基于遥感的漳平市杉松林地表燃物负荷量估测

杉木林和马尾松是分布于漳平市的两种易燃的可燃物类型.其可燃物负荷量的多少决定着林火行为.在利用ASTER计算杉木林小班平均植被覆盖度的基础上估算其郁闭度,通过与地面调查结果比较分析,校正已发生明显变化的小班郁闭度.根据已有估测方程,结合小班数据中的年龄、胸径和树高等因子对杉木地表可燃物负荷量进行了估测,据此对各杉木小班的燃烧性、燃烧强度及火险等级等因素进行初步评价.最后对漳平市另一易燃的可燃物马尾松类型开展地表可燃物负荷量估测及分析.     

漳平市森林可燃物类型变化遥感动态监测

森林可燃物类型的空间分布是林火蔓延、灭火可视化建模与仿真中需要考虑的重要因素之一。在对森林可燃物类型划分的研究进行回顾的基础上,提出了考虑树种信息的分类方法。根据漳平市2003年小班图层中的优势树种信息,获得该市2003年四种森林可燃物类型即竹林、阔叶树、杉木林以及马尾松的空间分布专题图。在此基础上,采用面向对象分类技术,对ASTER影像进行分类,探测每种可燃物类型的内部变化和外部变化。该技术利用影像分割技术构建分类对象,使每个对象具有光谱信息的同时,具备大小、形状、拓扑关系、类别层次等诸多信息。对分类结果进行评价的结果表明,利用面向对象分类技术,充分利用了光谱特征以及类别相关特征,提高了分类精度,分类的面积精度达到89.3%。由于影像分割过程应用了专题图层,对象的边界不会超越专题图层的边界,使得对现有图层的更新尤为容易。最后,利用矢量格式的遥感分类结果对原森林小班图层进行更新,获得新的可燃物类型图层,作为林火可视化模型的输入图层。该研究不仅提供了现势性强的森林可燃物类型图层,而且在不破坏原小班边界的基础上,对发生变化的区域进行刻画。对于森林资源管理者,提供了实地调查过程的目标区域的信息。     

城市地表覆盖组分及植被多样性与热环境关联性分析——以贵阳市为例

本文以贵阳市为研究对象,研究了地表覆盖组分及植被多样性对地表温度的影响。首先基于Landsat8 OLI多时相影像数据在GEE平台上实现了研究区域地表覆盖精细分类;然后结合不同季节8天合成的MODIS温度产品数据,利用时空统计分析、相关分析等方法分析了研究区不同地表覆盖类型地表温度时空分布特征,地表温度与不同地表覆盖组分、地表覆盖多样性和植被覆盖多样性的相关性。结果表明：贵阳市建成区主要分布有常绿阔叶林、常绿针叶林等植被,常绿阔叶林在不同季节对地表温度的降温效应明显,而不透水面对地表温度具有明显的增温效应,尤其以夏季最为显著;地表覆盖多样性与地表温度之间具有较强相关性,其中植被覆盖多样性较植被覆盖率对地表温度的影响更为显著,而不透水面的增加会明显降低植被多样性的影响。因此,要发挥城市绿地对城市温度和热岛效应的调节作用,建议可以适当增加常绿阔叶林的绿化面积,同时在空间上要提升植被多样性水平,能够较大程度改善城市热环境。     

森林资源空间数据更新模型

森林资源小班是森林资源二类调查、统计和经营、管理的基本单位,通过研究其变化的形式和影响其变化的因素,得到的结果可以适用于整个森林资源空间数据。本文依据小班对象的空间变化和属性变化,将其变化归纳为8种基本变化类型。对传统基态修正模型查询速率低下的问题予以解决:使用动态多级索引的方法进行模型的改进,同时,用面向对象的方法对小班空间数据更新的时空信息进行描述,构建了基于版本的多基态修正扩展时空数据模型(VBMBSEA-STDM),很大程度上提高了二类调查数据的准确性和更新速率,更加有利于实现历史查询和分析。在研究的基础上,通过GIS分析凉水实验林场1999年和2009年小班调查数据,得出小班数据的具体变化形式,构建森林资源空间数据更新时空模型,实现历史比较和回溯。     

遥感应用与地理信息系统动态报导 星载SAR应用研究

利用神经元网络森林类型分类系统,在暖温带地区对单时相或多时相的JERS-1SAR图象;对松林蓄积量估测,精度满足二类调查需要。热带林可区分林业与非林业用地、针叶林与阔叶林。     

结合遥感和气象数据的森林火险监测研究——以马来西亚半岛为例

结合遥感和气象数据,基于可燃物引燃能量的物理概念,利用火灾敏感性指数(FSI)进行森林火险监测和评估。计算FSI只需要两个参数,即可燃物的温度和湿度,两者均可以从遥感和气象数据获得。利用遥感数据进行可燃物类型制图,并根据可燃物易燃性对FSI的值进行调整计算。利用MODIS探测的火点数据对马来西亚半岛火灾前9天的FSI进行了比较分析,发现随着火灾发生日期的接近,FSI的值逐渐升高,这个趋势在火灾发生的前4天中表现的尤其明显。结果表明,FSI将遥感数据和气象数据两者紧密结合在一起,能够很好地对森林火险进行监测和预警,是一种性能优越的火险评价指数。FSI可用于计算可燃物的点燃概率以及在不同的生态系统中进行火险的比较,并且因其灵活性和可扩展性,有利于不同类型火险模型的集成和发展。     

GlobeLand 30和自发地理信息的对比分析研究

地表覆盖数据是关于土地利用信息的重要来源,在地理国情监测、生态环境保护等方面发挥着重要的作用,目前遥感影像解译、实地测量是该数据生产的主要手段,但是仍然存在一定的局限性。随着Web2.0、互联网技术以及各种GPS设备的快速发展传播,普通大众也可以参与公众制图,志愿者用户的参与能够有效判定地表类型的空间分布和属性特征,提高地表覆盖制图的分类精度。本文以自发地理信息中最成功的项目OpenStreetMap为例,与中国新研制的全球最高30m分辨率地表覆盖数据产品GlobeLand 30进行对比分析,首先对数据进行相应的预处理和拓扑检查,然后建立两种数据的要素对应关系,最后生成误差矩阵并分析两种数据的一致性。实验结果表明：① OpenStreetMap数据缺失的部分主要是耕地类型,其草地和水体要素比GlobeLand 30更加丰富;② 2种数据的一致性较好为75%左右,其中林地和人造地表的精度较高,耕地和水体次之,草地较差;③ 重点对不一致区域的地表类型进行判断验证,能够发现GlobeLand 30数据中的错误分类,为进一步修改和优化提供依据。本文研究表明,自发地理信息中包含丰富的地表覆盖信息,能够给地表覆盖制图及评价验证带来巨大的发展潜力。     

基于空间抽样的区域地表覆盖遥感制图产品精度评估——以中国陕西省为例

地表覆盖数据是研究气候变化、生态环境、地理国情和人文经济等方面不可或缺的基础信息,因此其质量的优劣将直接影响相关决策的可靠性。本文针对我国研制的首套30 m分辨率全球地表覆盖数据产品,通过分析其海量、多维、非均质等空间特点,对传统的制图产品精度评估方法提出了改进：以地表类型进行分层抽样,样本量的计算采用优化模型并以地类所占面积比为权重逐层分配,样本的布设则考虑层内对象之间的空间相关性,在地表数据自然分布的基础上,通过分析空间相关性指数来提高样本的代表性和精度评估结果的可靠性。因此,本文提出针对区域地表覆盖遥感制图产品的空间分层抽样方法,其将精度评估分成抽样方案（样本的定量估计）和布设方案（样本的空间布设）2个部分,并以中国陕西省地表覆盖产品为例进行区域精度评估实验分析,从全区7大类地表类型数据中抽取1467个样本,经过样本检验与精度计算,该区域产品的总体精度为79.96%,Kappa系数为0.74。实验结果表明,本文提出的基于空间抽样的精度评估方法可行可靠,实验区域产品质量较好,并为后续针对全球范围的地表覆盖产品精度评估方法提供了参考。     

基于空间抽样的区域地表覆盖遥感制图产品精度评估——以中国陕西省为例

地表覆盖数据是研究气候变化、生态环境、地理国情和人文经济等方面不可或缺的基础信息,因此其质量的优劣将直接影响相关决策的可靠性。本文针对我国研制的首套30 m分辨率全球地表覆盖数据产品,通过分析其海量、多维、非均质等空间特点,对传统的制图产品精度评估方法提出了改进:以地表类型进行分层抽样,样本量的计算采用优化模型并以地类所占面积比为权重逐层分配,样本的布设则考虑层内对象之间的空间相关性,在地表数据自然分布的基础上,通过分析空间相关性指数来提高样本的代表性和精度评估结果的可靠性。因此,本文提出针对区域地表覆盖遥感制图产品的空间分层抽样方法,其将精度评估分成抽样方案(样本的定量估计)和布设方案(样本的空间布设)2个部分,并以中国陕西省地表覆盖产品为例进行区域精度评估实验分析,从全区7大类地表类型数据中抽取1467个样本,经过样本检验与精度计算,该区域产品的总体精度为79.96%,Kappa系数为0.74。实验结果表明,本文提出的基于空间抽样的精度评估方法可行可靠,实验区域产品质量较好,并为后续针对全球范围的地表覆盖产品精度评估方法提供了参考。     

融合土地覆盖和土壤水分产品的近地表空气温度空间化方法

空气温度是评价人居环境的重要指标,与人类的生产生活息息相关;其观测对于水文、环境、生态和气候变化等方面的研究具有重要意义。传统的大范围空气温度观测数据一般通过气象站点获取,但由于气象观测站点空间分布离散稀疏的特点,所获取的数据不能精确描述空间连续的空气温度变化情况。因此,实现基于遥感数据的近地表空气温度精准估算具有重要的现实意义。本研究基于精细的地表覆盖类型、空间连续的土壤水分、地表温度（LST）数据,并结合其他辅助数据,构建了近地表空气温度空间化模型,并对近地表空气温度影响因子进行评估,发现地表覆盖类型对近地表空气温度的影响最大,土壤水分为最活跃的影响因素,经验证,模型精度较高,R²接近0.85,RMSE为0.5℃。本研究获取的精确空间连续的近地表空气温度信息,能够充分表达其空间异质性,为农业气象灾害灾变过程监测、农作物生长过程模拟、区域气候变化分析等研究提供良好的近地表空气温度数据支撑。     

Crown fuel characteristics and carbon emission from Japanese red pine stands burned by crown fire in Mt. Palgong, South Korea

Carbon emissions from forest fires are considered an important factor of ecosystem carbon balance and global climate change. Carbon emissions from Japanese red pine stands (Pinus densiflora S. et Z.) burned by crown fire were estimated at Mt. Palgong in Daegu Metropolitan City, and crown fuel characteristics, including crown bulk density, crown base height, and fuel moisture content of Japanese red pine, were analyzed. Total biomass combusted was calculated by subtracting the biomass of burned stands from that of unburned stands exhibiting similar stand structures and site environments. Ten trees in the unburned area and five trees in the burned area were cut by using direct harvesting techniques to estimate crown layer biomass. All biomass sampled was oven-dried and weighed. The dry weight ratios of stems, branches, and needles were 70%, 21%, and 9%, respectively. The available fuel load susceptible to combustion during the crown fire spread was equivalent to 55% of the crown layer biomass. The crown bulk density was 0.24 kg/ m 3 on average. The estimated amount of CO 2 was 23,454 kg CO 2 /ha for the crown layer. These results will be useful for calculating the amount of CO 2 emitted from forest fires and for developing a forest carbon model in P. densiflora forests.     

Carbon storage capacity of different plantation types under sandstorm source control program in Hebei Province,China

Afforestation and reforestation are effective and ecological ways of mitigating elevated atmospheric carbon dioxide(CO2) concentration and increasing carbon(C) storage in terrestrial ecosystems. In this study, we measured the above-ground(tree, herbaceous plants and litter) and below-ground(root and soil) C storage in an aspen plantation(Populus davidiana) monoculture(PD), a larch plantation(Larix pincipis-rupprechtii) monoculture(LP), a pine plantation(Pinus tabulaeformis) monoculture(PT), a larch and birch mixed plantation(L. pincipis-rupprechtii and Betula platyphlla mixed)(MLB), and an apricot plantation(Armeniaca sibirica) monoculture(AS) under the Desertification Combating Program in Hebei Province, the northern China. The objective was to assess the effect of afforestation species on ecosystem C pools of different plantation types. Results showed that C storage of LP stand(258.0 Mg/ha) and MLB(163.4 Mg/ha) were significantly higher than the C storage in PD(45.5 Mg/ha), PT(58.9 Mg/ha) and AS(49.4 Mg/ha), respectively. Soil C was the main carbon pool of the ecosystem C storage in the five plantation stands, ranging from 31.4 Mg/ha to 232.5 Mg/ha, which accounted for 69.0%–90.1% of the total ecosystem C storage. The C storage in tree layer was about 5.2%–23.2% of ecosystem C storage. The herbaceous plants and litter layers contained 1.0%–6.0% and 1.5%–3.3% of ecosystem C storage, respectively. Our results suggest that tree species should be incorporated to accurately develop regional C budget of afforestation program, and also imply that substantial differences in ecosystem C stocks among plantation types can facilitate decision making on C management.     

华南地区典型种植园地遥感分类研究

华南地区种植园地广泛分布,类型混杂多样,导致园地分布信息难以正确获取,为农业管理造成了较大困难。本研究基于Landsat8 OLI数据,通过数据融合、特征优化,应用随机森林算法构建面向对象的种植园地分类规则集,对华南地区典型经济作物香蕉、柑橘、葡萄、蒲葵、海枣、番木瓜和火龙果等进行类别识别,同时对比贝叶斯分类法、K最邻近分类法、支持向量机法、决策树分类法的分类效果。结果表明：数据融合会在一定程度上影响分类结果精度;植株形态、光谱特征接近,种植期交错是影响华南地区典型园地分类精度的重要原因;以中分辨率影像为数据源,面向对象的随机森林算法应用于种植园地分类研究总体精度可达88.05%,Kappa系数0.87,可以有效区分华南地区典型种植园地类别;相比于其他算法,随机森林算法在分类精度、可靠性和稳定性上具有一定优势,可为园地作物生长监测和种植管理提供科学依据。     

Influence of human pressure on forest resources and productivity at stand and tree scales: The case study of Yunnan pine in SW China

This paper examines human impact on stands and individual trees of Pinus yunnanensis growing near the small mountain villages of Pianshui and Yangjuan in southwestern Sichuan Province,China.In an effort to assess whether use of these forests was sustainable,we examined the effects of human use in two ways.First,we directly measured the effect of cutting branches,for fuel and fodder,on tree growth.We hypothesized that branch cutting would negatively impact tree growth.We established 12 plots on four hills and compared 14 pairs of trees,one tree in each pair with an apparently full crown and the other with a considerable portion of the crown removed.Second,we assessed stand and tree properties over a 500 m elevation gradient above the villages where we hypothesized that as elevation increases,stand and tree properties should show fewer human impacts.Although extensive branch cutting reduced the live crown,tree height and diameter,compensatory processes likely enabled trees to recover and to add basal area increments(BAIs) similar to those added by trees with full crowns.Trees and stands close to villages showed less growth and lower basal areas,respectively,than stands and trees at intermediate or distant elevations from villages.Areas relatively close to the villages showed considerable effects of human-related disturbances such as branch cutting,grazing,tree and shrub removal,losses of litter,and human and animal trails.Such areas had increased soil erosion and often loss of the ‘A’ horizon.Stands close to villages had younger trees,lower stand basal areas,smaller basal area increments,and more stumps.Our results suggest an increasingly vulnerable interface between occupants of these two villages and their surrounding forests.     

Biomass carbon storage and its sequestration potential of afforestation under natural forest protection program in China

Based on the data from China′s Seventh Forest Inventory for the period of 2004–2008, area and stand volume of different types and age-classes of plantation were used to establish the relationship between biomass density and age of planted forests in different regions of the country. Combined with the plantation area in the first-stage of the Natural Forest Protection(NFP) program(1998–2010), this study calculated the biomass carbon storage of the afforestation in the first-stage of the program. On this basis, the carbon sequestration potential of these forests was estimated for the second stage of the program(2011–2020). Biomass carbon storage of plantation established in the first stage of the program was 33.67 Tg C, which was majority accounted by protection forests(30.26 Tg C). There was a significant difference among carbon storage in different regions, which depended on the relationship of biomass carbon density, forest age and plantation area. Under the natural growth, the carbon storage was forecasted to increase annually from 2011 to 2020, reaching 96.03 Tg C at the end of the second-stage of the program in 2020. The annual growth of the carbon storage was forecasted to be 6.24 Tg C/yr, which suggested that NFP program has a significant potential for enhancing carbon sequestration in plantation forests under its domain.     

Plantation transformation alternatives determine carbon sequestration capacity – a case study with Pinus massoniana in southern China

It is widely accepted that global warming, which results from the increase of carbon dioxide(CO2) in the atmosphere, has a negative impact on human beings. Forests are the largest terrestrial ecosystem and play an important role in carbon sequestration. Many studies have documented that a mixed-species forest can sequester more carbon than single species forests, depending on the site conditions. Therefore, uneven-aged mixed-species forest management has been receiving more and more attention. In 2008, an experiment with five silvicultural models for Pinus massoniana(Chinese red pine) plantation, i.e., four transformation treatments(A1-A4) and one control treatment(A5) was conducted in the Experimental Center of Tropical Forestry of Chinese Academy of Forestry in Pingxiang City, in southwestern Guangxi Zhuang Autonomous Region, southern China. The four transformation treatments(A1-A4) enriched Castanopsis hystrix, Manglietia glance, Erythrophleum fordii and Quercus griffithii with differed richness and composition after thinning(removed 70% of trees), while no silvicultural treatment was used in the control treatment A5. In this study, we compared the carbon sequestration capacity of these five silvicultural models based on periodic annual increment and growth rate. Our results indicated that all the transformation treatments performed significantly better in carbon sequestration than the control treatment. A significant difference was also observed amongst the transformation treatments. Moreover, the transformation treatment A1 with enrichment species Castanopsis hystrix(350 trees·ha^-1) and Manglietia glance(350 trees·ha^-1) was determined to be the optimal model for maximum carbon sequestration because of its high tree-level growth rate and high economic value of enriched plantings, which could be popularized in other places. Our results further confirmed that management using mixed-species forests is a better approach to combat climate change than using monoculture forests.     

Wet Canopy Evaporation Rate of Three Stands in Western Sichuan, China

The wet canopy evaporation rate (Er) was calculated by Penman-Monteith combination model based on three assumptions and with meteorological variables 2 m above the canopy in three stands, dominated by spruce (SF), fir (FF) and birch (BF) trees, respectively, in the subalpine forests in western Sichuan, China over a growing season. The total amount of the E was 44.5 mm for SF, 88.5 mm for FF and 57.8 mm for BF, accounting for 9.2%, 16.6% and 10.2% of the gross rainfall, respectively, in the measuring period. There was the highest average monthly Er and percentage of E to gross rainfall for FF compared with SF and BF.Mean Er was 0.097 mm h^-1 (ranging from 0.028 to0.487 mm h^-1), 0.242 mm h^-1 (from 0.068 to 0.711 mm h^-1) and 0.149 mm h^-1 (from 0.060 to 0.576 mm h^-1) for SF, FF and BF, respectively. The highest average monthly Er occurred in June was 0.120 mm h^-1 for SF, 0.317 mm h^-1for FF and 0.169 mm h^-1 for BF, and the lowest value in October was 0.083 mm h^-1 for SF, 0.187 mm h^-1 for FF and 0.101 mm h^-1 for BF, respectively. The averages of Er from 8:00 to16:00 were significantly higher than those from 0:00 to 8:00 and from 16:00 to 0:00 for the three stands. The marked daily and monthly differences of Er were contributable to the variations of solar radiation, air temperature and relative humidity above the canopy.     

Effects of environmental factors on tree seedling regeneration in a pine-oak mixed forest in the Qinling Mountains, China

Natural seedling regeneration and tree establishment are affected by various environmental factors.In this study,we established eight,eight,six,and four independent forest stands(each stand was further divided into five subplots) respectively based on the altitudinal gradient,stand density,slope location,and slope aspect to investigate the effects of environmental factors on tree seedling regeneration in a pine-oak mixed forest.The results indicated that the seedling density was significantly higher at altitudes of 1,283 m to 1,665 m,whereas the sapling density did not differ with altitudes.The seedling and sapling density decreased significantly at 1,835 m.The seedling densities on the upper slopes were much higher than those on the middle and lower slopes,whereas the sapling density had no difference.The sapling density decreased gradually from the southwest(20°-75°) whereas it increased on the shady slopes to the northeast(40°).The seedling density increased from southwest(20°) to northeast(40°).The seedling and sapling densities increased with the stand density(850 trees ha-1to 1,525 trees ha-1) whereas the sapling density was significantly lower in stands(1,900 trees ha-1).Principal components analysis showed that the slope aspect and stand density had more important roles in tree regeneration in this study region compared with the other two factors.Therefore,our findings suggest that it will be beneficial to keep stands at a moderate density on shady slopes.Appropriate thinning of higher density stands is also expected to promote the natural regeneration of pine-oak mixed forest.     

基于改进线性光谱分离模型的植被覆盖度反演

线性光谱分离技术可以有效地提取像元水平上植被或其他端元(影像中的地物)的相对百分比,但是目前该技术在多光谱宽波段影像数据应用中,由于波段数量、波段宽度等的限制,估算精度离定量研究的水平仍有一定差距。鉴于此,本文提出了一种改进的线性光谱分离方法,该方法在对影像进行土地覆盖分类基础上进行分离,一方面同类土地覆盖类型内同种地物的光谱变异相对较小,更有利于端元选取;另一方面,分影像的地物种类数量明显少于整幅影像,更容易满足模型的适用条件,从而突破了波段数量限制,同时使地物光谱分离更具针对性,经过验证,该方法较传统分离方法相比植被覆盖度的反演精度可提高6.4%,用该方法实现了研究区的植被覆盖度的定量反演并对研究区植被覆盖度的空间结构进行了分析。