Evolutionary fingerprint, phylogenetic and forest structure of tropical montane Atlantic cloud forests along an elevation gradient

Environmental conditions can change markedly over geographical distances along elevation gradients,making them natural laboratories to study the processes that structure communities.This work aimed to assess the influences of elevation on Tropical Montane Cloud Forest plant communities in the Brazilian Atlantic Forest,a historically neglected ecoregion.We evaluated the phylogenetic structure,forest structure(tree basal area and tree density)and species richness along an elevation gradient,as well as the evolutionary fingerprints of elevation-success on phylogenetic lineages from the tree communities.To do so,we assessed nine communities along an elevation gradient from 1210 to 2310 m a.s.l.without large elevation gaps.The relationships between elevation and phylogenetic structure,forest structure and species richness were investigated through Linear Models.The occurrence of evolutionary fingerprint on phylogenetic lineages was investigated by quantifying the extent of phylogenetic signal of elevation-success using a genus-level molecular phylogeny.Our results showed decreased species richness at higher elevations and independence between forest structure,phylogenetic structure and elevation.We also verified that there is a phylogenetic signal associated with elevation-success by lineages.We concluded that the elevation is associated with species richness and the occurrence of phylogenetic lineages in the tree communities evaluated in Mantiqueira Range.On the other hand,elevation is not associated with forest structure or phylogenetic structure.Furthermore,closely related taxa tend to have their higher ecological success in similar elevations.Finally,we highlight the fragility of the tropical montane cloud forests in the Mantiqueira Range in face of environmental changes(i.e.global warming)due to the occurrence of exclusive phylogenetic lineages evolutionarily adapted to environmental conditions(i.e.minimum temperature)associated with each elevation range.

Influence of topographic aspect on floristic diversity, structure and treeline of afromontane cloud forests in the Bale Mountains,Ethiopia

We studied the influence of east and west aspects on floristic composition, diversity, structure and treeline of afromontane cloud forests at Rira in the Bale Mountains, southeast Ethiopia. In addition, we studied how aspect relates to and/or interacts with other topographic and edaphic factors in influencing vegetation diversity. Strati- fied systematic plot sampling was used to survey the floristic composition, diversity and structure of forests on east- and west-facing slopes. The sample plot size was 20 x 20 m and a total of 36 plots were inventoried. A total of 72 composite soil samples were collected and analysed. Woody species richness of the forest on the east-facing slope was 1.7 times higher than on the west-facing slope. Shannon, Simpson and Log-series alpha diversity indices and evenness of forests on the east-facing slope were sig- nificantly higher than on the west-facing slope. NMDS ordination indicated that the east- and west-facing slopes formed two clusters of species and aspect explained 55.2 % and 10.4 % of the variation in species richness and abun- dance, respectively. There was no significant difference between aspects in stand structure except in dominantheight, which was higher on the east-facing slopes. There was significant interaction between aspect and elevation in influencing woody species diversity. The four plant com- munity groups, which were identified using cluster and indicator species analysis were represented differently on the east and west aspects. The treeline on the east-facing slope （3352 m） was located about 110 m higher than on the west-facing slope （3240 m）. Soil moisture deficiency was unlikely to be a limiting factor on either site. Near the equator, east-west aspect was shown to have considerable impact on floristic composition, diversity, structure, and treeline position of montane forests. Diurnal cloud move- ment patterns and its impact on microclimate of slope aspect should be taken into account in future studies of cloud forest diversity, structure, and treeline position.

Land cover change and carbon stores in a tropical montane cloud forest in the Sierra Madre Oriental,Mexico

Tropical montane cloud forest is one of the ecosystems with the highest biomass worldwide, representing an important carbon store. Globally its deforestation index is –1.1%, but in Mexico it is higher than –3%. Carbon estimates are scarce globally, particularly in Mexico. The objective of this study was to simulate future land-cover scenarios for the Sierra Madre Oriental in Mexico, by analyzing past forest cover changes. Another objective was to estimate stored carbon in the two study areas. These objectives involve the generation of information that could be useful inputs to anti-deforestation public policy such as the REDD+ strategy. Remote sensing was used to measure land cover change and estimate carbon stocks. Satellite images from 2015, 2000 and 1986 were used, and Dinamica EGO freeware generatedmodels of future projections. Between 1986 and 2015, 5171 ha of forest were converted to pasture. The annual deforestation rates were –1.5% for Tlanchinol and –1.3% for the San Bartolo Tutotepec sites. Distance to roads and marginalization were highly correlated with deforestation. By 2030, an estimated 3608 ha of forest in these sites will have been converted to pasture. Stored carbon was estimated at 16.35 Mg C ha-1 for the Tlanchinol site and 12.7 Mg C ha-1 for the San Bartolo site. In the Sierra Madre Oriental deforestation due to land cover change(–1.4%) is higher than levels reported worldwide. Besides having high values of stored carbon(14.5 Mg C ha-1), these forests have high biodiversity. The models' outputs show that the deforestation process will continue if action is not taken to avoid the expansion of livestock pasturing. This can be done by paying incentives for forest conservation to the owners of the land. The results suggest that REDD+ is currently the most viable strategy for reducing deforestation rates in tropical montane cloud forests in Sierra Madre Oriental.

论新加坡滨海南花园的室内生态可持续设计——以“云雾森林”Cloud Forest园为例

位于新加坡河入海口的滨海南花园,作为新加坡向"花园中的城市"发展目标的起点之一,其中的代表性景观建筑"云雾森林"(Cloud Forest)充分展示了可持续性室内景观的运行与维持。通过对"云雾森林"的设计理念、结构、材料、区域组成及室内节能动力系统进行全面的分析,旨在能为读者全方位的展示一个可持续性室内景观设计的成功案例。同时,也向读者深度解读了该园林所导引的可持续互动之旅,凸显了作为公共空间的教育意义以及公众参与价值。

Effect of Land Use Change on Carbon Content and CO₂Flux of Cloud Forest Soils, Santa Elena, Costa Rica

We investigated the effects of land-use changes on soil carbon storage and soil CO2 flux by comparing soils from mature cloud forest and 31-year-old secondary forest, both in the Santa Elena Forest Reserve, a municipallyowned reserve at an elevation of 1600 to 1700 m near the town of Monteverde, and a clear-cut pasture near the reserve. Soils in the mature forest exhibit only weak horizonation but typically thick A horizons;they also consistently yield the highest carbon contents in the upper 30 cm. Soil CO2 flux was the highest in these soils, but also displayed the highest spatial variability. Secondary forest soils contain substantially less soil carbon than mature forest soils, but more than pasture soils. CO2 flux in the secondary forest soils was more similar to that of the mature forest, but displayed lower spatial variability. The pasture soils contain less soil carbon and produced lower CO2 flux levels than either of the forest soils. The pasture soils typically contain a well-defined coarse sandy layer 10 to 20 cm below the surface that we interpret as a sediment layer deposited across much of the landscape following a widespread erosion event, likely a consequence of the clear-cutting. Soil nitrogen concentrations are more than an order of magnitude lower than soil carbon concentrations, and display no trends between the different landscapes examined. Our preliminary results suggest that reforestation does restore soil carbon to clear-cut landscapes, but returning soil carbon levels to pre-land use levels occurs at a time scale of centuries, rather than decades.

热带云雾林植物功能性状对环境变化的年际响应

为了研究植物功能性状对环境变化的响应有助于理解植物对环境的适应策略,以海南热带雨林国家公园霸王岭片区的热带云雾林为研究对象,在2011年及2021年对21个20 m×20 m固定样地内DBH≥5 cm木本植物进行群落学调查,测定个体、物种和群落水平的植物功能性状及环境因子,分析植物功能性状和环境条件的变化,探讨环境因子对功能性状的影响规律.结果表明:(1)与2011年相比,2021年热带云雾林土壤全氮(TN)、有效氮(AN)含量显著增加,土壤有机质(SOM)、全磷(TP)、有效磷(AP)含量显著减少.空气最高温呈上升趋势,平均空气相对湿度、最低空气相对湿度呈下降趋势.对环境因子进行主成分分析,2011年负荷值较大的环境因子分别为土壤全氮、土壤全磷、土壤有效氮及最低空气相对湿度,2021年负荷值较大的环境因子为土壤有机质、土壤全磷、最高空气温度及最低空气相对湿度;(2)与2011年相比,2021年植物叶绿素含量和叶片厚度在个体水平显著增大,比叶面积显著减少;叶片厚度、叶绿素含量、茎全氮含量、叶片全氮含量在物种水平显著增大,茎可溶性糖含量、叶全磷含量显著减少;叶绿素含量、叶片厚度、茎全氮含量、茎全磷含量、叶全氮含量在群落水平显著增大,茎密度、茎可溶性糖含量、叶全磷含量显著减小;(3)多元逐步回归分析发现,土壤养分含量(如土壤有机质、全磷含量、有效氮含量等)和气候因子(如最高空气温度、最低空气相对湿度等)共同影响植物个体、物种及群落水平功能性状变化,土壤养分因子作用较大,土壤磷胁迫的加剧是影响各个群落组织水平功能性状变化关键因子.

森林火灾蔓延模拟及应急路径规划优选

【目的】利用云计算手段模拟林火蔓延过程,规划火灾应急路径,以促进救援力量快速、高效到达指定火灾地点,为布置林火阻隔措施预留充足时间。研究也为吉林省通化市东昌区的林火应急响应和预防规划提出建设性意见。【方法】以吉林省通化市东昌区为研究区域进行实地调研和模型推演。着眼于林火点单源和多源差异性,利用应急路径规划模型、Dijkstra算法模拟计算单源林火情况下最短救援路径,并纳入多源林火优先级理论,实现救援资源利用最大化。采用FARSITE模型对区域实例森林火灾进行蔓延速率模拟,获取林火发生后各阶段实时数据,综合考虑森林火点附近资源布局情况以及道路交通能力,进行救援路径的推算和防火道路的布局,以验证方法的可行性,并对森林火灾应急响应救援路径进行智能优选。【结果】在一般森林火灾发生时单个着火点的公式基础上,为2个及以上着火点引入优先级理论,再根据火情严重程度通过Dijkstra算法进行依次推演。此外,增加μ作为通行能力系数,以考虑路况、道路属性对救援通行时间的影响,得到新的路径公式。通过FARSITE模型,对2010年“5.3”一般森林火灾(20100503-220502-01)进行蔓延推演,在火源点起火5 h后,如果没有得到有效的人为干预,火情将演变为重大森林火灾,并蔓延到人居地区。由于风向影响,火源点东北侧的火情优先级高于北侧,若有条件同时进行救援,应急资源应按优先级倾斜。【结论】模拟结果表明,通过应急路径规划模型、Dijkstra算法和优先级理论,可以提高救援路径的获取速率,提前进行资源调配和人员安排,FARSITE模拟结合防火道路规划,联系火灾发生后各时间段的蔓延情况,能有效提升森林火灾决策效率,事半功倍地形成林火应急响应和预防规划方案。

激光点云滤波技术在植被分布调查中的应用

研究了机载激光雷达点云滤波技术在植被分布调查中的应用方法及效果。以RY县PK林场为案例分析对象,首先根据地质气候条件及测绘目标成像需求对航测设备、雷达设备进行选型;其次利用点云滤波技术,对PK林场点云数据进行采集与建模,并实现地块识别与分类;再次,针对十年间地块植被变异情况进行点云差异化建模,分析地块内植被空间变化情况;最后对研究成果进行误差分析,得出运用无人机航测设备、激光雷达测绘设备、点云模型建模系统等,可以为森林治理、土地调查等工作提供可靠的数据基础的结论。

无人机激光雷达杉木人工林碳储量估测

森林碳储量是衡量森林生态系统基本特征的重要指标,传统碳储量估测方式需要耗费大量的时间、人力、物力等,因此利用无人机搭载激光雷达获取机载激光数据并建立森林碳储量估测模型,为获取区域内森林碳储量提供参考,从而更好地对森林资源进行监测。以杉木人工林为研究对象,利用机载激光雷达点云数据获取点云特征变量,以此作为建模变量进行模型研建。通过不同筛选变量方式选择建模变量,分别建立非线性回归模型、线性回归模型和随机森林模型并进行模型评价,通过对比模型的决定系数(R^(2))、均方根误差(RMSE)和平均绝对误差(MAE)选出最优模型,用于后续研究。结果表明,1)3种模型拟合效果最佳的为随机森林模型,其R^(2)为0.94,RMSE为0.53 t/hm^(2),MAE为0.44 t/hm^(2)。非线性回归模型R^(2)为0.71,RMSE为0.66 t/hm^(2),MAE为0.56 t/hm^(2);线性回归模型R^(2)为0.67,RMSE为0.88 t/hm^(2),MAE为0.80 t/hm^(2)。2)共计提取101个点云特征变量,通过变量筛选发现高度变量、密度变量无论是相关性还是重要性,均大于强度变量。3)对比是否加入优选变量对随机森林模型预测精度的影响,加入优选变量后模型R^(2)没有变化,但RMSE与MAE小于未加入优选变量,精度有所提高。利用机载激光雷达点云数据获取的点云特征变量建立模型,对比非线性回归和线性回归模型,随机森林模型精度最高,用其估算得到研究区内碳储量为480.65 t/hm^(2),与实测值最相近。因此随机森林模型更适合于区域森林碳储量的估测。

海南热带云雾林幼苗动态参数对磷添加的响应

该研究以海南霸王岭热带云雾林幼苗群落为对象,结合原位磷添加实验和动态监测,设置对照组ck(0.0 gm^(-2)a^(-1))及5个不同水平的磷添加处理组,按磷添加浓度由低至高依次为p1(1.0gm^(-2)a^(-1))p2(2.0gm^(-2)a^(-1))、p3(4.0 gm^(-2)a^(-1))p4(8.0gm^(-2)a^(-1))p5(16.0gm^(-2)a^(-1))。从群落植物的生长、死亡与更新动态,探究热带云雾林幼苗对磷添加的生态响应。主要结果如下:基于幼苗动态监测数据的统计参数显示,在群落水平上,对照组与处理组之间、不同磷添加处理组之间的生长率、死亡率与增补率均无显著差异;种群水平上,处理组幼苗生长率(0.94±0.52)显著高于对照组(0.84±0.54),死亡率和增补率无显著差异。随磷添加浓度升高,种群动态参数的频数分布格局发生了一定改变,种群死亡率和增补率均呈“倒J型”分布(即测度值为0的种群居多数)但程度分别有所降低与升高;生长率均近似正态分布但偏斜程度发生改变。此外,不同磷添加处理下幼苗种群的生长率与死亡率、死亡率与增补率之间的关联性均不显著;增补率与生长率之间均呈负关联(r=-0.356~-0.140),但随磷添加浓度升高由显著转变为不显著。上述发现有望为高海拔热带森林生态系统的管理和动态预测提供科学依据。

河北省泥石流灾害易发性云模型评价方法:以邢台赵沟村泥石流为例

建立定量化的泥石流易发性评价方法对于泥石流灾害的防治工作具有重要意义。通过对河北省130处泥石流沟的详细勘查和数据整理,分析总结了河北省泥石流的区域性特征。在此基础上,选取评价因子,引入随机森林分类方法确定评价因子的权重,采用高斯正向云模型建立了评价因子与泥石流易发性等级的定量关系,最后,通过工程实例证明新方法的合理性。研究表明:河北省西部、东北部泥石流灾害点分别被600 mm和700 mm年降雨量等值线包围,北部地区泥石流灾害点则多位于年降雨量等值线400 mm与500 mm之间。河北省泥石流形成关键因子为降雨强度、汇水面积、沿沟松散物储量、沟岸山坡坡度、流域相对高差和沟谷比降等;河北省泥石流类型主要为中小型沟谷暴雨-水石流;工程实例分析表明基于随机森林赋权的泥石流易发性云模型评价方法评价结果与实际情况相符。

地面激光雷达扫描参数与点云简化对林木参数提取的影响

【目的】林木参数是森林蓄积量、森林生物量估算的基础指标,传统的人工调查方式费时费力,已难以适应新形势下数字化森林资源监测技术的要求。地面激光雷达扫描技术能够获取小尺度高分辨率的林分内部结构信息,为林分环境条件下林木胸径、树高提取提供一种新的思路。【方法】以芦头实验林场杉木林样地为研究对象,针对FARO Focus 3D X330三维激光扫描仪设计了7种不同的扫描组合方式对样地进行扫描,提出象限角点云简化思路进行参数提取和精度评价,探究不同扫描组合方式对林木胸径、树高参数提取精度与效率的影响。【结果】1)当扫描分辨率为1/2、质量为4X时,胸径参数提取精度最高;当扫描分辨率为1/4、质量为4X时,树高参数提取精度最高。2)在林木参数提取结果没有显著性差异的前提下,扫描分辨率为1/4、质量为4X的扫描参数工作效率最高。3)选取同时兼顾精度和效率的1/4扫描分辨率、质量4X的扫描结果,进行象限角点云简化,简化的点云能够准确地提取出林木胸径参数。【结论】研究结果对于具有相同或相似地理条件和树种的林地选择扫描参数和点云简化方式具有重要参考价值,可以提高内业工作效率,同时也为地面激光雷达野外样地调查提供方法和技术参考。

Tree community composition,structure and diversity along an elevational gradient in an Andean forest of Northern Ecuador

The Andean forests of northern Ecuador are known for their high levels of plant diversity relative to the area they occupy.Typically,these forests grow on steep slopes that lead to dramatic habitat gradients across short distances.These extreme habitat gradients make the Andean forest ecosystem an excellent natural laboratory for understanding the effect of elevation on forest community diversity,structure and composition.We established 31 plots(50 m×5 m)which are divided between two elevational transects in the cloud forest of the Siempre Verde Reserve in the western foothills of the Andes Mountains of northern Ecuador.All trees and tree ferns with a diameter at breast height(dbh)≥5 cm were measured and identified.We examined changes in community composition,structure,and diversity along and between the elevational transects and three elevational zones:low(2437–2700 m),middle(2756–3052 m),and high(3163–3334 m).We found four main trends associated with the elevational gradients at this site:(1)community composition differed between the two transects and among the three elevational zones according to N-MDS,ANOSIM,and percentage of shared species,with some species having limited distributions,(2)metrics of community structure showed opposite relationships with elevation,depending on the transect,with the only significant relationship(negative)found between basal area and elevation in the open trail transect,(3)alpha diversity,in general,peaked at mid-elevations,and(4)beta diversity consistently increased with distance between plots along elevation.The complexity of changes in community composition,structure,and alpha diversity along elevation may be related to the heterogeneity of the environment on a local scale,such as topography,soil composition,and even human impact,or to dispersal limitation and should be investigated further.These changes in community composition and the relatively high beta diversity found at this site exemplify the biological complexity of montane forest,reinforcing arguments from other studies on the importance of their conservation.

应用UAVSAR数据及改进极化水云模型对热带森林地上生物量反演

针对热带森林地上生物量遥感估测中容易饱和区域大尺度森林地上生物量估测精度低的问题,以非洲加蓬中部的洛佩达国家公园为研究区,以NASA提供的L波段全极化机载合成孔径雷达(SAR)数据和LiDAR网格化森林地上生物量产品为数据源开展森林生物量估测方法研究。采用极化分解方法提取森林的多种散射机制,从中选择反映森林结构差异的地面散射特征和森林体散射的特征构建体-地散射比,采用极化水云模型(PWCM)进行森林地上生物量反演和精度评价。为了提高PWCM模型的适应性,建模过程中根据体散射分量(V_(ol))分段进行模型的参数优化。结果表明:以Freeman三分量极化分解后得到的体散射(V_(ol))、表面散射(O_(dd))、地-干散射(D_(bl))为基础构建的6个体-地散射比在极化水云模型估算森林地上生物量中,以μ_(VG2)作为体-地散射比时估测效果最好,模型决定系数(R^(2))为0.60,均方根误差(R_(MSE))为127.78 Mg/hm^(2);在此基础上,进一步根据体散射分量分段优化极化水云模型,模型决定系数(R^(2))增加到0.74,均方根误差(R_(MSE))降低了约20%,预测精度从50.76%提升至60.28%,并改善了低值高估、高值低估问题,在地上生物量高达450 Mg/hm^(2)时未出现饱和现象。

近地面机载激光雷达在森林资源调查监测中的应用

从近地面机载激光雷达工作原理出发,分析该技术在林业调查中获取地形地貌、地表植被覆盖信息方面存在的优势,结合自主研发的点云数据处理软件所涉及的方法,描述了从机载激光雷达数据获取单木信息、林木树高、冠幅、株数和郁闭度的技术理论,最后根据机载激光雷达在林业上的应用情况,与传统的林业调查监测技术进行比较,得到当前机载激光雷达在调查树高、郁闭度、株数等方面有自身优势,可以替代部分传统工作,但在预测树高胸径模型、蓄积量调查、点云数据处理算法等方面仍然有待技术和理论进一步完善和改进的结论。

协同随机森林方法和无人机LiDAR空谱数据的盐沼植被“精灵圈”识别

“精灵圈”作为一种典型的空间自组织结构,对盐沼植被生态系统及其功能有重要影响。获取“精灵圈”的空间格局及时空变化,可为厘清其生态演化机理提供重要科学支撑。本文基于随机森林机器学习方法,结合无人机激光雷达(LiDAR)点云的空间信息与光谱信息,对盐沼植被“精灵圈”进行智能识别与提取。首先,利用激光雷达方程和Phong模型,消除距离、入射角以及镜面反射效应对强度数据的影响,并且通过校正后强度数据滤波分离植被点云与地面点云。然后,构造系列空间特征及几何变量,利用随机森林算法,对植被点云中的正常植被和“精灵圈”进行分类。结果表明:该方法无需人工经验设置参数,能够精确地从无人机LiDAR三维点云数据中快速自动识别“精灵圈”,总体精度为83.9%。本文为“精灵圈”时空分布反演提供了一种高精度的方法,也为基于机器学习的三维点云数据处理提供了技术借鉴。

Masting effect on alpha and beta avian diversity in fragmented forests of relict?endangered Mexican Beech(Fagus grandifolia subsp.mexicana)

Background:Tropical montane cloud forests are one of the most important hotspots on Earth and show presence of relict-endemic and endangered species,representing about 14%of the total tropical forest worldwide.Synchronous seed production or masting in tropical montane cloud tree species is a widespread reproductive strategy of decidu-ous and evergreen broad-leaved tree associations to decrease costs of reproduction and ensure offspring.Masting event maintains a high avian diversity,which can be modified by phenological process(seed production and non-seed production).Methods:The main aim of this study was to assess alpha and beta avian diversity and whether the composition of the trophic guild modifies among phenological processes and between two fragmented relict-endangered Mexican Beech(Fagus grandifolia subsp.mexicana)forests(Medio Monte and El Gosco)in the Mexican state of Hidalgo.In addi-tion,we evaluated beechnut production.Results:We recorded 36 bird species,11 of them included in some conservation risk status,and 5 endemic species.Alpha diversity values were dissimilar in avian richness(q=0)among phenological processes and between frag-mented beech forests.Avian communities among three phenological processes and between fragmented forests were structurally similar,dominated during immature seeds the Brown-backed Solitaire(granivores-insectivores-frugivores);during mature seeds the White-crowned Parrot(Pionus senilis,granivores-frugivores);and the Dwarf Jay(Cyanolyca nana,insectivores)was abundant during low seed quality.The complementarity index was high among phenological processes and low between forests.We found a high bird turnover value between immature seeds—mature seeds and during mature seeds—low seed quality.Furthermore,a similar pattern was recorded between the two study forests.Seed production showed a high number of undamaged beechnuts in Medio Monte,while in El Gosco beechnuts were attacked by insects.Conclusions:Our results reflect that masting phenological process and contrasting study forests'structure influence the shifts in alpha and beta diversity of seed and non-seed bird consumers.Our study reaffirms the importance of continuing studies throughout masting in all the Mexican Beech forests to address regional efforts in preserving the relict-ecological interactions.

基于无人机多维数据集的森林地上生物量估测模型研究

森林地上生物量(Aboveground biomass,AGB)是评价森林生长情况的重要指标。基于数字航空摄影(Digital aerial photography,DAP)生成的二维和三维数据,分别计算了41个点云高度变量和16个可见光植被指数,利用6种回归算法(随机森林(RF)、袋装树(BT)、支持向量回归(SVR)、Cubist、类别型特征提升(CatBoost)、极端梯度提升(XGBoost))分别构建了单一变量集和综合变量集AGB估测模型,探索了不同变量对于AGB估测模型的贡献。研究结果表明光谱数据集和点云数据集AGB预测模型精度最高分别为Cubist和XGBoost,R^(2)分别为0.5309和0.6395。组合数据集最高精度模型为XGBoost,R^(2)达到0.7601,XGBoost模型具有更高的AGB估测稳定性。研究还表明6种机器学习模型的贡献主要取决于所考虑的回归方法,所选择的特征个数和特征对模型的重要性在不同的模型中并不一致。DOM光谱特征在AGB的估测中具有更高的重要性。总体来说,二维和三维数据的结合能够有效提高森林AGB估测精度,基于无人机倾斜摄影获取的RGB影像能够实现森林AGB的快速无损估计。

基于GEE和Sentinel-2影像的杭州城市湿地精细化分类研究

基于Google Earth Engine(GEE)云平台,使用Sentinel-2卫星影像数据,采用随机森林算法面向杭州城市湿地类型进行遥感分类研究.研究结果表明:在进行湿地类型精细化分类时,相比于使用单一特征信息分类,多特征信息的组合可以显著提高分类精度,最优特征组合包括传统光谱特征、红边光谱特征、变换特征、纹理特征和地形特征,其总体精度为81.2%,Kappa系数为0.75;不同特征信息对于不同湿地类型的提取具有重要作用,传统光谱特征对于滩地和养殖池的识别非常有利,而红边光谱特征和变换特征分别对草本沼泽和湖泊的识别更为有效.此外,红边光谱特征和纹理特征的组合有利于运河/水渠的识别.通过特征对比实验可为湿地精细化分类的案例研究提供参考,湿地分类结果可为杭州城市湿地信息的遥感识别提供数据支撑.

山西天龙山自然保护区森林防火智慧监测系统架构及应用

通过构建森林防火智慧监测系统,解决山西天龙山自然保护区以传统可见光视频监控系统监测方式为主、卫星监测系统应用效率低下、智能烟火识别功能不健全、大数据预警监测能力不足等问题。森林防火智慧监测系统以物联网技术、云计算技术、大数据技术、低轨道卫星技术为依托,构建“端—边—云”协同林火监测架构,协同大数据处理中心,最大程度提高保护区森林防火监测能力。通过实践证实,森林防火智慧监测系统是一个具有高度智能化和可视化的监测平台,能够实现保护区森林火灾精准定位、信息快速响应、数据实时同步以及系统及时报警等功能,未来也将在森林资源监测、野生动植物保护、林业重点工程监管等领域得到广泛应用。