基于植被指数和神经网络的热带人工林地上蓄积量遥感估测

热带森林作为陆地生态系统的组成成分之一,研究其蓄积量估测对我们了解其在全球碳循环中的地位和作用有很重要的意义.但遥感估测森林生态参数的精度如何,还是个不确定的问题.利用LANDSAT-TM数据,基于森林清查数据和遥感技术,以尾叶桉和加勒比松为例,对中国南方地区人工林蓄积量估测进行了尝试研究.首先,通过测量样方胸径、树高,建立森林蓄积量估算模型.其次,通过对比分析不同植被指数与森林蓄积量之间的关系,选择合适植被指数组合,建立多元回归和神经网络模型.结果表明:单波段TM数据和大多数植被指数与蓄积量相关性并不好.神经网络比回归分析模拟效果好.而多元回归和神经网络模型大大提高预测精度.本研究方法对大面积的森林蓄积量估测具有一定的参考价值.     

神经网络模型森林生物量遥感估测方法的研究

森林生物量的估测是全球变化研究的基础,而遥感宏观、综合、动态、快速的特点决定了基于遥感的生物量模型为森林生物量估测的发展方向,目前的遥感生物量估测方法大多基于回归分析,需要预先假设、事后检验,仅为经验性的统计模型。神经网络的分布并行处理、非线性映射、自适应学习和容错等特性,使其具有独特的信息处理和计算能力,在机制尚不清楚的高维非线性系统体现出强大优势,可以用于遥感生物量估测。文章在野外调查的基础上,尝试应用BP网络和RBF网络技术,建立广州TM遥感影像数据与森林样方生物量实测数据之间的神经网络模型,通过训练和仿真,与生物量实测数据进行比较。结果表明,在独立样地估测中,人工神经网络估测的相对误差均小于15.18%,获得了满意的效果。而RBF网络与BP网络相比,在识别精度上、稳定性、速度上,均优于BP网络,其最大相对误差不超过10.12%,平均相对误差为4.76%。可见应用神经网络方法的“黑箱”操作虽然难以归纳出指导性规律,但可以获得很高的精度。尤其RBF网络,在训练完成后,可以应用该模型进行大区域生物量估算,对于森林的规划及管理具有深远意义。     

基于遥感技术的人工速生林二氧化碳吸收量估测

选择宁夏中卫市中冶·美利纸业集团的林纸一体化人工速生杨基地作为研究对象,以CBERS/CCD影像和同步实测样地数据为基础,利用相关分析方法筛选出相关系数分别为0.939和0.936的NDVI和RVI两个变量,构建了基于NDVI、RVI及NDVI和RVI的3个人工速生林地叶面积遥感回归估测模型,R2分别为0.882、0.877和0.885,并通过相关检验,估算出研究区不同林龄速生杨林地的叶面积。结合样地的实测净光合速率（PN,Net Photosynthetic Rate,μmol·m-2·s-1）,推算出不同林龄的速生杨单位叶面积的年二氧化碳净吸收量,最终估算出整个研究区的年固定碳（CO2）的净增量分别为326 648.66、315 688.73和322 509.04 t。通过与常规方法测得结果的比较,表明遥感结合地面实测数据估算林木固碳是可行的;从建立模型的R2值分析,根据不同植被指数建立的多元模型的精度要优于单一植被指数建立的一元模型;根据估测结果,随着林龄的增长,林木的固碳能力不断提高,但受不同生境的影响,增长速度存在空间差异。     

基于机器学习算法的干旱区绿洲地上生物量估算

植被地上生物量是反映陆地生态系统固碳能力的重要指标,利用遥感技术开展干旱区植被地上生物量估算与空间反演,可为荒漠绿洲生态系统的健康评价与碳储量估算提供重要依据。以野外调查和实地采样数据为基础,利用Landsat 8 OLI多光谱影像提取的7个植被指数和13个波段变量构成4种建模变量组合,采用支持向量机(Support Vector Machine,SVM)、反向传播神经网络(Back Propagation Neural Network,BPNN)、极端梯度提升(e Xtreme Gradient Boosting,XGBoost)和随机森林(Random Forest,RF)这4种机器学习算法对新疆渭干河-库车河三角洲绿洲地上生物量进行遥感估算和空间反演。结果表明,(1)由波段变量和随机蛙跳算法优选变量构建的植被地上生物量反演模型,其估测精度明显优于全变量和指数变量,预测能力更为稳定。与SVM和BPNN算法相比,XGBoost和RF算法构建的模型具有更好的估测效果,能更准确地估算研究区植被地上生物量。(2)在构建的估测模型中,波段变量结合RF算法模型的精度最高,稳定性最强,其建模集和...     

基于遥感估算的上海城市森林碳储量空间分布特征

城市森林在固碳释氧、应对气候变化方面发挥着重要作用,对其碳储量的估算为中国城市做好碳达峰、碳中和工作提供重要数据基础。以上海城市森林为研究对象,采用样地调查数据与Landsat OLI遥感影像,分别计算样地碳储量和遥感参数变量(波段数据、植被指数数据和纹理数据),构建基于多元逐步回归模型和普通克里格残差矫正相结合的估算模型,分析区域尺度城市森林碳储量和碳密度的空间分布特征。结果表明,(1)81个样地的碳储量范围为0.09—7.10 t,均值为2.14 t,数据有右偏分布和瘦尾特征,变异系数为0.75,样地类型多样;上海城市森林总碳储量为2.87 Mt,碳密度主要集中在13—40 t·hm~(-2)之间,均值为25.09 t·hm~(-2),整体呈现中部较低,东西部较高的态势,与土地利用强度及城市森林分布有关。(2)主城区碳储量为0.37 Mt,碳密度为1—69 t·hm~(-2),均值为21.77 t·hm~(-2);非主城区碳储量为2.50 Mt,占上海城市森林碳储量的87.11%,碳密度范围为0—89 t·hm~(-2),均值为25.66 t·hm~(-2)。(3)在选择波段反射率和遥感植被指数的同时,提取影像纹理特征并纳入模型,提高了影像的分类精度;采用多元逐步回归模型结合残差矫正的方法估算城市森林碳储量,使估算结果的均方根误差降低了10.29%,平均绝对误差降低了5.5%,提高了估算精度。     

热带人工林生物物理参数及生物量对RADARSAT SAR信号响应研究

合成孔径侧视雷达穿透冠层而获得树干信息的特点使得其具有在南方NDVI易饱和地区的生物量估测方面优于光学遥感。文章利用地面实测调查数据,首次系统地研究了热带人工林生物物理参数及生物量对RADARSAT-SAR信号响应。结果表明:RADARSAT-SAR后向散射系数与森林生物量、树高、胸径可以用对数模型模拟。对于本实验区人工速生林来说,在森林生物量小于40 t.hm-2时RADARSAT-SAR后向散射系数与森林生物量具有较强相关性。桉树树高与后向散射系数的相关系数大于桉树胸径与后向散射系数关系。松树胸径与后向散射系数相关系数高于树高与后向散射系数的相关系数。研究结果为SAR在热带森林研究中的应用提供了一定的理论基础。     

基于森林清查和遥感的城市森林净初级生产力估算

城市森林作为陆地生态系统的组成成分之一,研究其在全球碳循环中的地位和作用有很重要的意义.文章以上海城市森林为研究对象,开展基于基于森林清查数据和遥感技术的城市森林净初级生产力(NPP)估算研究.首先,根据选取上海市典型的森林植被类型,设置森林植被样方,测量反映植被生物学特性的特征参数,包括林龄、胸径、树高和叶面积指数(LAI)等,采用能反映林龄和蓄积量共同影响的生产力回归模型估算了样方NPP,建立了基于LAI的样方NPP回归模型;其次,利用一景相近时相的SPOT5影像,经进行几何纠正和辐射定标后,计算出能较好地反映植被特征和消除土壤背景影响的修正土壤调节植被指数(MSAVI),建立了基于MSAVI的区域森林LAI遥感估测模型;最后,根据建立的样方NPP回归模型以及区域LAI遥感估测模型,进行尺度化转换,估算出区域尺度上的上海城市森林净初级生产力.通过比较与前人运用传统方法研究估算的NPP,精度可达到89%,且本模型简单可行.因此本研究可为快速定量评估城市森林碳储量提供依据.     

辽东湾双台子河口湿地翅碱蓬（Suaeda salsa）生物量遥感反演研究

通过现场采集辽东湾双台子河口翅碱蓬（Suaeda salsa）光谱数据,收集标准样方（1 m×1 m）的翅碱蓬生物量（Biomass）,建立相关植被指数与翅碱蓬生物量的遥感反演回归算法,发现辽宁双台子河口湿地翅碱蓬生物量与植被指数PVI、SAVI和MSAVI的相关系数R2较高,直线回归方程相关系数分别达到0.626、0.698和0.679。同时,利用得到的估算模型,结合1990—2005年的双台子河口的TM影像数据,反演该区域翅碱蓬分布面积和生物量。结果发现：辽东湾双台子河口湿地翅碱蓬面积变化呈先降后升的趋势。通过植被指数（NDVI、RVI、PVI和MSAVI）和生物量的算法反演,发现翅碱蓬生物量曲线与分布面积曲线的变化趋势一致。     

基于MODIS数据的河南省冬小麦产量遥感估算模型

小麦是世界上最重要的粮食作物,小麦生产对中国的粮食保障起着十分重要的作用,及时、准确、大范围对小麦产量进行监测预报,对于农学经济发展和粮食政策制定具有极为重要的现实意义。对作物产量进行遥感监测的原理是建立在其遥感特征基础之上的,通过建立作物长势指标与遥感信息的定量关系,可实现对作物产量的监测预报。文章基于2009年MODIS遥感数据和气象数据,利用Arcgis和ENVI提取纯小麦像元,并提取纯小麦像元对应的NDVI、NPP和LAI,获取分县NDVI、NPP和LAI均值,利用统计软件对产量数据和分县遥感参数均值进行数据整理和分析,建立了河南省冬小麦产量估算模型。以往研究多采用遥感图像上某像元和地面调查点进行研究,具有很大的不确定性,文章以县为单位,对冬小麦平均单产和县域内冬小麦种植像元遥感参数的均值进行相关研究,提高了模型模拟精度。同时文章选用多种遥感参数和多项气象因子建立估产模型,避免了针对一个参数进行估产的局限性。在最佳时相的选择上,根据冯美辰（2010）以往的研究结果,从4月以后,5月8日和4月20Et植被指数和产量相关性最大,4月份之前冬小麦处于返青到拔节期,对产量来说还有很多不确定闪素,因此文章选用5月8El和4月20日进行冬小麦估产研究。结果表明,5月8日的估产模型优于4月20日,加入气象冈子的遥感气象估产模型优于只采用遥感参数进行估产的遥感模型。利用2010年产量数据对模型精度进行检验,遥感气象模型预测精度在70．2％N99．7％之间,平均精度为90．7％;遥感模型预测精度在68．1％到95．5％之间,平均精度为83．9％。表明遥感气象模型模拟精度更高,其精度可以满足大面积估产要求,可以对产量预报提供科学参考。     

基于多元线性模型、支持向量机(SVM)模型和地统计方法的地表水溶解性总固体(TDS)估算及其精度对比——以艾比湖流域为例

地表水溶解性总固体(TDS)是地表水各组分浓度的总指标,是地表水水化学特性长期演变的最终结果,也是表征水文地球化学作用过程的重要参数,TDS的高低直接影响地表水的含盐量.本研究以艾比湖流域为研究对象,结合实测地表水TDS数据;选用准同步的Landsat OLI数据,首先,利用光谱诊断指数选取与地表水TDS相关性较高的波段,其次,利用地统计方法、多元线性回归模型和支持向量机(SVM)模型对TDS进行预测,并对其结果进行精度比较.结果表明,SVM模型为最优估测模型,拟合决定系数R2为0.97,均方误差(RMSE)为50.59;多元线性回归模型的精度与SVM模型精度较为接近,拟合决定系数R2为0.9,RMSE为66.55;地统计克里格插值法预测精度最低,拟合决定系数R2为0.87,RMSE为95.73.遥感估测SVM模型预测值在大区域能较好地反映出艾比湖流域TDS的总体特征.该模型在水质遥感领域的应用中具有良好的可行性和有效性,其预测结果也与艾比湖流域水体TDS的实际分布相吻合,因此遥感估测SVM模型在水质估测中具有一定的应用潜力.     

Principal components derived from soil physicochemical data explained a land degradation gradient,and suggested the applicability of new indexes for estimation of soil productivity in the Sakaerat Environmental Research Station,Thailand

To find a principal component (PC) that quantifies the degree of soil degradation, we analyzed various physicochemical characteristics of soils over a land degradation gradient related to aboveground vegetation in the Sakacrat Environmental Research Station (SERS), Thailand. The aboveground vegetative types representing the degradation gradient were bare ground (BG, highly degraded), dry dipterocarp forest (DDF, moderately disturbed) and dry evergreen forest (DEF, the original vegetation). Soils under these vegetative types were sampled in February (dry season). March just after temporal precipitation) and June (rainy season) 2001. Through the period of this research, the degradation was consistently explained by sandy texture, high bulk density, lower pH, high exchangeable acidity, poor mineral and organic nutrients and dryness. Principal component analysis (PCA) was applied to determine significant principal components (PCs) that clarify the differences in soil properties between the vegetative types and between the timing of soil sampling. The PC loadings suggested that the first PC was the component that indicates total fertility of soil in the site, while the fifth PC indicates the dry to wet seasonal transition. The first PC was named the total fertility component (TFC). The linear regression between the TFC score and recently proposed indexes, the soil fertility index (SFI) or the soil evaluation factor (SEF), was highly significant (p < 0.001), indicating that the SFI and the SEF are applicable to measuring total fertility of soils in the SERS.     

基于GWR的中国地级城市SO2年均质量浓度模拟

中国城市空气污染问题已经引起广泛关注。目前相关研究很多,但是以空间位置为拟合参数,对空气质量进行回归模拟的研究较少。以2010年中国地级以上城市SO2年均质量浓度为因变量,分别应用普通线性回归和地理加权回归（GWR）模型模拟SO2年均质量浓度,其中地理加权回归方法考虑了空间位置的影响并以此作为回归参数。回归的自变量指标体系包括气象要素（多年平均温度、光照、降水）、植被覆盖（NDVI）、地形要素（坡度、坡向、起伏度）、人为因素（GDP、能源消费）几个方面。由于各指标之间存在较强的相关性,用主成分分析方法计算得到温度、日照、降水、NDVI表征的气象植被综合指标,高程、坡度、起伏度表征的地形综合指标,和GDP、能源消费表征的人为因素综合指标。用3个综合指标值作为自变量进行回归模拟。普通回归结果较差,其r^2为0.11,矫正的r^2为0.10;GWR模型模拟结果相对较好,其拟合优度显著提高,r^2为0.66,矫正的r^2为0.47。因此,地理加权回归适合进行此类拟合,普通线性回归不适合。通过对比地理加权回归模拟的各个城市的拟合优度,发现年均质量浓度数值较高的地区拟合效果较差,这些地区主要集中在中国华北和南部部分地区。与基于机理的模型相比,GWR 模型和其各具优缺点,GWR 的优势主要表现在数据及其格式化要求低,计算机软硬件条件要求低,运算速度快等。     

Spectral characteristics of plant communities from salt marshes: A case study from Chongming Dongtan, Yangtze estuary, China

The spectral reflectance of recently formed salt marshes at the mouth of the Yangtze River, which are undergoing invasion by Spartina alterniflora, were assessed to determine the potential utility of remotely sensed data in assessing future invasion and changes in species composition. Following a review of published research on remote sensing of salt marshes, 53 locations along three transects were sampled for paired data on plant species composition and spectral reflectance using a FieldSpec? Pro JR Field Portable Spectroradiometer. Spectral data were processed concerning reflectance, and the averaged reflectance values for each sample were reanalysed to correspond to a 12-waveband bandset of the Compact Airborne Spectral Imager. The spectral data were summarised using principal components analysis (PCA) and the relationships between the vegetation composition, and the PCA axes of spectral data were examined. The first PCA axis of the reflectance data showed a strong correlation with variability in near infrared reflectance and ‘brightness’, while the second axis was correlated with visible reflectance and ‘greenness’. Total vegetation cover, vegetation height, and mudflat cover were all significantly related to the first axis. The implications of this in terms of the ability of remote sensing to distinguish the various salt marsh species and in particular the invasive species S. alterniflora were discussed. Major differences in species with various physiognomies could be recognised but problems occurred in separating early colonising S. alterniflora from other species at that stage. Further work using multi-seasonal hyperspectral data might assist in solving these problems.     

Postfire stand structure in a semiarid savanna: cross-scale challenges estimating biomass.

Algorithms relating remotely sensed woody cover to biomass are often the basis for large-scale inventories of aboveground carbon stocks. However, these algorithms are commonly applied in a generic fashion without consideration of disturbances that might alter vegetation structure. We compared field and remote sensing estimates of woody biomass on savannas with contrasting disturbance (fire) histories and assessed potential errors in estimating woody biomass from cover without considering fire history. Field surveys quantified multilayer cover (MLC) of woody and succulent plants on sites experiencing wildfire in 1989 or 1994 and on nearby unburned (control) sites. Remote sensing estimates of the woody cover fraction (WCF) on burned and control sites were derived from contemporary (2005) dry-season Landsat Thematic Mapper imagery (during a period when herbaceous cover was senescent) using a probabilistic spectral mixture analysis model. Satellite WCF estimates were compared to field MLC assessments and related to aboveground biomass using allometry. Field-based MLC and remotely sensed WCFs both indicated that woody cover was comparable on control areas and areas burned 11-16 years ago. However, biomass was approximately twofold higher on control sites. Canopy cover was a strong predictor of woody biomass on burned and control areas, but fire history significantly altered the linear cover-biomass relationship on control plots to a curvilinear relationship on burned plots. Results suggest predictions of woody biomass from "generic" two-dimensional (2-D) cover algorithms may underestimate biomass in undisturbed stands and overestimate biomass in stands recovering from disturbance. Improving the accuracy of woody-biomass estimates from field and/or remotely sensed cover may therefore require disturbance-specific models or detection of vegetation height and transforming 2-D vegetation cover to 3-D vegetation volume.     

Modelling the spatial distribution of montane and subalpine forests in the central Alps using digital elevation models

The analysis of complex interactions between spatial distribution patterns of site factors and vegetation types is crucial for understanding high mountain ecosystems, especially in the view of a changing climate. Therefore, in the present study, a GIS and remote sensing-based approach is followed to produce a vegetation map for a study area in the Western Alps (Switzerland). Two major forest alliances are chosen for analysis: subalpine coniferous forest Vaccinio-Piceion/Larici-Pinetum cembrae and montane oak forest Quercion pubescenti-petraeae. As spatial information on site factors is commonly lacking in mountain areas, the use of a digital elevation model (DEM) is a potential substitute for use in vegetation analyses: it highly correlates with temperature, moisture, geomorphological processes and disturbance factors. Thus, it is important to analyse the capabilities of a DEM for indicating habitat conditions in a landscape characterised by high topodiversity and a patchwork of microclimatic habitats.For the purpose of identifying the potential of landform parameters for the indication of forest habitat structures in the present study, 24 primary and secondary landform parameters have been derived, indicating temperature and moisture distribution, exposure towards wind, snow, etc. Quantitative analyses were performed using statistical means such as contingency correlation coefficients and principal components analysis. The results formed the basis for the development of parallel-epiped-vegetation models (PED) used to simulate the spatial distribution patterns of the subalpine coniferous and the montane oak forest. It can be shown that topographic variables derived from a DEM at a spatial resolution of 25 m are very useful for indicating habitats of large forest types. Additionally potential forest sites in the cultural landscape, removed by human logging, can be reconstructed.Inaccuracies within the simulation results can partly be attributed to the insufficient parameterisation of geomorphologic activity and to poor spatial resolution of the DEM as compared to the vegetation data. Although the lack of information on the human dimension leads to some uncertainties in the interpretation of spatial patterns of vegetation, the exclusive use of topographic variables in vegetation models for the indication of forest habitats is very promising.     

Simulating dynamics of managed monsoon evergreen broad-leaved forest in south China

The forest succession model FORDYN is developed based on TREEDEV model. TREEDEV is a process-based tree growth model, that calculates tree growth based on carbon and nitrogen balance, and is calculated using on the photo-production of leaves, respiration, nitrogen content of all organisms and that in soil, and other losses due to respiration, litter and renewal of stems, branches, leaves and roots. In the FORDYN model succession is divided into three phases called early, middle and late succession, and the transition between these three succession phases is distinguished by a difference in leaf area index. As a verification of the model we used the characteristics and available data of a monsoon evergreen broad-leaved forest in Dinghushan Biosphere Reserve (DHS-BR). The model was validated with natural forest data. In addition, a sensitivity analysis was performed in which 30 independent variables were varied and analyzed in connection with their influence on 16 dependent variables describing forest conditions. The simulation results describe the changes in total biomass, carbon and nitrogen change in plant–litter–soil system of an undisturbed monsoon evergreen broad-leaved forest during succession. We compared these findings with simulation in which different logging management strategies were used. The results show that having a longer logging cycle, delaying the first logging time and a smaller logging fraction the scenario can contribute to a sustainable forest development, while still having a positive economic yield.     

On the use of principal component analysis (PCA) for evaluating interannual vegetation anomalies from SPOT/VEGETATION NDVI temporal series

In this work, we discuss the use of principal component analysis (PCA) for evaluating the vegetation interannual anomalies. The analysis was preformed on a temporal series (1999–2002) of the yearly Maximum Value Composit of SPOT/VEGETATION NDVI acquired for Sicily Island. The PCA was used as a data transform to enhance regions of localized change in multi-temporal data sets. This is a direct result of the high correlation that exists among images for regions that do not change significantly and the relatively low correlation associated with regions that change substantially. Both naturally vegetated areas (forest, shrub-land, herbaceous cover) and agricultural lands have been investigated in order to extract the most prominent natural and/or man-induced alterations affecting vegetation behavior. Our findings suggest that PCA can provide valuable information for environmental management policies involving biodiversity preservation and rational exploitation of natural and agricultural resources.     

Natural resource depletion and the economic performance of local districts: suggestions from a within-country analysis

SUMMARY

Studying the interactions among ecological factors and economic performance measured at the local scale is necessary to suggest policies able to mitigate natural resource depletion in complex ecosystems, like those in the Mediterranean region. The aim of this paper is to explore the relationship among natural resource depletion due to land degradation (LD) and some economic characteristics of local labour systems in Italy, a country where many areas, especially southern regions, are vulnerable. LD was estimated using an environmental sensitive area index, covering the national territory at a fine scale, which takes into account ecological factors such as climate, soil and vegetation. Economic indicators used here consider per capita income, land productivity, production and labour productivity by sector (agriculture, industry and services). A regression analysis was built-up at the LLS scale using an index change over time (1990–2000) as dependent variable and economic variables as predictors. A principal component analysis (PCA) was carried out to synthesise the outputs of the regression analysis. Results indicate a negative relationship among per capita income and LD over the whole study area. However, other variables showed a significant correlation with the dependent variable, highlighting the importance of local factors to increase land vulnerability. In order to clarify the contribution of economic factors to natural resource depletion it is necessary to drive integrate policies to combat LD in dry Mediterranean areas.