共查询到20条相似文献，搜索用时 31 毫秒
1.
In this study, we tested whether the inclusion of the red-edge band as a covariate to vegetation indices improves the predictive accuracy in forest carbon estimation and mapping in savanna dry forests of Zimbabwe. Initially, we tested whether and to what extent vegetation indices (simple ratio SR, soil-adjusted vegetation index and normalized difference vegetation index) derived from high spatial resolution satellite imagery (WorldView-2) predict forest carbon stocks. Next, we tested whether inclusion of reflectance in the red-edge band as a covariate to vegetation indices improve the model's accuracy in forest carbon prediction. We used simple regression analysis to determine the nature and the strength of the relationship between forest carbon stocks and remotely sensed vegetation indices. We then used multiple regression analysis to determine whether integrating vegetation indices and reflection in the red-edge band improve forest carbon prediction. Next, we mapped the spatial variation in forest carbon stocks using the best regression model relating forest carbon stocks to remotely sensed vegetation indices and reflection in the red-edge band. Our results showed that vegetation indices alone as an explanatory variable significantly (p < 0.05) predicted forest carbon stocks with R2 ranging between 45 and 63% and RMSE ranging from 10.3 to 12.9%. However, when the reflectance in the red-edge band was included in the regression models the explained variance increased to between 68 and 70% with the RMSE ranging between 9.56 and 10.1%. A combination of SR and reflectance in the red edge produced the best predictor of forest carbon stocks. We concluded that integrating vegetation indices and reflectance in the red-edge band derived from high spatial resolution can be successfully used to estimate forest carbon in dry forests with minimal error. 相似文献
2.
Sentinel-2数据的冬小麦地上干生物量估算及评价 总被引：3，自引：0，他引：3
作物生物量快速精确的监测对于农业资源的合理利用与农田的精准管理具有重要意义。近年来,遥感技术因其独特的优势已被广泛用于作物生物量的估算中。本文主要针对不同宽波段植被指数在冬小麦生物量(文中的生物量均是指地上干生物量)估算方面的表现进行探索。首先利用欧洲空间局最新的Sentinel-2A卫星数据提取出17种常见的植被指数,之后分别构建其与相应时期内采集的冬小麦地上生物量间的最优估算模型,通过分析两者间的相关性与敏感性,获取适宜进行生物量估算的指数。最后,绘制了研究区的生物量空间分布图。结果表明,所选的植被指数均与生物量显著相关。其中,红边叶绿素指数(CI_(re))与生物量的估算精度最高(决定性系数R~2为0.83;均方根误差RMSE为180.29 g·m~(–2))。虽然相关性较高,但部分指数,如归一化差值植被指数(NDVI)等在生物量较高时会出现饱和现象,从而导致生物量的低估。而加入红边波段的指数不仅能够延缓指数的饱和趋势,而且能够提高反演精度。此外,通过敏感性分析发现,归一化差值指数和比值指数分别在作物生长的早期和中后期对生物量的变化保持较高的敏感性。由于红边比值指数(SR_(re))和MERIS叶绿素敏感指数(MTCI)在冬小麦全生长季内一直对生物量的变化保持高灵敏性,二者是生物量估算中最为稳定的指数。 相似文献
3.
The challenge of assessing and monitoring the influence of rangeland management practices on grassland productivity has been hampered in southern Africa, due to the lack of cheap earth observation facilities. This study, therefore, sought to evaluate the capability of the newly launched Sentinel 2 multispectral imager (MSI) data, in relation to Hyperspectral infrared imager (HyspIRI) data in estimating grass biomass subjected to different management practices, namely, burning, mowing and fertilizer application. Using sparse partial least squares regression (SPLSR), results showed that HyspIRI data exhibited slightly higher grass biomass estimation accuracies (RMSE = 6.65 g/m2, R2 = 0.69) than Sentinel 2 MSI (RMSE = 6.79 g/m2, R2 = 0.58) across all rangeland management practices. Student t-test results then showed that Sentinel 2 MSI exhibited a comparable performance to HyspIRI in estimating the biomass of grasslands under burning, mowing and fertilizer application. In comparing the RMSEs derived using wave bands and vegetation indices of HyspIRI and Sentinel, no statistically significant differences were exhibited (α = 0.05). Sentinel (Bands 5, 6 and 7) and HyspIRI (Bands 730 nm, 740 nm, 750 nm, 710 nm), as well as their derived vegetation indices, yielded the highest predictive accuracies. These findings illustrate that the accuracy of Sentinel 2 MSI data in estimating grass biomass is acceptable when compared with HyspIRI. The findings of this work provide an insight into the prospects of large-scale grass biomass modeling and prediction, using cheap and readily available multispectral data. 相似文献
4.
This study tested the use of machine learning techniques for the estimation of above-ground biomass (AGB) of Sonneratia caseolaris in a coastal area of Hai Phong city, Vietnam. We employed a GIS database and multi-layer perceptron neural networks (MLPNN) to build and verify an AGB model, drawing upon data from a survey of 1508 mangrove trees in 18 sampling plots and ALOS-2 PALSAR imagery. We assessed the model’s performance using root-mean-square error, mean absolute error, coefficient of determination (R2), and leave-one-out cross-validation. We also compared the model’s usability with four machine learning techniques: support vector regression, radial basis function neural networks, Gaussian process, and random forest. The MLPNN model performed well and outperformed the machine learning techniques. The MLPNN model-estimated AGB ranged between 2.78 and 298.95 Mg ha?1 (average = 55.8 Mg ha?1); below-ground biomass ranged between 4.06 and 436.47 Mg ha?1 (average = 81.47 Mg ha?1), and total carbon stock ranged between 3.22 and 345.65 Mg C ha?1 (average = 64.52 Mg C ha?1). We conclude that ALOS-2 PALSAR data can be accurately used with MLPNN models for estimating mangrove forest biomass in tropical areas. 相似文献
5.
Estimating tropical biomass is critical for establishment of conservation inventories and landscape monitoring. However, monitoring biomass in a complex and dynamic environment using traditional methods is challenging. Recently, biomass estimates based on remotely sensed data and ecological variables have shown great potential. The present study explored the utility of remotely sensed data and topo-edaphic factors to improve biomass estimation in the Eastern Arc Mountains of Tanzania. Twenty-nine vegetation indices were calculated from RapidEye data, while topo-edaphic factors were taken from field measurements. Results showed that using topo-edaphic variables or vegetation indices, biomass could be predicted with an R2 of 0.4. A combination of topo-edaphic variables and vegetation indices improved the prediction accuracy to an R2 of 0.6. Results further showed a decrease in biomass estimates from 1162 ton ha?1 in 1980 to 285.38 ton ha?1 in 2012. This study demonstrates the value of combining remotely sensed data with topo-edaphic variables in biomass estimation. 相似文献
6.
Mbulisi Sibanda Onisimo Mutanga Timothy Dube Thulile S Vundla Paramu L Mafongoya 《地理信息系统科学与遥感》2019,56(1):68-86
This study assessed the strength of Sentinel-2 multispectral instrument (MSI) derived Red Edge (RE) bands in estimating Leaf Area Index (LAI) and mapping canopy storage capacity (CSC) for hydrological applications in wattle infested ecosystems. To accomplish this objective, this study compared the estimation strength of models derived, using standard bands (all bands excluding the RE band) with those including RE bands, as well as different vegetation indices. Sparse Partial Least Squares (SPLSR) and Partial Least Squares Regression (PLSR) ensembles were used in this study. Results showed that the RE spectrum covered by the Sentinel-2 MSI satellite reduced the estimation error by a magnitude of 0.125 based on simple ratio (RE SR) vegetation indices from 0.157 m2· m?2 based on standard bands, and by 0.078 m2· m?2 based on red edge normalised difference vegetation (NDVI-RE). The optimal models for estimating LAI to map CSC were obtained based on the RE bands centered at 705 nm (Band 5), 740 nm (Band 6), 783 nm (Band 7) as well as 865 nm (Band 8a). A root mean square error of prediction (RMSEP) of 0.507 m2· m?2 a relative root mean square error of prediction (RRMSEP) of 11.3% and R2 of 0.91 for LAI and a RMSEP of 0.246 m2/m2 (RRMSEP = 7.9%) and R2 of 0.91 for CSC were obtained. Overall, the findings of this study underscore the relevance of the new copernicus satellite product in rapid monitoring of ecosystems that are invaded by alien invasive species. 相似文献
7.
Mariano Garcia Sassan Saatchi Antonio Ferraz Carlos Alberto Silva Susan Ustin Alexander Koltunov Heiko Balzter 《Carbon balance and management》2017,12(1):4
Background
Accurate estimation of aboveground forest biomass (AGB) and its dynamics is of paramount importance in understanding the role of forest in the carbon cycle and the effective implementation of climate change mitigation policies. LiDAR is currently the most accurate technology for AGB estimation. LiDAR metrics can be derived from the 3D point cloud (echo-based) or from the canopy height model (CHM). Different sensors and survey configurations can affect the metrics derived from the LiDAR data. We evaluate the ability of the metrics derived from the echo-based and CHM data models to estimate AGB in three different biomes, as well as the impact of point density on the metrics derived from them.Results
Our results show that differences among metrics derived at different point densities were significantly different from zero, with a larger impact on CHM-based than echo-based metrics, particularly when the point density was reduced to 1 point m?2. Both data models-echo-based and CHM-performed similarly well in estimating AGB at the three study sites. For the temperate forest in the Sierra Nevada Mountains, California, USA, R2 ranged from 0.79 to 0.8 and RMSE (relRMSE) from 69.69 (35.59%) to 70.71 (36.12%) Mg ha?1 for the echo-based model and from 0.76 to 0.78 and 73.84 (37.72%) to 128.20 (65.49%) Mg ha?1 for the CHM-based model. For the moist tropical forest on Barro Colorado Island, Panama, the models gave R2 ranging between 0.70 and 0.71 and RMSE between 30.08 (12.36%) and 30.32 (12.46) Mg ha?1 [between 0.69–0.70 and 30.42 (12.50%) and 61.30 (25.19%) Mg ha?1] for the echo-based [CHM-based] models. Finally, for the Atlantic forest in the Sierra do Mar, Brazil, R2 was between 0.58–0.69 and RMSE between 37.73 (8.67%) and 39.77 (9.14%) Mg ha?1 for the echo-based model, whereas for the CHM R2 was between 0.37–0.45 and RMSE between 45.43 (10.44%) and 67.23 (15.45%) Mg ha?1.Conclusions
Metrics derived from the CHM show a higher dependence on point density than metrics derived from the echo-based data model. Despite the median of the differences between metrics derived at different point densities differing significantly from zero, the mean change was close to zero and smaller than the standard deviation except for very low point densities (1 point m?2). The application of calibrated models to estimate AGB on metrics derived from thinned datasets resulted in less than 5% error when metrics were derived from the echo-based model. For CHM-based metrics, the same level of error was obtained for point densities higher than 5 points m?2. The fact that reducing point density does not introduce significant errors in AGB estimates is important for biomass monitoring and for an effective implementation of climate change mitigation policies such as REDD + due to its implications for the costs of data acquisition. Both data models showed similar capability to estimate AGB when point density was greater than or equal to 5 point m?2.8.
Jorge Andres Morandé Christine M. Stockert Garrett C. Liles John N. Williams David R. Smart Joshua H. Viers 《Carbon balance and management》2017,12(1):5
Background
Quantifying terrestrial carbon (C) stocks in vineyards represents an important opportunity for estimating C sequestration in perennial cropping systems. Considering 7.2 M ha are dedicated to winegrape production globally, the potential for annual C capture and storage in this crop is of interest to mitigate greenhouse gas emissions. In this study, we used destructive sampling to measure C stocks in the woody biomass of 15-year-old Cabernet Sauvignon vines from a vineyard in California’s northern San Joaquin Valley. We characterize C stocks in terms of allometric variation between biomass fractions of roots, aboveground wood, canes, leaves and fruits, and then test correlations between easy-to-measure variables such as trunk diameter, pruning weights and harvest weight to vine biomass fractions. Carbon stocks at the vineyard block scale were validated from biomass mounds generated during vineyard removal.Results
Total vine C was estimated at 12.3 Mg C ha?1, of which 8.9 Mg C ha?1 came from perennial vine biomass. Annual biomass was estimated at 1.7 Mg C ha?1 from leaves and canes and 1.7 Mg C ha?1 from fruit. Strong, positive correlations were found between the diameter of the trunk and overall woody C stocks (R2 = 0.85), pruning weights and leaf and fruit C stocks (R2 = 0.93), and between fruit weight and annual C stocks (R2 = 0.96).Conclusions
Vineyard C partitioning obtained in this study provides detailed C storage estimations in order to understand the spatial and temporal distribution of winegrape C. Allometric equations based on simple and practical biomass and biometric measurements could enable winegrape growers to more easily estimate existing and future C stocks by scaling up from berries and vines to vineyard blocks.9.
Leaf area index (LAI) and biomass are important indicators of crop development and the availability of this information during the growing season can support farmer decision making processes. This study demonstrates the applicability of RapidEye multi-spectral data for estimation of LAI and biomass of two crop types (corn and soybean) with different canopy structure, leaf structure and photosynthetic pathways. The advantages of Rapid Eye in terms of increased temporal resolution (∼daily), high spatial resolution (∼5 m) and enhanced spectral information (includes red-edge band) are explored as an individual sensor and as part of a multi-sensor constellation. Seven vegetation indices based on combinations of reflectance in green, red, red-edge and near infrared bands were derived from RapidEye imagery between 2011 and 2013. LAI and biomass data were collected during the same period for calibration and validation of the relationships between vegetation indices and LAI and dry above-ground biomass. Most indices showed sensitivity to LAI from emergence to 8 m2/m2. The normalized difference vegetation index (NDVI), the red-edge NDVI and the green NDVI were insensitive to crop type and had coefficients of variations (CV) ranging between 19 and 27%; and coefficients of determination ranging between 86 and 88%. The NDVI performed best for the estimation of dry leaf biomass (CV = 27% and r2 = 090) and was also insensitive to crop type. The red-edge indices did not show any significant improvement in LAI and biomass estimation over traditional multispectral indices. Cumulative vegetation indices showed strong performance for estimation of total dry above-ground biomass, especially for corn (CV ≤ 20%). This study demonstrated that continuous crop LAI monitoring over time and space at the field level can be achieved using a combination of RapidEye, Landsat and SPOT data and sensor-dependant best-fit functions. This approach eliminates/reduces the need for reflectance resampling, VIs inter-calibration and spatial resampling. 相似文献
10.
Land surface temperature (LST) is an important aspect in global to regional change studies, for control of climate change and balancing of high temperature. Urbanization is one of the influencing factors increasing land surface and atmospheric temperature, by the emission of greenhouse gases (e.g. CO2, NO and methane). In the present study, LST was derived from Landsat-8 of multitemporal data sets to analyse the spatial structure of the urban thermal environment in relation to the urban surface characteristics and land use–land cover (LULC). LST is influenced by the greenhouse gases i.e. CO2 plays an important role in increasing the earth’s surface temperature. In order to provide the evidence of influence of CO2 on LST, the relationship between LST, air temperature and CO2 was analysed. Landsat-8 satellite has two thermal bands, 10 and 11. These bands were used to accurately to calculate the temperature over the study area. Results showed that the strength of correlation between ground monitoring data and satellite data was high. Based on correlation values of each month April (R2 = 0.994), May (R2 = 0.297) and June (R2 = 0.934), observed results show that band 10 was significantly correlating with air temperature. Relationship between LST and CO2 levels were obtained from linear regression analysis. band 11 was correlating significantly with CO2 values in each of the months April (R2 = 0.217), May (R2 = 0.914) and June, (R2 = 0.934), because band 11 is closer to the 15-micron band of CO2. From the results, it was observed that band 10 can be used for calculating air temperature and band 11 can be used for estimation of greenhouse gases. 相似文献
11.
氮素是植被整个生命周期的必要元素,红树林冠层氮素含量（CNC）遥感估算对红树林健康监测具有重要意义。以广东湛江高桥红树林保护区为研究区,本文旨在基于Sentinel-2影像超分辨率重建技术进行红树林CNC估算和空间制图。研究首先基于三次卷积重采样、Sen2Res和SupReMe算法实现Sentinel-2影像从20 m分辨率到10 m的重建;然后以重建后的影像和原始20 m影像为数据源构建40个相关植被指数,采用递归特征消除法（SVM-RFE）确定CNC估算的最优变量组合,进而构建CNC反演的核岭回归（KRR）模型;最后选取最优模型实现CNC制图。研究结果表明：基于Sen2Res和SupReMe超分辨率算法的重建影像不仅与原始影像具有很高的光谱一致性,且明显提高了影像的清晰度和空间细节。红树林CNC反演波段主要集中在红（B4）、红边（B5）、近红外波段（B8a）以及短波红外波段（B11和B12）,与“红边波段”相关的植被指数（RSSI和TCARIre1/OSAVI）也是红树林CNC反演的有效变量。基于3种方法重建后10 m的影像构建的模型反演精度（R2val>0.579）均优于原始20 m的影像（R2val=0.504）;基于Sen2Res算法重建影像构建的反演模型拟合精度（R2val=0.630,RMSE_val=5.133,RE_val=0.179）与基于三次卷积重采样重建影像的模型拟合精度（R2val=0.640,RMSE_val=5.064,RE_val=0.179）基本相当,前者模型验证精度（R2cv=0.497,RMSE_cv=5.985,RE_cv=0.214）较高且模型变量选择数量最为合理。综合重建影像光谱细节及模型精度,基于Sen2Res算法重建的Sentinel-2影像在红树林CNC估算中具有良好的应用潜力,能为区域尺度红树林冠层健康状况的精细监测提供有效的方法借鉴和数据支撑。 相似文献
12.
The influence of morphophysiological variation at different growth stages on the performance of vegetation indices for estimating plant N status has been confirmed. However, the underlying mechanisms explaining how this variation impacts hyperspectral measures and canopy N status are poorly understood. In this study, four field experiments involving different N rates were conducted to optimize the selection of sensitive bands and evaluate their performance for modeling canopy N status of rice at various growth stages in 2007 and 2008. The results indicate that growth stages negatively affect hyperspectral indices in different ways in modeling leaf N concentration (LNC), plant N concentration (PNC) and plant N uptake (PNU). Published hyperspectral indices showed serious limitations in estimating LNC, PNC and PNU. The newly proposed best 2-band indices significantly improved the accuracy for modeling PNU (R2 = 0.75–0.85) by using the lambda by lambda band-optimized algorithm. However, the newly proposed 2-band indices still have limitations in modeling LNC and PNC because the use of only 2-band indices is not fully adequate to provide the maximum N-related information. The optimum multiple narrow band reflectance (OMNBR) models significantly increase the accuracy for estimating the LNC (R2 = 0.67–0.71) and PNC (R2 = 0.57–0.78) with six bands. Results suggest the combinations of center of red-edge (735 nm) with longer red-edge bands (730–760 nm) are very efficient for estimating PNC after heading, whereas the combinations of blue with green bands are more efficient for modeling PNC across all stages. The center of red-edge (730–735 nm) paired with early NIR bands (775–808 nm) are predominant in estimating PNU before heading, whereas the longer red-edge (750 nm) paired with the center of “NIR shoulder” (840–850 nm) are dominant in estimating PNU after heading and across all stages. The OMNBR models have the advantage of modeling canopy N status for the entire growth period. However, the best 2-band indices are much easier to use. Alternatively, it is also possible to use the best 2-band indices to monitor PNU before heading and PNC after heading. This study systematically explains the influences of N dilution effect on hyperspectral band combinations in relating to the different N variables and further recommends the best band combinations which may provide an insight for developing new hyperspectral vegetation indices. 相似文献
13.
One of the challenges in fighting plant invasions is the inefficiency of identifying their distribution using field inventory techniques. Remote sensing has the potential to alleviate this problem effectively using spectral profiling for species discrimination. However, little is known about the capability of remote sensing in discriminating between shrubby invasive plants with narrow leaf structures and other cohabitants with similar ecological niche. The aims of this study were therefore to (1) assess the classification performance of field spectroradiometer data among three bushy and shruby plants (Artemesia afra, Asparagus laricinus, and Seriphium plumosum) from the coexistent plant species largely dominated by acacia and grass species, and (2) explore the performance of simulated spectral bands of five space-borne images (Landsat 8, Sentinel 2A, SPOT 6, Pleiades 1B, and WorldView-3). Two machine-learning classifiers (boosted trees classification and support vector machines) were used to classify raw hyperspectral (n = 688) and simulated multispectral wavelengths. Relatively high classification accuracies were obtained for the invasive species using the original hyperspectral bands for both classifiers (overall accuracy, OA = 83–97%). The simulated data resulted in higher accuracies for Landsat 8, Sentinel 2A, and WorldView-3 compared to those computed for bands simulated to SPOT 6 and Pleiades 1B data. These findings suggest the potential of remote-sensing techniques in the discrimination of different plant species with similar morphological characteristics occupying the same niche. 相似文献
14.
Background
Forest landscape restoration (FLR) has been adopted by governments and practitioners across the globe to mitigate and adapt to climate change and restore ecological functions across degraded landscapes. However, the extent to which these activities capture CO2 with associated climate mitigation impacts are poorly known, especially in geographies where data on biomass growth of restored forests are limited or do not exist. To fill this gap, we developed biomass accumulation rates for a set of FLR activities (natural regeneration, planted forests and woodlots, agroforestry, and mangrove restoration) across the globe and global CO2 removal rates with corresponding confidence intervals, grouped by FLR activity and region/climate.Results
Planted forests and woodlots were found to have the highest CO2 removal rates, ranging from 4.5 to 40.7 t CO2 ha?1 year?1 during the first 20 years of growth. Mangrove tree restoration was the second most efficient FLR at removing CO2, with growth rates up to 23.1 t CO2 ha?1 year?1 the first 20 years post restoration. Natural regeneration removal rates were 9.1–18.8 t CO2 ha?1 year?1 during the first 20 years of forest regeneration, followed by agroforestry, the FLR category with the lowest and regionally broad removal rates (10.8–15.6 t CO2 ha?1 year?1). Biomass growth data was most abundant and widely distributed across the world for planted forests and natural regeneration, representing 45% and 32% of all the data points assessed, respectively. Agroforestry studies, were only found in Africa, Asia, and the Latin America and Caribbean regions.Conclusion
This study represents the most comprehensive review of published literature on tree growth and CO2 removals to date, which we operationalized by constructing removal rates for specific FLR activities across the globe. These rates can easily be applied by practitioners and decision-makers seeking to better understand the positive climate mitigation impacts of existing or planned FLR actions, or by countries making restoration pledges under the Bonn Challenge Commitments or fulfilling Nationally Determined Contributions to the UNFCCC, thereby helping boost FLR efforts world-wide.15.
《地理信息系统科学与遥感》2013,50(4):463-491
This research examined the utility of Airborne Imaging Spectrometer for Applications (AISA) hyperspectral imagery for estimating the biomass of three forest crops—sycamore, sweetgum, and loblolly pine—planted in experimental plots with a range of fertilization and irrigation treatments on the Savannah River Site near Aiken, South Carolina. Both vegetation index (VI) and red-edge positioning (REP) approaches were investigated to estimate the biomass associated with 12 treatment conditions. The optimum band pairs using the VI approach for biomass estimation were located mainly in the visible, NIR, and/or water absorption region around 970 nm, depending on the treatment conditions. Both the selected hyperspectral variables (i.e., VI and REP) resulted in good performance for biomass estimation for a range of treatment conditions except for those associated with loblolly pine. The hyperspectral variables were also examined to determine if they were able to identify the optimum fertilization treatment level. For the fertilization treatment conditions with good biomass estimation (R 20.9), their optimum treatment levels were successfully identified. 相似文献
16.
Leaf and canopy nitrogen (N) status relates strongly to leaf and canopy chlorophyll (Chl) content. Remote sensing is a tool that has the potential to assess N content at leaf, plant, field, regional and global scales. In this study, remote sensing techniques were applied to estimate N and Chl contents of irrigated maize (Zea mays L.) fertilized at five N rates. Leaf N and Chl contents were determined using the red-edge chlorophyll index with R2 of 0.74 and 0.94, respectively. Results showed that at the canopy level, Chl and N contents can be accurately retrieved using green and red-edge Chl indices using near infrared (780–800 nm) and either green (540–560 nm) or red-edge (730–750 nm) spectral bands. Spectral bands that were found optimal for Chl and N estimations coincide well with the red-edge band of the MSI sensor onboard the near future Sentinel-2 satellite. The coefficient of determination for the relationships between the red-edge chlorophyll index, simulated in Sentinel-2 bands, and Chl and N content was 0.90 and 0.87, respectively. 相似文献
17.
High-data dimensionality is a common problem in hyperspectral data processing. Consequently, remote sensing techniques that reduce the number of bands are considered essential tools for most hyperspectral applications. The aim of this study was to examine the utility of the random forest ensemble to select the optimal subset of hyperspectral bands to predict the age of Pinus patula stands. Airborne AISA Eagle hyperspectral image data were collected over the study area. The random forest ensemble was used to test whether the forward or backward variable selection methods could identify the optimal subset of bands. Results indicate that both the selection methods produced high-predictive accuracies (root mean square error = 3.097 years). However, the backward variable selection method utilized 206 bands for the final model, while the forward variable selection utilized only a small subset of non-redundant bands (n = 9) while preserving the highest model accuracy (R 2 = 0.6). 相似文献
18.
WorldView-2纹理的森林地上生物量反演 总被引：1，自引：0，他引：1
使用高空间分辨率卫星WorldView-2的多光谱遥感影像,构建植被指数和纹理因子等遥感因子与森林地上生物量的关系方程,并计算模型估测精度和均方根误差,探索高分辨率数据的光谱与纹理信息在温带森林地上生物量估测应用中的潜力。以黑龙江省凉水自然保护区温带天然林及天然次生林为研究对象,通过灰度共生矩阵(GLCM)、灰度差分向量(GLDV)及和差直方图(SADH)对高分辨率遥感影像进行纹理信息提取,并利用外业调查的74个样地地上生物量与遥感因子建立参数估计模型。提取的遥感因子包括6种植被指数(比值植被指数RVI、差值植被指数DVI、规一化植被指数NDVI、增强植被指数EVI、土壤调节植被指数SAVI和修正的土壤调节植被指数MSAVI)以及3类纹理因子(GLCM、GLDV和SADH)。为避免特征变量个数较多对估测模型造成过拟合,利用随机森林算法对提取的遥感因子进行特征选择,将最优的特征变量输入模型参与建模估测。采用支持向量回归(SVR)进行生物量建模及验证,结果显示选入模型的和差直方图均值(sadh_mean)、灰度共生矩阵方差(glcm_var)和差值植被指数(DVI)等遥感因子对森林地上生物量有较好的解释效果;植被指数+纹理因子组合的模型获得较精确的AGB估算结果(R2=0.85,RMSE=42.30 t/ha),单独使用植被指数的模型精度则较低(R~2=0.69,RMSE=61.13 t/ha)。 相似文献
19.
Estimation of green grass/herb biomass from airborne hyperspectral imagery using spectral indices and partial least squares regression 总被引：7，自引：0，他引：7
Moses Azong Cho Andrew Skidmore Fabio Corsi Sipke E. van Wieren Istiak Sobhan 《International Journal of Applied Earth Observation and Geoinformation》2007,9(4):414-424
The main objective was to determine whether partial least squares (PLS) regression improves grass/herb biomass estimation when compared with hyperspectral indices, that is normalised difference vegetation index (NDVI) and red-edge position (REP). To achieve this objective, fresh green grass/herb biomass and airborne images (HyMap) were collected in the Majella National Park, Italy in the summer of 2005. The predictive performances of hyperspectral indices and PLS regression models were then determined and compared using calibration (n = 30) and test (n = 12) data sets. The regression model derived from NDVI computed from bands at 740 and 771 nm produced a lower standard error of prediction (SEP = 264 g m−2) on the test data compared with the standard NDVI involving bands at 665 and 801 nm (SEP = 331 g m−2), but comparable results with REPs determined by various methods (SEP = 261 to 295 g m−2). PLS regression models based on original, derivative and continuum-removed spectra produced lower prediction errors (SEP = 149 to 256 g m−2) compared with NDVI and REP models. The lowest prediction error (SEP = 149 g m−2, 19% of mean) was obtained with PLS regression involving continuum-removed bands. In conclusion, PLS regression based on airborne hyperspectral imagery provides a better alternative to univariate regression involving hyperspectral indices for grass/herb biomass estimation in the Majella National Park. 相似文献
20.
Coastal wetlands are among the most productive ecosystems globally but have experienced dramatic degradation and loss within the past several decades. Vegetation biomass of coastal wetlands is not only the key component of blue carbon storage but also plays an important role in vertical accretion, important for maintaining these habitats under relative sea-level rise. Remote sensing offers a cost-effective approach to study vegetation biomass at a broad spatial scale. We developed statistical models to predict peak aboveground green biomass of Spartina alterniflora and Juncus roemerianus, two dominant species of salt marshes using WorldView-2 satellite imagery at the Grand Bay National Estuarine Research Reserve (NERR) on the Mississippi coast in the northern Gulf of Mexico. The model accounted for nested data structures in the sampled biomass, assimilated uncertainties from data, parameters and model structures, and helped determine the best vegetation index among a variety of commonly-used indices to predict aboveground green biomass. We developed a series of mixed-effects models, which included different combinations of fixed effect(s), random intercept, and random slope(s). The fixed effects were species and one of the 60 vegetation indices derived from a WorldView-2 image obtained on 6 October 2012. The random effect used was site. We implemented the models in a Bayesian framework and selected the best model structure and vegetation index based on minimum posterior predictive loss and deviance information criterion. The results showed that the best vegetation index to predict peak green biomass was the green chlorophyll index derived from the reflectance values of band 8 (near-infrared) and band 3 (green), and its effect on biomass prediction varied among sites. The inclusion of species as a fixed effect improved the model prediction. The study demonstrated the need to account for spatial dependence of data in developing a robust model, and the importance of the second WorldView-2 near-infrared band (860–1040 nm) in predicting aboveground green biomass for the Grand Bay NERR. The analysis using mixed-effects modeling in Bayesian inference which coherently combined field and WorldView-2 data with uncertainties accounted for provides a robust and nondestructive tool for resource managers to monitor the status of coastal wetlands at a high spatial resolution in a timely manner. Through this study, we hope to emphasize the importance of appropriately accounting for nested data structures using mixed-effects models and promote wider application of Bayesian inference to facilitate assimilation of uncertainties in remote sensing applications. 相似文献