基于垂直带谱的太白山区山地植被遥感信息提取

遥感数据因其全覆盖的优势被广泛应用于山地植被信息的调查和研究。为了实现山区植被类型的高精度提取,本文以太白山区为实验区,结合山地植被的垂直地带性分布规律,利用太白山植被垂直带谱、高分辨率遥感影像（GF1/GF2/ZY3）和1:1万的数字表面模型（Digital Surface Model, DSM）数据,进行了多层次、多尺度的影像分割,构建了具有植被垂直带谱信息的地形约束因子,并据此进行样本选择和面向对象的分类,分类总精度达92.9%,kappa系数达到0.9160。该方法相比于未辅以垂直带谱信息的分类,总精度提高了10%。研究结果表明,分类过程中加入具有垂直带谱信息的地形约束因子,能显著地提高样本选择的效率和准确率,为后续的植被分类提供了精度的保证。通过人机交互的方式,将垂直带谱知识应用到分类中,可以有效地提高山地植被分类的精度。     

Diversity and geographical pattern of altitudinal belts in the Hengduan Mountains in China

This paper analyses the diversity and spatial pattern of the altitudinal belts in the Hengduan Mountains in China. A total of 7 types of base belts and 26 types of altitudinal belts are identified in the study region. The main altitudinal belt lines, such as forest line, the upper limit of dark coniferous forest and snow line, have similar latitudinal and longitudinal spatial patterns, namely, arched quadratic curve model with latitudes and concave quadratic curve model along longitudinal direction. These patterns can be together called as “Hyperbolic-paraboloid model”, revealing the complexity and speciality of the environment and ecology in the study region. This result further validates the hypnosis of a common quadratic model for spatial pattern of mountain altitudinal belts proposed by the authors. The spatial pattern of altitudinal belts is closely related with moisture-related exposure effect in the Hengduan Mountains. Different combinations (spectra) of altitudinal belts and different base belt types appear in windward and leeward flanks and even in the same flanks of different ranges. This is closely related with the parallel mountain ranges of the Hengduan Mountains, which, at nearly right angle with the moving direction of prevailing moisture-laden air masses from west and east, hold up the warm and humid monsoon wind from moving into the core region and result in different moisture conditions in windward and leeward flanks. However, how to quantitatively describe the moisture-related exposure effect needs further study. In addition, the data quality and data accuracy at present also affect to some extent the result of quantitative modeling and should be improved with RS/GIS in the future.     

Woody species diversity and forest structure from lowland to montane forest in Hyrcanian forest ecoregion

Alborz Mountains host Caspian Hyrcanian forest ecoregion along the northern slopes and forest steppe ecoregion in highlands. Hyrcanian forest covers the southeastern part of Caucasus biodiversity hotspot and is of great biogeographic importance. Altitudinal pattern and correlation between woody species biodiversity (DIV), forest structure ((stem density (DEN), mean basal area (MBA) and mean height class (MHC)) and disturbance (DIS) were explored along 2,400 m altitudinal gradient in Hyrcanian relict forest, Central Alborz Mountains. Vegetation changes from lowland forest (LoF) to mid- altitude forest (MiF) and montane forest (MoF) in this area. The altitudinal gradient was divided into twelve 200 m elevational belts. Point centered quarter method (PCQM) with 96 sampling points and 83 vegetation samples by plot method (PM) were used to record field data. Shannon-Wiener index and Pearson coefficient were used for diversity and correlation analysis. The results showed that DEN decreased linearly, MBA and MHC showed relatively hump shaped and DIS showed a reverse hump shaped pattern of change along altitudinal gradient. Woody species diversity decreased non-steadily from LoF to MoF. Transitional vegetations of Carpinus-Fagus and Fagus-Quercus represented higher diversity of woody taxa compared to adjacent homogenous communities. Significant correlation was observed between altitude and all parameters: DEN with MBA, DIS and DIV; MBA with DIS; MHC with DIS along with DIV; and DIS with DIV at the study area scale. Surprisingly,correlation between studied parameters differed within each vegetation type. Altitude probably acts as a proxy for human and environmental driving forces in this area. Stability of warm and wet condition, season length, soil depth along with forest accessibility probably influences the altitudinal pattern of the studied parameters. Disturbance affects forest structure and consequently diversity; especially in lowlands. The obtained results recommend using both forest biodiversity and mensuration data in management process of forest ecosystems.     

Vegetation-environment relationships between northern slope of Karlik Mountain and Naomaohu Basin, East Tianshan Mountains

Based on data from 22 sample plots and applying the Canonical Correspondence Analysis (CCA),this paper discusses the vegetation-environment relationships between the northern slope of Karlik Mountain and Naomaohu Basin,which is situated in the easternmost end of the Tianshan Mountains,Xinjiang Uygur Autonomous Region,China.For the zonal vegetation,community diversity of mountain vegetation is higher than that of the desert vegetation due to environmental factors.The CCA ordination diagram revealed that the composition and distribution of vegetation types are mainly determined by altitude,soil pH and soil salt content.With increasing elevation,the soil pH and total salt content decrease but the contents of soil organic matter,soil water,total nitrogen and total phosphorus increase gradually.In the CCA ordination diagrams,the sample plots and main species can be divided into five types according to their adaptations to the environmental factors.Type I is composed of desert vegetation distributed on the low mountains,hills,plains and deserts below an elevation of 1900 m;type II is distributed in the mountain and desert ecotone with an elevation of 1900-2300 m,and includes steppe desert,desert steppe and wetland meadow;type III is very simply composed of only salinized meadow;type IV is distributed above an elevation of 2300 m,containing mountain steppe,meadow steppe,subalpine meadow and alpine meadow;type V only contains salinized meadow.The results show that with increasing elevation,species combination changes from the xerophytic shrubs,semi-shrubs and herbs distributed in the low altitude zone with arid climate to the cold-tolerant perennial herbs growing in the high altitudinal zone with cold climate.     

Digital extraction of altitudinal belt spectra in the West Kunlun Mountains using SPOT-VGT NDVI and SRTM DEM

In this paper, a digital identification method for the extraction of altitudinal belt spectra of montane natural belts is presented. Acquiring the sequential spectra of digital altitudinal belts in mountains at an acceptable temporal frequency and over a large area requires extensive time and work if traditional methods of field investigation are to be used. Such being the case, often the altitudinal belts of a whole mountain or the belts at a regional scale are represented by single points. However, single points obviously cannot accurately reflect the spatial variety of altitudinal belts. In this context, a digital method was developed to extract the spectra of altitudinal belts from remote sensing data and SRTM DEM in the West Kunlun Mountains. By means of the 1km resolution SPOT-4 vegetation 10-day composite NDVI, the horizontal distribution of altitudinal belts were extracted through supervised classification, with a total classification accuracy of 72.23%. Then, a way of twice-scan was used to realize the automatic transition of horizontal maps to vertical belts. The classification results of remote-sensing data could thus be transformed automatically to sequential spectra of digital altitudinal belts. The upper and lower lines of the altitudinal belts were then extracted by vertical scanning of the belts. Relationships between the altitudinal belts based on the montane natural zones concerning vegetation types and the geomorphological altitudinal belts were also discussed. As a tentative method, the digital extraction method presented here is effective at digitally identifying altitudinal belts, and could be helpful in rapid information extraction over large-scale areas.     

秦巴山地山体基面高度的提取及分布

秦巴山地是中国的南北分界线,也是黄河和长江的分水岭,其山体效应的定量化影响秦巴山地山体垂直带的分布格局、非地带性因素的作用强度和机理,以及中国暖温带和北亚热带的具体位置的确定。山体基面高度是影响山体效应最重要和关键的地形因子,其定量化和数字化提取是秦巴山地山体效应定量化研究的重要内容。本研究针对秦巴山地山体效应的定量化研究,使用30 m分辨率的STRM-1数据,分别基于山体特征线和流域分区2种方法提取了秦巴山地的山体基面高度分区,并根据地形起伏度和坡度,确定基面范围,计算了山体基面高度值。结果表明：① 基于山体特征线的方法将秦巴山地分为93个基面高度分区,基于流域分区的方法将秦巴山地分为209个基面高度分区,根据2种分区结果提取的基面高度值相差不大且均体现了秦巴山地地势的特点;② 秦巴山地山体基面高度从东向西呈阶梯状递增的趋势;③ 从南到北,秦巴山地的东段和中段均呈先增高后降低的趋势,即从大巴山向北至汉江谷地降低,再向北至秦岭升高;④ 山地的不同侧翼的山体基面高度不同,秦岭南坡的基面高度（1000~1809 m）明显高于北坡（850~1300 m）。秦巴山地山体基面高度与其植被带分布上限联系密切,实现山体基面高度的数字化提取,为山体效应的定量化研究提供了重要的技术支持。     

Tree population dynamics of three altitudinal vegetation communities on Mount Cameroon (1989�C2004)

Changes in permanent sample plots in the lowland,submontane and montane forests on Mount Cameroon(4,095 m above sea level),an active volcano,are described for 15 years from 1989 to 2004.Throughout the study period,the stocking level of trees with a diameter at breast height(DBH) ≥ 10 cm in the three forests were lower than in pan-tropical stands suggesting a significant impact of volcanic and human-related activities on the vegetation communities on the mountain.Annual mortality rates in the submontane and montane forests were consistent with those reported for comparable altitudinal ranges in the Blue Mountains of Jamaica.The annual mortality rate was higher in the lowland forest than other lowland sites included.Divergence between recruitment and mortality rates was large suggesting that the three vegetation communities have not reached their climax.The seven-year difference in half-life of large trees(with a DBH ≥ 50 cm) in the submontane and montane forests suggests an altitudinal effect on turnover of larger trees that in turn contributes to the frequent small stature of high altitude forests.There was little evidence of an altitudinal effect on species turnover and growth rate.This finding supports generalizations about the zero effect of growth on the stature of high altitude trees.Understanding forest dynamics is crucially important in the management of tropical montane environmentsand in this instance particularly so given the recent creation of the Mount Cameroon National Park.     

IMPACT OF ROAD CONSTRUCTION ON VEGETATION ALONGSIDE QINGHAI-XIZANG HIGHWAY AND RAILWAY

Based on the field investigation in August 2001 and August 2002, digital China Vegetation Map in 2001 and Qinghai-Xizang(Tibet) Plateau Vegetation Regionalization Map in 1996, vegetation characteristics along two sides of Qinghai-Xizang highway and railway are studied in this paper. Meanwhile, the impact of Qinghai-Xizang highway and railway constructions on the vegetation types are analyzed using ARCVIEW. ARC/1NFO and PATCH ANALYSIS. It was found that: 1) Qinghai-Xizang highway and railway span 9 latitudes, 12 longitudes and 6 physical geographic regions (East Qinghai and Qilian mountain steppe region, Qaidam mountain desert region,South Qinghai-Xizang alpine meadow steppe region. Qiangtang alpine steppe region, Golog-Nagqu alpine shrubmeadow region and South Xizang mountain shrub steppe region); 2) the construction of Qinghai-Xizang highway and railway destroyed natural vegetation and landscape, especially in 50m-wide buffer regions along both sides of the roads, it was estimated that the net primary productivity deceased by about 30 504.62t/a and the gross biomass deceased by 432 919.25-1 436 104.3t. The losing primary productivity accounted for 5.70% of the annual primary productivity within lkm-wide buffer regions (535 005.07-535 740.11t/a), and only 0.80%-0.89% of that within 10km-wide buffer regions (3 408 950.45-3 810 480.92t/a). The losing gross biomass was about 9.47%-17.06% of the gross biomass within lkm-wide buffer regions (7 502 971.85-25 488 342.71t), and only 1.47%-2.94% of that within 10km-wide buffer regions (43 615 065.35-164 150 665.37t).     

GF-2在城市绿地分类及景观格局度量中的应用

城市植被是维护城市生态系统平衡的重要组成元素,不同类型的植被具有不同的生态服务效益,然而基于植被精细分类的城市绿地景观格局度量研究较少。因此本文以北京城市副中心为研究区,使用夏、冬两季GF-2卫星遥感影像,基于随机森林特征优选和面向对象分类,划分了植被类型,并在此基础上使用景观指数法和移动窗口法分别度量了功能区尺度和栅格尺度绿地景观特征。研究结果表明：针对GF-2数据,使用多尺度分割后影像对象的光谱、纹理特征可以有效地提取植被信息;不同时相的影像能反映各类型植被的物候特征,相比于单时相数据,其分类精度更高,达到了87.7%;各功能区绿地景观格局特征差异较大,城市绿心拥有最丰富的植被且分布集中,商业区植被覆盖度和多样性均较低,绿地景观破碎;北京城市副中心景观多样性及各类型植被分布特征的空间差异显著,当前副中心绿地景观格局已形成规划的基本轮廓,但城市绿心和老城区公共绿地的建设仍有不足。研究探明了北京城市副中心绿地建设现状,证明了国产GF-2卫星在城市生态环境监测中具有较高的应用价值,有助于推动GF-2在城市生态环境领域的应用,并为副中心建设中的绿地系统监测和优化提供参考。     

Digital Spectra and Analysis of Altitudinal Belts in Tianshan Mountains, China

Based on the framework of the geo-info spectra of montane altitudinal belts, this paper firstly reviews six classification systems for the spectra of mountain altitudinal belts in China and considers that detailed regional study of altitudinal belts is the key for reaching standardization and systemization of mountain altitudinal belts. Only can this further identify and resolve problems with the study of altitudinal belts. The factors forming the spectra of altitudinal belts are analyzed in the Tianshan Mountains of China, and a digital altitudinal belt system is constructed for the northern flank, southern flank, the heartland, and Ili valley in the west. The characteristics of each belt are revealed with a summarization of the pattern of areal differentiation of altitudinal belts.     

Alternative stable states in mountain forest ecosystems: the case of European larch(Larix decidua) forests in the western Alps

European larch(Larix decidua) forests of the western Alps form extensivecultural landscapes whose resilience to global changes is currently unknown. Resilience describes the capacity of ecological systems to maintain the same state, i.e., the same function, processes, structure, and composition despite disturbances, environmental changes and internal fluctuations. Our aim is to explore the resilience of larch forests to changes in climate and land use in the western Italian Alps.To do so, we examined whether larch forests can be described as an alternative stable state in mountain forest ecosystems. We used tree basal area data obtained from field forest inventories in combination with topography, forest structure, land use, and climate information. We applied three different probabilistic methods: frequency distributions, logistic regressions, and potential analyses to infer the resilience of larch forests relative to that of other forest types.We found patters indicative of alternative stable states: bimodality in the frequency distribution of the percent of larch basal area, and the presence of an unstable state, i.e., mixed larch forests, in the potential analyses. We also found:(1) high frequency ofpurelarchforestsathighelevation,(2)the probability of pure larch forests increased mostly with elevation, and(3) pure larch forests were a stable state in the upper montane and subalpine belts.Our study shows that the resilience of larch forests may increase with elevation, most likely due to the altitudinal effect on climate. Under the same climate conditions, land use seems to be the main factor governing the dominance of larch forests. In fact, subalpine larch forests may be more resilient, and natural succession after land abandonment, e.g., towards Pinuscembra forests, seems slower than in montane larch forests. In contrast, in the upper montane belt only intense land use regimes characterized by open canopies and forest grazing may maintain larch forests.We conclude that similar approaches could be applied in other forest ecosystems to infer the resilience of tree species.     

Alternative stable states in mountain forest ecosystems: the case of European larch (<Emphasis Type="Italic">Larix decidua</Emphasis>) forests in the western Alps

European larch (Larix decidua) forests of the western Alps form extensive cultural landscapes whose resilience to global changes is currently unknown. Resilience describes the capacity of ecological systems to maintain the same state, i.e., the same function, processes, structure, and composition despite disturbances, environmental changes and internal fluctuations. Our aim is to explore the resilience of larch forests to changes in climate and land use in the western Italian Alps. To do so, we examined whether larch forests can be described as an alternative stable state in mountain forest ecosystems. We used tree basal area data obtained from field forest inventories in combination with topography, forest structure, land use, and climate information. We applied three different probabilistic methods: frequency distributions, logistic regressions, and potential analyses to infer the resilience of larch forests relative to that of other forest types. We found patters indicative of alternative stable states: bimodality in the frequency distribution of the percent of larch basal area, and the presence of an unstable state, i.e., mixed larch forests, in the potential analyses. We also found: (1) high frequency of pure larch forests at high elevation, (2) the probability of pure larch forests increased mostly with elevation, and (3) pure larch forests were a stable state in the upper montane and subalpine belts. Our study shows that the resilience of larch forests may increase with elevation, most likely due to the altitudinal effect on climate. Under the same climate conditions, land use seems to be the main factor governing the dominance of larch forests. In fact, subalpine larch forests may be more resilient, and natural succession after land abandonment, e.g., towards Pinus cembra forests, seems slower than in montane larch forests. In contrast, in the upper montane belt only intense land use regimes characterized by open canopies and forest grazing may maintain larch forests. We conclude that similar approaches could be applied in other forest ecosystems to infer the resilience of tree species.     

集成虚拟森林环境与景观指数计算模型的森林景观格局分析

森林景观格局分析是森林景观优化配置与实施规划的基础。本文以森林资源二类调查数据为基础,采用森林景观定量空间分析方法,在自主开发而成的森林景观可视化系统VisForest支持下,利用VS2008开发平台、ArcEngine组件与OSG图形渲染引擎,集成景观指数计算模型和地理信息系统的分析方法,形成森林景观格局分析与景观三维可视化模拟系统,实现了景观格局分析的定量化和景观可视化;并以闽侯县白沙国有林场2013年的数据为例,按优势树种划分森林景观类型,计算分析了景观组成结构、斑块特征和景观异质性等。结果表明,杉木林、马尾松林、湿地松林、木荷林、非林地是白沙林场的优势景观类型,总体景观类型丰富、景观异质性较高;木荷作为隔离带树种,以小面积狭长斑块居多,其形状指数和分形维数最大,斑块形状复杂而不规则;森林景观的三维可视化模拟为研究景观格局提供了一个直观、交互的展示平台。     

Omni-directional distribution patterns of montane coniferous forest in the Helan Mountains of China

Slope aspect has significant effect on the development and distribution of montane forest,especially in arid and semiarid regions.This paper,using SPOT5 images and 1:50,000 DEM,digitally extracts and analyzes the spatial information of montane coniferous forest(mainly Qinghai spruce),and thereby explores how the upper and lower limits and the altitudinal range of coniferous forest vary and how the area of coniferous forest is related with annual insolation with all aspects in the Helan Mountains.The results show that: 1) In the eastern flank,the lower limit of coniferous forest is between 1,600 m and 2,000 m a.s.l.,and the upper limit between 2,800 m and 3,000 m a.s.l.;in the western flank,the lower limit of coniferous forest is between 2,000 m and 2,300 m a.s.l.,and the upper limit between 2,800 m and 3,100 m a.s.l.2) The altitudinal ranges of coniferous forest are 806~1,435 m,438~1,140 m for eastern flank and western flank,respectively.3) The area of coniferous forest takes on a normal distribution with aspect,and it has a close relationship with annual insolation.This distribution model developed in this paper quantitatively reveals the significant effect of slope aspect on the distribution of coniferous forest in arid and semi-arid land.     

山地垂直带谱信息系统应用分析与技术改进

山地垂直带信息系统虽实现了山地垂直带谱的首次数字集成,但系统中还存在标识不清、垂直带颜色不易变动、大量带谱显示时可视程度较差等问题。本文以山地垂直带谱空间可视化为研究对象,探讨系统的技术改进,包括山地垂直带谱图形数据的立体标识、翻转页面、颜色可定制、数据轴可操作等功能的实现。山地垂直带信息系统空间可视化功能的进一步完善,不仅可以有效地管理山地垂直带谱数据,而且可方便直观地显示山地垂直带谱数据,有效地集成山地垂直带谱数据并且充分地发掘山地垂直带谱空间变化规律,也为建立界面友好的世界山地垂直带信息系统奠定了基础。     

山地垂直带谱数字识别的技术实现和图谱构建

山地垂直带谱是地学信息图谱的一个重要组成部分。"带谱数字识别"就是利用现代数字方法和数据,对客观存在的山地垂直自然带谱进行提取,获取比较完备的山地垂直带连续图谱模式。这是从传统山地垂直带研究走向地学信息图谱研究的重要步骤,有助于使我们对山地垂直谱的认识上升到地学信息图谱的高度。本文探索MATLAB语言快速、准确地实现各种识别算法,结合VB.NET构建的用户操作界面,实现带谱的信息提取、分类集成和多样化表达。山地垂直自然带谱的数字识别分为3种不同的模式:山系单侧、山体单峰和山体多峰;依据各自不同的识别算法,可以得到山地连续的带谱模式。本文全面介绍了数字带谱、带谱数字识别的含义,以及数字识别模型的原理、算法和原型系统等实现过程。     

京津冀地区植被时空动态及定量归因

作为气候变化的敏感指示器,植被的物候、生长、空间分布格局等特征及其动态变化主要取决于气候环境中的水热条件,因此在气候变化背景下,气候-植被关系成为了全球变化研究的前沿和热点问题。本文综合平均温度、降水、水汽压、湿度、日照时数、SPEI等气候因子,坡度、坡向海拔等地形因子及人为活动因子,应用地理探测器方法针对2006-2015年京津冀地区不同季节NDVI、不同地貌类型区、不同植被类型区生长季NDVI的定量归因研究,揭示了过去10年间植被时空分布格局,及植被对气候、非气候因素响应的季节差异与区域差异,以期为生态工程的建设与修复提供参考意义。趋势分析表明：①2006-2015年京津冀地区NDVI呈现增加趋势,但存在显著的空间差异,如山地生长季NDVI的增长速率大于平原、台地、丘陵等地;②基于地理探测器的定量归因结果表明,降水是年尺度上NDVI空间分布的主导因子（解释力39.4%）,土地利用与降水的交互作用对NDVI的影响最为明显（q=58.2%）;③NDVI对气候因子的响应存在季节性及区域性差异,水汽压是春季NDVI空间分布的主导因子,湿度是夏、秋两季的主导因子,土地利用是冬季的主导因子;④影响因子对生长季NDVI的解释力因不同地貌类型区、不同植被类型区而差异显著。     

Vegetation geo-climatic zonation in the rocky mountains,Northern Utah,USA

We developed a vegetation geo-climatic zonation incorporating the zonal concept, gradient and discriminant analysis in Wasatch Range, northern Utah, USA. Mountainous forest ecosystems were sampled and described by vegetation, physiographic features and soil properties. The Snowpack Telemetry and National Weather Service Cooperative Observer Program weather station networks were used to approximate the climate of sample plots. We analysed vegetation and environmental data using clustering, ordination, classification, and ANOVA techniques to reveal environmental gradients affecting a broad vegetation pattern and discriminate these gradients. The specific objective was to assess and classify the response of the complex vegetation to those environmental factors operating at a coarse-scale climatic level. Ordination revealed the dominant role of regional, altitude-based climate in the area. Based on vegetation physiognomy, represented by five tree species, climatic data and taxonomic classification of zonal soils, we identified two vegetation geo-climatic zones： （1） a montane zone, with Rocky Mountain juniper and Douglas-fir; and （2） a subalpine zone, with Engelmann spruce and subalpine fir as climatic climax species. Aspen was excluded from the zonation due to its great ecological amplitude. We found significant differences between the zones in regional climate and landformgeomorphology/soils. Regional climate was represented by elevation, precipitation, and air and soil temperatures; and geomorphology by soil types. This coarsescale vegetation geo-climatic zonation provides a framework for a comprehensive ecosystem survey, which is missing in the central Rocky Mountains of the United States. The vegetationgeoclimatic zonation represents a conceptual improvement on earlier classifications. This framework explicitly accounts for the influence of the physical environment on the distribution of vegetation within a complex landscape typical of the central Rocky Mountains and in mountain ranges elsewhere.     

中国陆地地貌基本形态类型定量提取与分析

中国陆地地貌基本形态类型由海拔和地势起伏度两个指标组合划分而成,其中海拔分4级,地势起伏度分7级,组合后的基本形态类型共25类。按照数字地貌分类体系,海拔和起伏度拥有全国普适性的分级指标,基于DEM数据可获得两指标的分级类型。试验表明,基于SRTM-DEM(水平分辨率90m,相当于1:25万比例尺)数据可得出全国普适性的采用单元为4km²,利用1:10万、1:400万等比例尺数据进行DEM试验,得出我国存在0.4、4、12、18、21km²五种不同规模的采样单元,并分别对应着不同的比例尺。因我国地貌复杂多样,仅利用DEM数据所获得的海拔和地势起伏度分级数据不能完全反映不同地域的地貌特征,故利用遥感等多源数据,综合多种信息获得的地貌类型,可很好地反映出我国的海拔4级分级特征和地势起伏度7级空间分布,进而获得全国陆地的25种基本形态类型的面积及空间分布格局。     

MODIS EVI时序数据重建方法及拟合分析

植被遥感监测中长时间序列数据择优的重建方法,已成为当今一个研究热点。本文以东北地区5种主要植被覆盖类型为例,在定性分析TIMESAT提供的3种常用重建方法对EVI（Enhanced Vegetation Index）时序曲线重建效果的基础上,定量对比研究了各方法,对原始高质量EVI点真实值的保真性,及对原始曲线整体特征的保持度。结果表明：S-G（Savitzky-Golay）滤波对原始曲线生长季的峰值及宽度重建效果较好,但容易因过度拟合保留过多噪声,特别是草地和灌丛类型;非对称性高斯函数（AG）和双Logistic曲线（DL）方法相似,对草地、灌丛和耕地的重建结果更接近真实值,但AG拟合对波峰处异常值的处理结果较差,重建后波峰表现低平。3种算法对原始EVI时序数据的保真性和对原始时序数据曲线特征的保持度,都表现出与植被类型分布相关的空间分布格局。分析结果表明,在东北地区,AG算法对草原和灌丛的重建效果最好,DL算法对耕地重建效果最优,S-G算法最适合对落叶阔叶林和落叶针叶林进行重建处理。