共查询到20条相似文献，搜索用时 843 毫秒
1.
《国土资源遥感》2016,(2)
科学合理地定量评估林火烈度,对揭示林火干扰下森林生态系统的变化,以及植被的恢复与管理具有重要意义。以美国科罗拉多大峡谷国家公园北缘的Poplar Fire为实验区,利用Landsat5 TM影像,结合实地调查的综合火烧指数(composite burn index,CBI),分析评价了归一化差值植被指数(normalized difference vegetation index,NDVI),归一化火烧指数(normalized burn ration,NBR),差分归一化植被指数(differenced normalized difference vegetation index,ΔNDVI)和差分归一化火烧指数(differenced normalized burn ration,ΔNBR)4种遥感指数对林火烈度评估的适应性。结果表明,4种遥感指数对识别不同等级林火烈度存在一定的差异。在未过火区和轻度火灾区,单一遥感指数的精度略高于差分遥感指数,其中NBR的提取精度最高,分别达到了66.7%和80%;在中度火灾区和重度火灾区,差分遥感指数的精度高于单一遥感指数,ΔNBR的提取精度最高,分别达到了100%和90%。总体上,基于差分遥感指数的林火烈度制图精度总体高于单一遥感指数,其中ΔNBR的总体制图精度最高,达到了86.2%。因此,ΔNBR是林火烈度分析与评估的适宜遥感指数。 相似文献
2.
川西亚高山针叶林位于中国西南地区,受多云、多雨、多雾的影响,难以通过卫星影像进行植被分类的研究。为了解决这一难题,本研究选取川西亚高山针叶林的典型区域王朗自然保护区作为研究区,使用多旋翼无人机获取研究区域北部高分辨率RGB影像,结合卷积神经网络进行植被分类;为进一步挖掘卷积神经网络在无人机遥感影像上的潜力,选择语义分割方法 （U-Net）进行分类,并根据不同分辨率的无人机影像和不同尺寸下的样本集构建植被分类模型,建立森林指纹库。结果表明：（1）结合无人机可见光影像和CNN模型进行分类能够获得高精度分类结果。在空间分辨率为5 cm,尺寸为256×256像素的情况下达到最优,总体精度为93.21%,Kappa系数为0.90;(2)选择合适的尺寸大小能够提高模型的分类精度。在5 cm的空间分辨率下,尺寸为128×128像素的模型总体精度为82.30%,Kappa系数为0.76;尺寸为256×256像素的模型总体精度为93.21%,Kappa系数为0.90;(3)超高空间分辨率的升高对模型精度的提升是有限的。当空间分辨率从10 cm升到5 cm时,模型的总体精度提高了0.02,Kappa系数提... 相似文献
3.
《国土资源遥感》2021,(3)
不透水层是表征人类活动的重要指标,及时精确的不透水层信息对区域生态环境保护有重要意义。以山东省东营市为研究区,探索了一种基于多源Sentinel-1/2影像和随机森林的不透水层提取方法。通过对比实验发现,随机森林结合地表反射率特征、纹理特征和后向散射系数能够降低暗不透水层和亮不透水层与裸土的混淆现象,可以有效改善不透水层的估算精度(总体精度达到93.37%,Kappa系数达到0.925 8)。研究结果揭示了随机森林协同Sentinel-1和Sentinel-2数据在不透水层信息提取方面有着广泛的应用前景,为融合多源数据对黄河三角洲区域遥感监测提供了参考。 相似文献
4.
提出了一种基于深度学习技术的遥感分类方法，它能有效解决中分辨率影像在分类过程中出现的像元混分问题。研究选用2016年5月12日武汉市Landsat 7 ETM+遥感影像，基于GoogleNet模型中的Inception V3网络结构，借助迁移学习方法，构建出遥感分类模型，实现了对武汉市主城区4类典型地物（不透水层、植被、水体和其他用地）的自动分类提取，并将分类结果与传统最大似然分类（ML）结果进行了对比分析。研究表明：基于深度学习方法的遥感影像总体分类精度高达88.33%，Kappa系数为0.834 2，明显优于传统ML方法总体分类精度83%和Kappa系数0.755 0，而且有效抑制了地物在分类过程中出现的像元混分现象。 相似文献
5.
6.
南美洲湿地面积广且类型多样，但湿地制图相关研究匮乏，通过遥感手段可为南美洲全域湿地制图提供科学技术支撑。本研究依托GEE (Google Earth Engine)平台面向南美洲湿地提出一种多源多特征集成的湿地制图方法。研究选取南美洲典型湿地地区为研究区，首先利用已有土地覆盖数据集提出一种有效的湿地样本采集流程以保证样本质量，其次结合哨兵1号、哨兵2号和SRTM数据构建多源特征集合，并基于随机森林的递归特征消除算法(RF＿RFE)进行特征优选，构建不同特征组合方案对比多源特征对湿地分类结果的影响，最后采用随机森林算法对研究区湿地进行分类提取。研究结果表明，设计样本采集方案可有效提高样本质量，多源特征集合能够提升湿地分类精度，特征优选能够减少特征冗余并提升分类精度。研究区分类总体精度为85.62%,Kappa系数为0.8333，其中湿地类别的精度最低为69.85%，最高为95.18%。 相似文献
7.
针对从中分辨率遥感影像中提取果园种植信息所面临的特征维数高、提取精度低等问题,采用随机森林算法进行特征优选,进而提取脐橙果园种植信息.以重庆奉节为研究区,首先选取春、秋、冬3季Landsat 8 OLI遥感数据生成光谱特征、植被指数特征和纹理特征,并结合地形因子构建初始特征集;再通过随机森林算法确定优选特征集;最后基于随机森林分类器,利用初始特征集和优选特征集进行脐橙果园的提取实验,并对各实验结果进行了精度评估和对比分析.研究结果表明,基于随机森林算法构建的优选特征集能在降低特征维度的情况下获得较高的精度,总体精度和Kappa系数分别为90.71％和0.89. 相似文献
8.
利用大丰市沿海滩涂湿地区域的高光谱影像和同时期的机载LIDAR数据,结合影像的光谱信息,采用随机森林算法(RF)对研究区进行湿地植被精细分类,并分析和评价分类模型参数设置对总体精度的影响,最后与SVM分类结果进行对比。结果表明:随机森林分类方法的总体精度为90.3%、卡帕(Kappa)系数为0.874;与传统的SVM分类方法相比,RF法均提高了4种湿地植被的生产者精度和使用者精度。通过分析RF分类模型参数设置对总体精度的影响,得出当生长树个数为30、生长树深度为30时,分类精度最高。 相似文献
9.
针对新疆南疆大规模枣园的检测识别,本文提出了一种基于泛化迁移深度学习的枣园目标检测识别方法。以GF-6卫星影像数据为基础制作了Jujube数据集,并将其泛化扩充增强;以Faster R-CNN体系为基础,利用多态协同模式实现数据集的有效关联和优化重构,进行检测识别模型的迁移深度学习以提高对目标对象检测识别的泛化能力。结果表明,模型算法的验证识别精确率、召回率和调和平均值分别达0.979、0.952和0.965,在应用测试中,3个指标平均值均大于0.929,优于传统检测方法,且本文模型方法总体分类精度为0.97,Kappa系数为0.93,均高于面向对象最邻近法,能够有效地满足研究区规模化枣园目标检测识别的精度和效率的要求,为精细化枣园田间管理提供基础依据。 相似文献
10.
基于时间序列统计特性的森林变化监测 总被引：1，自引：0，他引：1
森林动态变化分析对揭示生态系统环境变化及植被恢复和布局重建等具有重要意义,时间序列的遥感数据为森林监测提供了基础数据。本文根据森林植被的统计学特性,在暗目标法的基础上,利用归一化植被指数NDVI实现森林样本自动选择;并融合NDVI构建了新的综合森林特征指数(Integrated Forest Z-Score,IFZ);以时间序列的IFZ分析森林动态信息,实现森林变化动态监测。以三峡大坝及周边区域森林为研究区,利用2001年至2012年每年生长季节(5月—10月)的Landsat TM影像检验本文算法。基于2002年、2006年和2010年三期7月—9月的Quick Bird影像的精度分析结果发现:研究区森林变化检测的总体精度可达96.53%,Kappa系数为0.9512。在添加NDVI指数后构建的IFZ提高了总体监测精度。其中,毁林类别的检测精度提高显著,漏检率和误检率分别为2.74%和3.64%;干扰后重建的森林类别的检测精度有一定提高,其漏检率和误检率分别为10.79%和10.51%。研究结果表明,改进暗目标法能提高森林样本的选样效率,添加NDVI的IFZ能提高森林动态变化的识别度。此外,本算法不仅能定性识别森林变化,而且能定量提供森林干扰发生时间和干扰强度。 相似文献
11.
《地理信息系统科学与遥感》2013,50(4):392-405
Wildland fires are common in rangelands worldwide. The potential for high-severity fires to affect long-term changes in rangelands is considerable, and for this reason assessing fire severity shortly after the fire is critical. Such assessments are typically carried out following Burned Area Emergency Response teams or similar protocols. These data can then be used by land managers to plan remediation and future land uses. To complement these procedures and explore fire severity modeling of sagebrush steppe rangelands, we compared models developed using (1) post-fire imagery only with (2) differenced imagery (pre-fire minus post-fire imagery). All models were developed from Classification Tree Analysis (CTA) techniques using Satellite Pour l'Observation de la Terre 5 (SPOT 5) imagery and Shuttle Radar Topography Mission (SRTM) elevation data. The results indicate that both techniques produced similar fire severity models (model agreement = 98.5%) and that little improvement in overall accuracy was gained by using differenced imagery (0.5%). We suggest the use of CTA models developed using only the post-fire imagery. The analyses and techniques described in this paper provide land managers with tools to better justify their recommendations and decisions following wildland fires in sagebrush steppe ecosystems and provide remote sensing scientists with information that will potentially improve future modeling efforts. 相似文献
12.
Qingmin Meng Ross K. Meentemeyer 《International Journal of Applied Earth Observation and Geoinformation》2011
Most of fire severity studies use field measures of composite burn index (CBI) to represent forest fire severity and fit the relationships between CBI and Landsat imagery derived differenced normalized burn ratio (dNBR) to predict and map fire severity at unsampled locations. However, less attention has been paid on the multi-strata forest fire severity, which represents fire activities and ecological responses at different forest layers. In this study, using field measured fire severity across five forest strata of dominant tree, intermediate-sized tree, shrub, herb, substrate layers, and the aggregated measure of CBI as response variables, we fit statistical models with predictors of Landsat TM bands, Landsat derived NBR or dNBR, image differencing, and image ratioing data. We model multi-strata forest fire in the historical recorded largest wildfire in California, the Big Sur Basin Complex fire. We explore the potential contributions of the post-fire Landsat bands, image differencing, image ratioing to fire severity modeling and compare with the widely used NBR and dNBR. Models using combinations of post-fire Landsat bands perform much better than NBR, dNBR, image differencing, and image ratioing. We predict and map multi-strata forest fire severity across the whole Big Sur fire areas, and find that the overall measure CBI is not optimal to represent multi-strata forest fire severity. 相似文献
13.
J. Norton N. Glenn M. Germino K. Weber S. Seefeldt 《International Journal of Applied Earth Observation and Geoinformation》2009,11(5):360-367
Remote sensing indices of burn area and fire severity have been developed and tested for forest ecosystems, but not sparsely vegetated, desert shrub-steppe in which large wildfires are a common occurrence and a major issue for land management. We compared the performance of remote sensing indices for detecting burn area and fire severity with extensive ground-based cover assessments made before and after the prescribed burning of a 3 km2 shrub-steppe area. The remote sensing indices were based on either Landsat 7 ETM+ or SPOT 5 data, using either single or multiple dates of imagery. The indices delineating burned versus unburned areas had better overall, User, and Producer's accuracies than indices delineating levels of fire severity. The Soil Adjusted Vegetation Index (SAVI) calculated from SPOT had the greatest overall accuracy (100%) in delineating burned versus unburned areas. The relative differenced Normalized Burn Ratio (RdNBR) using Landsat ETM+ provided the highest accuracies (73% overall accuracy) for delineating fire severity. Though SPOT's spatial resolution likely conferred advantages for determining burn boundaries, the higher spectral resolution (particularly band 7, 2.21 μm) of Landsat ETM+ may be necessary for detecting differences in fire severity in sparsely vegetated shrub-steppe. 相似文献
14.
《地理信息系统科学与遥感》2013,50(2):276-299
After 110 years of sustained fire suppression, the 2000 Jasper fire consumed about 33,785 ha (83,500 acres), or 12% of the Black Hills National Forest. We mapped the severity of the Jasper fire using Landsat imagery and then investigated post-fire vegetation regeneration conditions using field data, Quickbird imagery, and regression modeling. We found that fire scar and severity could be delineated and mapped accurately based on remotely sensed and field-acquired data. Results also revealed that vegetative recovery relative to burn severity, topography, and soil factors could be modeled effectively using local geographically weighted regression (GWR). Further regeneration assessment revealed that severely or heavily burned areas were more rapidly re-vegetated with grasses, forbs, and woody shrubs in the short term. The field survey showed that prescribed burns retard crown fires and that after eight years ponderosa pine seedlings have not re-established. 相似文献
15.
《地理信息系统科学与遥感》2013,50(4):388-410
Soil erosion is a prominent cause of land degradation and desertification in Mediterranean countries. The detrimental effects of soil erosion are exemplified in climate (in particular climate change), topography, human activities, and natural disasters. Forest fires, which are an integral part of Mediterranean ecosystems, are responsible for the destruction of above-and below-ground vegetation that protects against soil erosion. Under this perspective, the estimation of potential soil erosion, especially after fire events, is critical for identifying watersheds that require management to prevent sediment loss, flooding, and increased ecosystem degradation. The objective of this study was to model the potential post-fire soil erosion risk following a large and intensive wildland fire, in order to prioritize protection and management actions at the watershed level in a Mediterranean landscape. Burn severity and preand post-fire land cover/uses were mapped using an ASTER image acquired two years before the fire, air photos acquired shortly after the fire, and a Landsat TM image acquired within one month after-fire. We estimated pre-and post-fire sediment loss using an integrated GIS-based approach, and additionally we analyzed landscape erosion patterns. The overall accuracy of the severity map reached 83%. Severe and heavy potential erosion classes covered approximately 90% of the total area following the fire, compared to 55% before. The fire had a profound effect on the spatial erosion pattern by altering the distribution of the potential erosion classes in 21 out of 24 watersheds, and seven watersheds were identified as being the most vulnerable to post-fire soil erosion. The spatial pattern of the erosion process is important because landscape cover heterogeneity induced especially by fire is a dominant factor controlling runoff generation and erosion rate, and should be considered in post-fire erosion risk assessment. 相似文献
16.
Forest fires are one of the most important causes of environmental alteration in Mediterranean countries. Discrimination of different degrees of burn severity is critical for improving management of fire-affected areas. This paper aims to evaluate the usefulness of land surface temperature (LST) as potential indicator of burn severity. We used a large convention-dominated wildfire, which occurred on 19–21 September, 2012 in Northwestern Spain. From this area, a 1-year series of six LST images were generated from Landsat 7 Enhanced Thematic Mapper (ETM+) data using a single channel algorithm. Further, the Composite Burn Index (CBI) was measured in 111 field plots to identify the burn severity level (low, moderate, and high). Evaluation of the potential relationship between post-fire LST and ground measured CBI was performed by both correlation analysis and regression models. Correlation coefficients were higher in the immediate post-fire LST images, but decreased during the fall of 2012 and increased again with a second maximum value in summer, 2013. A linear regression model between post-fire LST and CBI allowed us to represent spatially predicted CBI (R-squaredadj > 85%). After performing an analysis of variance (ANOVA) between post-fire LST and CBI, a Fisher's least significant difference test determined that two burn severity levels (low-moderate and high) could be statistically distinguished. The identification of such burn severity levels is sufficient and useful to forest managers. We conclude that summer post-fire LST from moderate resolution satellite data may be considered as a valuable indicator of burn severity for large fires in Mediterranean forest ecosytems. 相似文献
17.
Careful evaluation of forest regeneration and vegetation recovery after a fire event provides vital information useful in land management. The use of remotely sensed data is considered to be especially suitable for monitoring ecosystem dynamics after fire. The aim of this work was to map post-fire forest regeneration and vegetation recovery on the Mediterranean island of Thasos by using a combination of very high spatial (VHS) resolution (QuickBird) and hyperspectral (EO-1 Hyperion) imagery and by employing object-based image analysis. More specifically, the work focused on (1) the separation and mapping of three major post-fire classes (forest regeneration, other vegetation recovery, unburned vegetation) existing within the fire perimeter, and (2) the differentiation and mapping of the two main forest regeneration classes, namely, Pinus brutia regeneration, and Pinus nigra regeneration. The data used in this study consisted of satellite images and field observations of homogeneous regenerated and revegetated areas. The methodology followed two main steps: a three-level image segmentation, and, a classification of the segmented images. The process resulted in the separation of classes related to the aforementioned objectives. The overall accuracy assessment revealed very promising results (approximately 83.7% overall accuracy, with a Kappa Index of Agreement of 0.79). The achieved accuracy was 8% higher when compared to the results reported in a previous work in which only the EO-1 Hyperion image was employed in order to map the same classes. Some classification confusions involving the classes of P. brutia regeneration and P. nigra regeneration were observed. This could be attributed to the absence of large and dense homogeneous areas of regenerated pine trees in the study area. 相似文献
18.
S. Veraverbeke S. Lhermitte W.W. Verstraeten R. Goossens 《International Journal of Applied Earth Observation and Geoinformation》2011
Burn severity is an important parameter in post-fire management. It incorporates both the direct fire impact (vegetation depletion) and ecosystem responses (vegetation regeneration). From a remote sensing perspective, burn severity is traditionally estimated using Landsat's differenced normalized burn ratio (dNBR). In this case study of the large 2007 Peloponnese (Greece) wildfires, Landsat dNBR estimates correlated reasonably well with Geo composite burn index (GeoCBI) field data of severity (R2 = 0.56). The usage of Landsat imagery is, however, restricted by cloud cover and image-to-image normalization constraints. Therefore a multi-temporal burn severity approach based on coarse spatial, high temporal resolution moderate resolution imaging spectroradiometer (MODIS) imagery is presented in this study. The multi-temporal dNBR (dNBRMT) is defined as the 1-year integrated difference between burned pixels and their unique control pixels. These control pixels were selected based on time series similarity and spatial context and reflect how burned pixels would have behaved in the case no fire had occurred. Linear regression between downsampled Landsat dNBR and dNBRMT estimates resulted in a moderate-high coefficient of determination R2 = 0.54. dNBRMT estimates are indicative for the change in vegetation productivity due to the fire. This change is considerably higher for forests than for more sparsely vegetated areas like shrub lands. Although Landsat dNBR is superior for spatial detail, MODIS-derived dNBRMT estimates present a valuable alternative for burn severity mapping at continental to global scale without image availability constraints. This is beneficial to compare trends in burn severity across regions and time. Moreover, thanks to MODIS's repeated temporal sampling, the dNBRMT accounts for both first- and second-order fire effects. 相似文献
19.
S. Lhermitte J. Verbesselt W.W. VerstraetenS. Veraverbeke P. Coppin 《ISPRS Journal of Photogrammetry and Remote Sensing》2011,66(1):17-27
Several remote sensing studies have discussed the potential of satellite imagery as an alternative for extensive field sampling to quantify fire-vegetation impact over large areas. Most studies depend on Landsat image availability with infrequent image acquisition dates and consequently are limited for assessing intra-annual fire-vegetation dynamics or comparing different fire plots and dates. The control pixel based regeneration index (pRI) derived from SPOT-VEGETATION (VGT) normalized difference vegetation index (NDVI) is used in this study as an alternative to the traditional bi-temporal Landsat approach based on the normalized burn ratio (NBR). The major advantage of the pRI is the use of unburnt control plots which allow the expression of the intra-annual variation due to regeneration processes without external influences. In the comparison of Landsat and VGT data, (i) the inter-annual differences between the bi-temporal and control plot approach were contrasted and (ii) metrics of pRI were derived and compared with the inter-annual dynamics of both VGT and Landsat data. Results of these comparisons, demonstrate the overall similarity between NBR and NDVI data, stress the importance of the elimination of external influences (e.g., phenological variations), and emphasize the failure of including post-fire vegetation responses in bi-temporal Landsat assessments, especially in quickly recovering ecotypes with a strong annual phenological cycle such as savanna. This highlights the importance of using high frequency multi-temporal approaches to estimate fire-vegetation impact in temporally dynamic vegetation types. 相似文献
20.
Abstract Wildfire is a major disturbance agent in Mediterranean Type Ecosystems (MTEs). Providing reliable, quantitative information on the area of burns and the level of damage caused is therefore important both for guiding resource management and global change monitoring. Previous studies have successfully mapped burn severity using remote sensing, but reliable accuracy has yet to be gained using standard methods over different vegetation types. The objective of this research was to classify burn severity across several vegetation types using Landsat ETM imagery in two areas affected by wildfire in southern California in June 1999. Spectral mixture analysis (SMA) using four reference endmembers (vegetation, soil, shade, non‐photosynthetic vegetation) and a single (charcoal‐ash) image endmember were used to enhance imagery prior to burn severity classification using decision trees. SMA provided a robust technique for enhancing fire‐affected areas due to its ability to extract sub‐pixel information and minimize the effects of topography on single date satellite data. Overall kappa classification accuracy results were high (0.71 and 0.85, respectively) for the burned areas, using five canopy consumption classes. Individual severity class accuracies ranged from 0.5 to 0.94. 相似文献