Identification of Forest Vegetation Using Vegetation Indices

Abstract

Spectral feature of forest vegetation with remote sensing techniques is the research topic all over the world, because forest plays an important role in human beings' living environment. Research on vegetation classification with vegetation index is still very little recently. This paper proposes a method of identifying forest types based on vegetation indices, because the contrast of absorbing red waveband with reflecting near-infrared waveband strongly for different vegetation types is recognized as the theoretic basis of vegetation analysis with remote sensing. Vegetation index is highly related to leaf area index, absorbed photosynthetically active radiation and vegetation cover. Vegetation index reflects photosynthesis intensity of plants and manifests different forest types. According to reflectance data of forest canopy and soil line equation NIR=1.506R+0.0076 in Jingyuetan, Changchun of China, many vegetation indices are calculated and analyzed. The result shows that the relationships between vegetation indices and forest types are that perpendicular vegetation index (PVI) identifies broadleaf forest and coniferous forest the most easily; the next is transformed soil-adjusted vegetation index (TSVI) and modified soil-adjusted vegetation index (MSVI), but their calculation is complex. Ratio vegetation index (RVI) values of different coniferous forest vary obviously, so RVI can classify conifers. Therefore, the combination of PVI and RVI is evaluated to classify different vegetation types.     

基于决策树—山体阴影模型的植被信息提取研究

植被是土地覆被分类的重要内容,植被分类对研究流域生态具有重要参考价值。鄂西犟河流域是南水北调中线工程的重要水源地,该区以中低山和丘陵为主,地形起伏剧烈,导致遥感影像存在大量阴影,制约了植被分类的精度。基于LandSAT OLI影像,使用ArcGIS的Hillshade模块,输入DEM数据和传感器具体参数,计算得到影像成像时刻阴影的准确分布;统计野外采集地物样本点在MNDWI、NDVI和RVI等指数上的差异;结合决策树分类法,分别设定阴影和非阴影下6类样本的阈值进行分类。结果表明:(1)该方法总分类精度能够达到92.93%,Kappa系数为0.912;(2)阴影和非阴影区植被具备明显的同物异谱和异物同谱特征,表现为阳面的植被指数整体高于阴面;3类林地的RVI值由高往低依次为:灌木,混交林和针叶林。(3)传统经验模型在不同纬度的适用性不同,无法精确提取阴影的范围,降低了分类精度;决策树—山体阴影模型作为一种智能分类方法,能够还原Landsat OLI影像准确的阴影分布,提高山地和丘陵等阴影面积大、形状复杂区域的植被分类精度。     

基于C5.0决策树和时序HJ-1A/B CCD数据的神农架林区植被分类

我国神农架林区海拔高、气候复杂,森林类型多样,结构破碎,森林遥感分类难度较大。将2013年时间序列HJ-1A/B CCD遥感影像作为数据源,计算出植被指数(NDVI、DVI、RVI)和主成分第一分量(PC1),使用DEM数据生成地形因子(高程、坡度、坡向),构建植被分类时序因子集。运用C5.0决策树分类法将神农架林区植被细分为七类:针叶林;针阔混交林;落叶阔叶林;常绿和落叶阔叶混交林;常绿阔叶林;灌丛和草甸。结果表明:该方法的总体精度为72.7%,Kappa系数为0.67;在6~8月,针叶林、草甸和灌丛的植被指数明显低于常绿阔叶林、常绿和落叶阔叶混交林、落叶阔叶林和针阔混交林,对分类的贡献较大,称为植被分类的"窗口期"。PC1、NDVI和高程因子对神农架林地的区分度较高,而坡度、坡向和RVI因子对分类帮助不大。作为一种智能分类方法,C5.0决策树分类方法应用于30m分辨率的时间序列HJ-1A/B CCD数据,能够将地貌复杂的神农架林区植被分为七类,提高了类别精度,具有更高的应用价值。     

基于随机森林和Sentinel-2影像数据的低山丘陵区土地利用分类——以重庆市江津区李市镇为例

精准的土地利用信息是土地资源监测和管理的基础。为提高低山丘陵区域的土地利用分类精度,选取重庆市江津区李市镇为研究案例,基于随机森林方法,以Sentinel-2影像数据和地形因子为数据源,提取3种变量（传统遥感数据,红边遥感数据和地形因子）,合计23个特征指标,构建3个具有不同输入变量的组合模型,以提取研究区土地利用信息,分析变量的重要性。结果表明：（1）传统遥感数据模型中顺序添加红边遥感数据和地形因子,总体分类精度分别为86.54%,87.19%,88.61%;Kappa系数分别为 0.800 9,0.810 2,0.831 4;（2）对模型精度有重要影响的特征指标依次是波段B2(Blue),B4(Red),B3(Green),改进归一化差异水体指数(MNDWI)和波段B5(Vegetation Red Edge 1);(3）基于随机森林的遥感数据和地形因子的组合方法,是获取研究区高精度土地利用信息的一种有效手段。研究成果可以为地形复杂区域的土地利用分类提供参考。     

唐家河保护区主要植被类型景观格局的梯度分析

在遥感图像基础上,利用GIS技术从景观指数方面定量分析了唐家河自然保护区主要植被类型在东西、南北和西北至东南3个方向上的梯度变化。结果表明：次生落叶阔叶林、常绿落叶阔叶混交林、针阔叶混交林与针叶林的梯度变化明显,并且在各方向上具有不同的变化特征。其中,常绿落叶阔叶混交林从西北至东南方向的梯度变化最为复杂, 斑块数量与面积分别呈“升-降-升-降”与“升-降-升”的波动变化趋势,而边界密度与平均最近距离呈先升后降的变化趋势,两端破碎度高但连接性好,中部相反。针叶林从北至南的梯度变化最为简单,斑块面积减少,破碎度与复杂度降低,南北两端分布较多,中部较少。唐家河自然保护区植被景观格局在不同方向上的梯度变化研究为地震后该地区的植被保护与管理提供重要指导意义。     

庐山不同森林植被类型土壤特性及其健康评价

土壤作为森林生态系统的一个重要因子,评价森林土壤健康状况对森林健康的维护经营以及森林系统功能的发挥具有重要意义。在系统调查和分析庐山8种森林植被类型土壤特性的基础上,评价指标分别从物种多样性以及不同的森林土壤特性中进行筛选,包括物种多样性系数、枯落物层厚度、腐殖质层厚度、土层厚度、容重、粘粒含量、有机质、p H值、阳离子交换量、全氮、水解氮、有效磷、速效钾、磷酸酶活性等指标,基于SPSS19.0软件对所获得数据进行差异性检验和相关分析,确定各项指标的权重,应用合适的土壤健康评分函数,将测得的指标值转换为对应指标的分值,最后通过加权综合法,计算其土壤健康指数,并对不同森林植被类型土壤健康状况进行评价。结果表明,8种森林植被类型下最终的土壤健康指数大小排序为:针阔混交林(0.78)常-落混交林(0.72)灌丛(0.69)常绿阔叶林(0.67)落叶阔叶林(0.64)竹林(0.59)马尾松林(0.53)黄山松林(0.46)。     

唐家河保护区主要植被类型景观格局的梯度分析

在遥感图像基础上，利用GIS技术从景观指数方面定量分析了唐家河自然保护区主要植被类型在东西、南北和西北至东南3个方向上的梯度变化。结果表明：次生落叶阔叶林、常绿落叶阔叶混交林、针阔叶混交林与针叶林的梯度变化明显，并且在各方向上具有不同的变化特征。其中，常绿落叶阔叶混交林从西北至东南方向的梯度变化最为复杂， 斑块数量与面积分别呈“升—降—升—降”与“升—降—升”的波动变化趋势，而边界密度与平均最近距离呈先升后降的变化趋势，两端破碎度高但连接性好，中部相反。针叶林从北至南的梯度变化最为简单，斑块面积减少，破碎度与复杂度降低，南北两端分布较多，中部较少。唐家河自然保护区植被景观格局在不同方向上的梯度变化研究为地震后该地区的植被保护与管理提供重要指导意义。     

基于MODIS数据的安徽省植被水分利用效率时空变化及影响因素

水分利用效率是衡量生态系统碳水循环耦合程度的重要指标。基于MODIS数据、土地覆盖类型数据和气象数据,估算安徽省植被水分利用效率(WUE）,采用趋势分析法和相关分析法对安徽省2000~2014年植被WUE的时空格局、变化趋势及影响因素进行研究。研究表明:（1）不同植被类型的WUE年均值差异明显,常绿阔叶林和常绿针叶林的WUE均值较高,分别达到1.66和1.69 gC?mm-1?m-2,而耕地的年均WUE最低,各植被类型的年均WUE按照“常绿针叶林>常绿阔叶林>灌木>草地>落叶阔叶林>针阔混交林>耕地”的顺序递减。植被年均WUE具有较强的空间分异性规律,整体上呈现南北高中间低的趋势,植被WUE的高值区主要分布在大别山区和皖南山区,分布范围与常绿针叶林、常绿阔叶林的分布范围基本一致。（2）安徽省2000~2014年植被WUE年内变化呈现出“增加-减小-增加-减小”的M状“双峰型”趋势,具有明显的季节差异,呈现出春季＞秋季＞夏季＞冬季的特征,各季节植被WUE的均值分别占植被WUE的32.58%、24.91%、29.27%、13.24%。（3）安徽省植被WUE动态变化受到降水影响显著的区域占比3.88%;气温显著影响的区域占比2.19%;降水显著影响的地区主要分布在林地范围内,温度显著影响的地区则位于耕地范围内,降水和气温综合显著影响所占面积最小,为0.11%;而植被WUE受气温和降水影响均不显著占比为93.82%;整体上,安徽省大部分地区的植被WUE变化主要受非气候因素影响。     

Application of remote sensing techniques toward the role of traditional water bodies with respect to vegetation conditions

Sri Lanka being an agrarian country, the role of water is important for agricultural production. In Sri Lanka, various tank cascade systems, earthen dams and distribution canals have been accepted as few of the most complex ancient traditional water systems of the world. Rainfall, surface water, groundwater and runoff are linked with each other, they have close interactions to land cover classes such as forests and agriculture. The monitoring of vegetation conditions can show subsurface manifestations of groundwater. In this study, an effort to understand the role of traditional water reservoirs and groundwater recharge was made using remote sensing techniques. We have analyzed various vegetation indices such as Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI-2), Soil-Adjusted Vegetation Index (SAVI), tasselled cap transformation analysis (TCA brightness, greenness and wetness) and their relations with the existence of soil, vegetation and water. Result shows that EVI, SAVI, and TCA-based Greenness Index indicates good relationship with the vegetation conditions as compared to other indices. Therefore, these indices could play a crucial role in depicting the interaction between soil, vegetation, and water. However, multi-temporal observations can provide significant results about these interactions more accurately.     

基于ETM影像的植被指数在土地利用中的特征——以雅安市雨城区为例

利用LandSAT7号卫星的ETM数据、土地利用现状数据，在GIS空间分析工具与ERDAS空间建模语言的支持下，提取ETM影像的RVI、NDVI和MSAVI等3类植被指数，在ArcGIS中应用样带统计分析工具，统计并分析3类植被指数在各土地利用类型中的分布与特征。结果表明：RVI与NDVI、MSAVI在各土地利用中的特征相似，都能较好的反映土地利用类型的覆盖程度，但各植被指数的表现形式有异，即RVI值高的，NDVI、MSAVI值较低，代表的植被覆盖度也较低，反之亦然；3类植被指数能较好的区分出农田用地、非植被覆盖区、植被覆盖区，而林地与灌木林、早地与园地则不易区分。     

Application of Remote Sensing Detection and GIS in Analysis of Vegetation Pattern Dynamics in the Yellow River Delta

Regional vegetation pattern dynamics has a great impact on ecosystem and climate change. Remote sensing data and geographical information system （GIS） analysis were widely used in the detection of vegetation pattern dynamics. In this study, the Yellow River Delta was selected as the study area. By using 1986, 1993, 1996, 1999 and 2005 remote sensing data as basic information resource, with the support of GIS, a wetland vegetation spatial information dataset was built up. Through selecting the landscape metrics such as class area （CA）, class percent of landscape （PL）, number of patch （NP）, largest patch index （LPI） and mean patch size （MPS） etc., the dynamics of vcgetation pattern was analyzed. The result showed that the change of vegetation pattern is significant from 1986 to 2005. From 1986-1999, the area of the vegetation, the percent of vegetation, LPI and MPS decreased, the NP increased, the vegetation pattern tends to be fragmental. The decrease in vegetation area may well be explained by the fact of the nature environment evolution （Climate change and decrease in Yellow River runoff） and the increase in the population in the Yellow River Delta. However, from 1999 2005, the area of the vegetation, the percent of vegetation, LPI and MPS increased, while the NP decreased. This trend of restoration may be due to the implementation of water resources regulation for the Yellow River Delta since 1999.     

Application of Remote Sensing Detection and GIS in Analysis of Vegetation Pattern Dynamics in the Yellow River Delta

Abstract

Regional vegetation pattern dynamics has a great impact on ecosystem and climate change. Remote sensing data and geographical information system (GIS) analysis were widely used in the detection of vegetation pattern dynamics. In this study, the Yellow River Delta was selected as the study area. By using 1986, 1993, 1996, 1999 and 2005 remote sensing data as basic information resource, with the support of GIS, a wetland vegetation spatial information dataset was built up. Through selecting the land-scape metrics such as class area (CA), class percent of landscape (PL), number of patch (NP), largest patch index (LPI) and mean patch size (MPS) etc., the dynamics of vegetation pattern was analyzed. The result showed that the change of vegetation pattern is significant from 1986 to 2005. From 1986–1999, the area of the vegetation, the percent of vegetation, LPI and MPS decreased, the NP increased, the vegetation pattern tends to be fragmental. The decrease in vegetation area may well be explained by the fact of the nature environment evolution (Climate change and decrease in Yellow River runoff) and the increase in the population in the Yellow River Delta. However, from 1999–2005, the area of the vegetation, the percent of vegetation, LPI and MPS increased, while the NP decreased. This trend of restoration may be due to the implementation of water resources regulation for the Yellow River Delta since 1999.     

Application of Remote Sensing Detection and GIS in Analysis of Vegetation Pattem Dynamics in the Yellow River Delta

Regional vegetation pattem dynamics has a great im- pact on ecosystem and climate change.Remote sensing data and geographical information system (GIS) analysis were widely used in the detection of vegetation pattern dynamics.In this study,the Yellow River Delta was selected as the study area.By using 1986, 1993,1996,1999 and 2005 remote sensing data as basic informa- tion resource,with the support of GIS,a wetland vegetation spa- tial information dataset was built up.Through selecting the land- scape met...     

Generality of relationships between leaf pigment contents and spectral vegetation indices in Mallorca (Spain)

Vegetation indices are calculated from reflectance data of discrete spectral bands. The reflectance signal in the visible spectral range is dominated by the optical properties of photosynthetic pigments in plant leaves. Numerous spectral indices have been proposed for the estimation of leaf pigment contents, but the efficacy of different indices for prediction of pigment content and composition for species-rich communities is unknown. Assessing the ability of different vegetation indices to predict leaf pigment content we identify the most suitable spectral indices from an experimental dataset consisting of field-grown high light exposed leaves of 33 angiosperm species collected in two sites in Mallorca (Spain) with contrasting leaf anatomy and pigment composition. Leaf-level reflectance spectra were recorded over the wavelength range of 400 – 900 nm and contents of chlorophyll a, chlorophyll b, total carotenoids, and anthocyanins were measured in 33 species from different plant functional types, covering a wide range of leaf structures and pigment content, five-fold to more than 10-fold for different traits. The best spectral region for estimation of leaf total chlorophyll content with least interference from carotenoids and anthocyanins was the beginning of near-infrared plateau well beyond 700 nm. Leaves of parallel-veined monocots and pinnate-veined dicots had different relationships between vegetation indices and pigments. We suggest that the nature and role of “far-red” chlorophylls which absorb light at longer wavelengths than 700 nm constitute a promising target for future remote sensing studies.

     

Vegetation distribution on Tibetan Plateau under climate change scenario

The impact of climate change on distribution of vegetation is an important aspect in studies on the responses of ecosystems to the climate change. Particularly in the sensitive environments of the Tibetan Plateau, vegetation distribution may be significantly affected by climate change. In this research, the coupled biogeography and biogeochemistry model, BIOME4, was modified according to the features of vegetation distribution on the Plateau, and the Kappa statistic was used to evaluate the modeling results by comparing the simulated vegetation distribution with the existing 1:1,000,000 vegetation map of China. The comparison showed that modified model was appropriate for simulating the overall vegetation distribution on the Plateau. With the improved BIOME4 model, possible changes in the vegetation distribution were simulated under climate change scenarios. The simulated results suggest that alpine meadows, steppes, and alpine sparse/cushion vegetation and deserts would shrink, while shrubs, broad-leaved forests, coniferous-broad-leaved mixed forests, and coniferous forests would expand. Among these types, shrubs, alpine meadows, and steppes would change the most. The shrubs vegetation would expand toward the northwest, replacing most alpine meadows and part of steppes, and thus causing their shrinkages. Yet broad-leaved forests and coniferous-broad-leaved mixed forests demonstrated smaller changes in their distributions. For all the forest types, the area of coniferous forests would increase the most by spreading to the interior of the Plateau.     

2000～2015年丹江口库区植被覆盖时空变化趋势及其成因分析

基于MODIS NDVI、Landsat遥感影像及气象观测数据,应用趋势分析、偏相关分析和土地利用转移矩阵等方法,阐明了2000~2015年丹江口库区植被覆盖时空变化趋势,并探讨了气候变化和人类活动对库区植被覆盖的影响。结果表明：（1）近16 a来丹江口库区植被覆盖度呈增加的趋势,增速为4.73%/10 a（p < 0.001）;（2）40.94%的区域植被覆盖度增加显著,主要分布在库周丘陵和平原地带;10.04%的区域植被覆盖度减少显著,主要位于西北部伏牛山区及库区建成区周边;49.02%的区域变化不显著;（3）丹江口库区植被覆盖度受气候变化影响不显著,但受人类活动影响较大,其中灌草地和农业用地转变为林地是库区植被覆盖度升高的主要原因,农业用地转变为水体和建设用地是部分区域植被覆盖度降低的重要因素,这些土地利用/覆被变化主要受造林、退耕还林、水库蓄水以及建设活动的驱动;（4）生态建设工程和项目的实施对库区植被覆盖度的稳步增加起到了积极作用。 关键词： 植被覆盖度;变化趋势;土地利用;丹江口库区     

Sensitivity analysis assessment of remotely based vegetation indices to improve water resources management

Normalized Difference Vegetation Index (NDVI) is estimated from Landsat 8 sensor acquired in June 2013 to drive four different water-related indices calculated as NDVI derivatives. Different vegetation indices (VIs) have been extracted exclusively in estimation of different VIs: Leaf Area Index, Water Supply Vegetation Index, Crop Water Shortage Index, and Drought Severity Index in addition to estimation of daily evapotranspiration (ET). Sensitivity analysis assesses the contributions of the inputs to the total uncertainty in the analysis outcomes. Vegetation indices are complex and intercepted, therefore the interceptions of the five different vegetation indices are considered in the current study. A comparative analysis of Gaussian process emulators for performing global sensitivity analysis was used to conduct a variance-based sensitivity analysis to identify which uncertain inputs are driving the output uncertainty. The results showed that the interconnections between different VIs vary, but the extent of the features sensitivity is uncertain. Findings from the current work conducted are anticipated to contribute decisively toward an inclusive VIs assessment of its overall verification. Daily ET is the less sensitive and more certain index followed by Drought Vegetation Index.     

2000～2015年川西高原植被EVI海拔梯度变化及其对气候变化的响应

川西高原植被系统受地形因子影响在垂直方向上空间特征差异明显。以MODIS EVI遥感数据作为植被动态监测指数,结合高程数据分析2000～2015年川西高原植被EVI沿海拔梯度的变化规律,然后根据川西高原内部及附近39个气象站点的气温和降水资料开展川西高原植被EVI变化对气候变化的响应研究。结果表明:(1)川西高原近16年生长季植被EVI以0.8%/10 a的速率波动增加,沿海拔梯度具有先升高后降低的特点,垂直分布特征差异显著;(2)川西高原植被EVI变化趋势整体处于稳定状态,改善面积多于退化面积。在1 000 m的低海拔区域,由于人类活动的干扰,植被退化严重;中等海拔范围内水热条件充足,利于植被生长,植被逐渐得到改善,局部地区有轻微退化现象;在4 000 m的高海拔地带,植被EVI波动幅度较低并趋于稳定;(3)不同高程区间内植被EVI变化受气候影响不同,川西高原高海拔地区植被生长主要受气温控制,而中等海拔地区受降水影响较大。(4)在0.05显著性水平下,川西高原植被EVI变化受非气候因子驱动的面积分布较广,约84.22%;受气候因子驱动的面积占比为15.78%,气温对植被生长和分布的驱动作用强于降水驱动作用。     

基于高分一号影像的江汉平原表层土壤湿度指数反演研究

土壤湿度指数遥感监测在农业生产中具有重要的作用。为探讨国产高分一号（GF 1）遥感数据在江汉平原农情参数快速获取中的适用性,以潜江市2017年3月8日的GF 1 WFV影像和106个采样点的土壤湿度实测数据为数据源,选择垂直干旱指数（PDI）、改进型垂直干旱指数（MPDI）和植被调整垂直干旱指数（VAPDI）,对土壤湿度指数反演的效果进行比较和验证。研究结果表明：PDI、MPDI、VAPDI与土壤湿度实测含水量的决定系数分别达到0.649、0.802和0.821,实测土壤含水量验证精度评价也表明各模型均能满足反演的精度要求,说明基于GF 1 WFV影像开展江汉平原的大尺度土壤湿度反演是可行的;在植被覆盖中等区域,MPDI和VAPDI能够在一定程度上克服混合像元对土壤湿度光谱信息的影响,反演的精度要比PDI高,但在高植被覆盖度区,采用垂直植被指数（PVI）修正的VAPDI不易出现植被覆盖饱和现象,具有更高的反演精度;基于3种指数模型反演的土壤湿度指数空间异质性基本一致,但MPDI、VAPDI对土壤湿度变化更为敏感,能反映出不同植被覆盖类型下土壤湿度的实际水平。研究结果可为江汉平原大范围和动态监测表层土壤湿度指数提供理论基础和实践参考。 关键词： 高分一号;土壤湿度指数;PDI;MPDI;VAPDI;江汉平原     

基于面向对象的石漠化山区植被信息提取及分布特征研究

重庆东北部地区是重庆岩溶石漠化重点治理区,该区地形复杂,山高坡陡,植被退化现象严重。了解该地区的植被分布特征,对该区环境的改善和石漠化治理具有十分重要的现实意义。基于Landsat OLI等数据,运用面向对象分类方法对研究区植被信息进行提取,然后对分类后的数据进行统计、制图和分析,并在空间布局上进行总结和探讨,旨在了解该区域植被的空间分布特征和规律。结果表明：（1）在eCognition 9.0软件中进行多尺度分割,再结合地物类型特征使用隶属度函数法进行分类,该方法符合山地地物类型的分布规律和特点,分类精度达到81.35%;（2）研究区属典型的中山地区,海拔在500~1 500 m之间的地区约占64.49%,林地和耕地是该区域主要的地物类型,所占总面积为8 872.22 km2,占研究区总面积的96.50%,各地物类型分布受地形地势的影响较大;（3）综合研究区地形因子(高程和坡度)与植被分布的相关性可知,耕地和草灌主要分布在高程为200~1 500 m且坡度等级在2~4级（5°~35°）之间,该区域人类活动频繁,故受人类活动影响较大,植被覆盖度低,群落生长不稳定,容易受到干扰。针阔混交林主要分布在高程>500 m且坡度等级在2~4级（5°~35°）之间。马尾松林、阔叶林和柏木林主要分布在高程大于500 m且坡度等级在2~5级（5°~45°）之间。 关键词： 多尺度分割;面向对象分类;地形因子;植被空间分布