pattern metrics for the assessment of land cover condition and landscape change dynamics. The availability of a wide variety of sensors allows for characterisation of land cover at multiple spatial scales, and increases the need for practical scaling techniques that permit the comparison of pattern estimates across different spatial resolutions. Previous research has reported on scaling functions describing the variations of different landscape pattern metrics with spatial resolution; this may be particularly useful in downscaling spatial pattern characteristics, but no quantitative results or independent validation have been reported yet in this respect. We analysed a wide set of landscape data derived from remotely sensed images covering different study areas, sensor spatial resolutions, and classification approaches (pixel-based and object-based), which were aggregated to coarser resolutions through majority filters. We considered eight landscape pattern metrics for which predictable scaling functions have been reported, and compared the subpixel estimates provided by those scaling functions (when fitted to the metric values for different ranges of spatial resolution above the pixel level) with the true value of the metric at the subpixel resolution. We found that for metrics like mean patch size, landscape shape index or edge length, quite accurate subpixel estimates were achieved in all the datasets, even for relatively large downscaling factors. However, the opposite was the case for several of the metrics for which a predictable scaling behaviour had been previously described. The most accurate subpixel estimates were obtained when only a narrow range of spatial resolutions (closest to the subpixel resolution) was used to fit the scaling function, suggesting that the scaling functions are not fully scale invariant. We also found that the performance of available scaling functions is much lower in object-based data (in comparison with per-pixel classified data) for ranges of spatial resolution below the characteristic minimum mapping unit of the interpreted or segmented image. We conclude that scaling functions may be useful and reasonably accurate for estimating pattern metrics at the subpixel level, but only if the specific scaling recommendations and limitations reported in this study are taken into account.</p>

DOI

[24]
张娜. 生态学中的尺度问题——尺度上推[J].生态学报,2007,27(10):4252-4265.尺度推绎是生态学理论和应用的核心。如何在一个异质景观中进行尺度推绎仍然是一个悬而未决的科学难题，是对当今生态学家在全球变化背景下研究环境问题的重大挑战。就目前的研究，一般可分为四大类尺度推绎途径：空间分析法（如分维分析法和小波分析法）、基于相似性的尺度上推方法、基于局域动态模型的尺度上推方法、随机（模型）法。基于相似性的尺度上推方法来源于生物学上的异量关联，可将其思想延伸至空间上，研究物种丰富度、自然河网、地形特征、生态学格局或过程变量和景观指数等。基于局域动态模型的尺度上推方法需要首先确定是否进行跨尺度推绎，以及是否考虑空间单元之间的水平相互作用和反馈，然后再应用具体的方法或途径，如简单聚合法、有效值外推法、直接外推法、期望值外推、显式积分法和空间相互作用模拟法等。随机（模型）法以其它尺度上推方法为基础，根据研究的是单个景观，还是多个景观，采用不同的途径。理解、定量和降低尺度推绎结果的不确定性已经变得越来越重要，但相关研究仍然极少。以上所有有关尺度推绎的方法、途径和结果分析共同构成了尺度推绎的概念框架。

[ Zhang N.Scale issues in ecology:upscaling[J]. Acta Ecologica Sinica, 2007,27(10):4252-4265. ]

[25]
Arganaraz J P, Entraigas I.Scaling functions evaluation for estimation of landscape metrics at higher resolutions[J]. Ecological Informatics, 2014,22:1-12.

[26]
高燕. 云蒙湖流域土地利用变化的尺度效应研究[D].济南:山东师范大学, 2014.

[ Gao Y.Scale effect research on land use change in Yunmeng Lake Watershed[D]. Jinan: Shandong Normal University, 2014. ]

[27]
郑达贤,陈佳源,吴幼恭,等.闽东南沿海一些地理环境特征的再认识[J].地理学报,1991,46(4):405-414.本文认为,闽东南沿海的气候为 南亚热带半湿润气候;土壤属半干润老成土;原生植被不可能是湿润季雨林,而可能是具有旱生特征的常绿硬叶阔叶和针叶混交林或森林灌丛;在土地利用上以旱作 为特征,应把该区与台湾海峡一起作为一个自然区划单元划出,建议称为闽东南沿海与台湾海峡半岛岛屿州。

DOI

[ Zheng D X, Chen J Y, Wu Y G, et al.Reconsideration on characters of geographical environmrnt in the South-Eastern Coast Region of Fujian[J]. Acta Geographical Sinica. 1991,46(4):405-414. ]

[28]
黄彩霞,李小梅,沙晋明.主题分辨率对NDVI空间格局的影响[J].生态学报,2011,31(18):5414-5420.景观格局与尺度的相互关系一直是景观生态学的研究重点。景观生态学中的尺度内涵包括时间、空间和分析（观察）等。近年来景观格局与空间、时间尺度关系受到广泛关注,而分析尺度对格局的影响探讨较少。地理空间信息的主题分辨率反映了地理空间信息的制图细节,即专题图的分类数。以NDVI（Normalized Difference Vegetation Index）为专题图内容,选择福州行政区、福州城区和永泰县为研究区域,通过改变NDVI分类数,研究主题分辨率对空间格局的影响。NDVI格局指数包括：斑块数量（NP）、最大斑块占景观面积比例（LPI）、斑块平均大小（AREA-MN）、面积加权平均形状指数（SHAPE-AM）、面积加权平均分维数（FRAC-AM）、蔓延度指数（CONTAG）、散步与并列指数（IJI）、香农多样性指数（SHDI）。结果表明格局指数对主题分辨率（分类数）的连续响应特征如下：初始条件敏感区（分类数2—4类）、敏感响应区（分类数为4—8类）、适合分析尺度区（分类数8—12类）、不敏感区（分类数≥12类）；格局指数与主题分辨率的关系有对数增长、线性增长和幂函数下降等。分析尺度对NDVI空间格局影响的本质原因是相对应的生态系统等级结构的存在

[ Huang C X, Li X M, Sha J M.The impact of thematic resolution on NDVI spatial pattern. Acta Ecologica Sinica, 2011,31(18):5414-5420. ]

[29]
McGarigal K. Fragstats: Spatial pattern analysis program for categorical maps[EB/OL]. www.umass.edu/landeco/2014.

[30]
布仁仓,胡远满,常禹,等.景观指数之间的相关分析[J].生态学报,2005,25(10):2764-2775.应用辽宁省1997~1998年的TM 5影像数据,编制了景观类型图,以78个县市区为单位,分割成78个景观,共计算39个景观格局指数,对它们进行了相关分析。总面积是最基本的景观指数,它决定景观总边界长度、斑块数、类型密度等基本指数,同时与多个指数有显著的相关关系(相关系数绝对值大于0.75)。形状指数的独立性强,极少数指数与其它指数有显著的相关关系;多样性指数和蔓延度指数之间信息重复量最多,都表示景观的异质性,但多样性指数以面积百分比表示景观异质性,而蔓延度指数以类型之间相邻边界的百分比表示景观异质性。研究发现,如果两个指数之间存在显著的相关关系,而由它们两个构成的指数与它们之间没有显著的相关关系。如果指数平均值之间存在显著的相关关系,则它们的变异系数之间不存在显著的相关关系。景观指数间的相关系数不仅与景观格局本身有关,还与空间尺度,分类系统、计算公式及其参数、计算单元和生态学意义关系密切。指数之间影响因子的相同之处越多,它们之间存在显著相关关系的概率越大。

DOI

[ Bu R C, Hu Y M, Chang Y, et al.A correlation analysis on landscape metrics[J]. Acta Ecologica Sinica, 2005,25(10):2764-2775. ]

[31]
Saura S.Effects of minimum mapping unit on land cover data spatial configuration and composition[J]. International Journal of Remote Sensing, 2002,23(22):4853-4880.

[32]
Comber A, Fisher P, Wadsworth R.You know what land cover is but does anyone else? an investigation into semantic and ontological confusion[J]. International Journal of Remote Sensing, 2005,26(1):223-228.

Options
Outlines

/

〈 〉