Assessing the attributes of high-density Eucalyptus globulus stands using airborne laser scanner data

This article presents an airborne Light Detection and Ranging (LiDAR)-based method to extract interesting stand attributes for forest management in high-density Eucalyptus globulus Labill. plantations. An adaptive morphological filter (AMF) for classifying terrain LiDAR points in forested areas is used to classify LiDAR points; canopy cover (CC), number of LiDAR-detected trees per hectare (N _LD) and individual tree height (h _tree) were calculated using the canopy height model (CHM); and several statistics and metrics extracted from the CHM and the normalized height of the LiDAR data cloud (NHD) were incorporated into the linear and multiplicative models for estimating mean height (H _m), dominant height (H _d), mean diameter (d _m), quadratic mean diameter (d _g), number of stems per hectare (N), basal area (G) and volume (V). The height accuracy results of the LiDAR-derived digital terrain model (DTM), root mean square error (RMSE)?=?0.303 m, revealed that the developed filter behaved well. The values of the RMSE for CC, N _LD and h _tree were 13.2%, 733.3 stems ha^–1 and 1.91 m, respectively. The regressions explained 78% of the variance in ground-truth values for H _m (RMSE?=?1.33 m); 92% for H _d (RMSE?=?1.18 m); 71% for d _m (RMSE?=?1.68 cm); 73% for d _g (RMSE?=?1.66 cm); 49% for N (RMSE?=?667 stems ha^–1); 78% for G (RMSE?=?5.30 m² ha^–1); and 81% for V (RMSE?=?53.6 m³ ha^–1).     

Modelling forest stand volume and tree density using Landsat ETM+ data

Estimation of stand volume and tree density in a large area using remotely sensed data has considerable significance for sustainable management of natural resources. In this paper, we explore likely relationships between forest stand characteristics and Landsat Enhanced Thematic Mapper Plus (ETM+) reflectance values. We used multivariate regression technique to predict stand volume and tree density. The result showed that a linear combination of greenness and difference vegetation index (DVI) were better predictors of stand volume (adjusted R² = 43%; root mean square error (RMSE) = 97.4 m³ ha^?1) than other ETM+ bands and vegetation indices. In addition, the regression model with ETM4 (near infrared band) and ETM5 (first shortwave band) as independent variables was a better predictor of tree density (adjusted R² = 73.4%; RMSE = 170.13 ha^?1) than other combinations of ETM⁺ bands and vegetation indices. Results obtained from this study demonstrate the significant relationship between forest stand characteristics and ETM+ reflectance values and the utility of transformed bands in modelling stand volume and tree density. Based on the results of this study, we conclude that ETM+ data are useful to estimate forest volume and density and to gain insights into its structural characteristics in our study area. Forest managers could use ETM+ data for gaining insights into stand characteristics and generating maps required for developing forest management plans and identifying locations within stands that require treatments and other interventions.     

Model-assisted regional forest biomass estimation using LiDAR and InSAR as auxiliary data: A case study from a boreal forest area

There is a need for accurate inventory methods that produce relevant and timely information on the forest resources and carbon stocks for forest management planning and for implementation of national strategies under the United Nations Collaborative Program on Reduced Emissions from Deforestation and Forest Degradation in Developing Countries (REDD). Such methods should produce information that is consistent across various geographical scales. Airborne scanning Light Detection and Ranging (LiDAR) is among the most promising remote sensing technologies for estimation of forest resource information such as timber volume and biomass, while acquisition of three dimensional data with Interferometric Synthetic Aperture Radar (InSAR) from space is seen as a relevant option for inventory in the tropics because of its ability to “see through the clouds” and its potential for frequent updates at low costs. Based on a stratified probability sample of 201 field survey plots collected in a 960 km² boreal forest area in Norway, we demonstrate how total above-ground biomass (AGB) can be estimated at three distinct geographical levels in such a way that the estimates at a smaller level always sum up to the estimate at a larger level. The three levels are (1) a district (the entire study area), (2) a village, local community or estate level, and (3) a stand or patch level. The LiDAR and InSAR data were treated as auxiliary information in the estimation. At the two largest geographical levels model-assisted estimators were employed. A model-based estimation was conducted at the smallest level. Estimates of AGB and corresponding error estimates based on (1) the field sample survey were compared with estimates obtained by using (2) LiDAR and (3) InSAR data as auxiliary information. For the entire study area, the estimates of AGB were 116.0, 101.2, and 111.3 Mg ha⁻¹, respectively. Corresponding standard error estimates were 3.7, 1.6, and 3.2 Mg ha⁻¹. At the smallest geographical level (stand) an independent validation on 35 large field plots was carried out. RMSE values of 17.1-17.3 Mg ha⁻¹ and 42.6-53.2 Mg ha⁻¹ were found for LiDAR and InSAR, respectively. A time lag of six years between acquisition of InSAR data and field inventory has introduced some errors. Significant differences between estimates and reference values were found, illustrating the risk of using pure model-based methods in the estimation when there is a lack of fit in the models. We conclude that the examined remote sensing techniques can provide biomass estimates with smaller estimated errors than a field-based sample survey. The improvement can be highly significant, especially for LiDAR.     

Aboveground forest biomass derived using multiple dates of WorldView-2 stereo-imagery: quantifying the improvement in estimation accuracy

ABSTRACT

The aim of this study was to investigate the capabilities of two date satellite-derived image-based point clouds (IPCs) to estimate forest aboveground biomass (AGB). The data sets used include panchromatic WorldView-2 stereo-imagery with 0.46 m spatial resolution representing 2014 and 2016 and a detailed digital elevation model derived from airborne laser scanning data. Altogether, 332 field sample plots with an area of 256 m² were used for model development and validation. Predictors describing forest height, density, and variation in height were extracted from the IPC 2014 and 2016 and used in k-nearest neighbour imputation models developed with sample plot data for predicting AGB. AGB predictions for 2014 (AGB₂₀₁₄) were projected to 2016 using growth models (AGB_{Projected_2016}) and combined with the AGB estimates derived from the 2016 data (AGB₂₀₁₆). AGB prediction model developed with 2014 data was also applied to 2016 data (AGB_{2016_pred2014}). Based on our results, the change in the 90^th percentile of height derived from the WorldView-2 IPC was able to characterize forest height growth between 2014 and 2016 with an average growth of 0.9 m. Features describing canopy cover and variation in height derived from the IPC were not as consistent. The AGB₂₀₁₆ had a bias of ?7.5% (?10.6 Mg ha^?1) and root mean square error (RMSE) of 26.0% (36.7 Mg ha^?1) as the respective values for AGB_{Projected_2016} were 7.0% (9.9 Mg ha^?1) and 21.5% (30.8 Mg ha^?1). AGB_{2016_pred2014} had a bias of ?19.6% (?27.7 Mg ha^?1) and RMSE of 33.2% (46.9 Mg ha^?1). By combining predictions of AGB₂₀₁₆ and AGB_{Projected_2016} at sample plot level as a weighted average, we were able to decrease the bias notably compared to estimates made on any single date. The lowest bias of ?0.25% (?0.4 Mg ha^?1) was obtained when equal weights of 0.5 were given to the AGB_{Projected_2016} and AGB₂₀₁₆ estimates. Respectively, RMSE of 20.9% (29.5 Mg ha^?1) was obtained using equal weights. Thus, we conclude that combination of two date WorldView-2 stereo-imagery improved the reliability of AGB estimates on sample plots where forest growth was the only change between the two dates.     

Estimating biomass carbon stocks for a Mediterranean forest in central Spain using LiDAR height and intensity data

Biomass fractions (total aboveground, branches and foliage) were estimated from a small footprint discrete-return LiDAR system in an unmanaged Mediterranean forest in central Spain. Several biomass estimation models based on LiDAR height, intensity or height combined with intensity data were explored. Raw intensity data were normalized to a standard range in order to remove the range dependence of the intensity signal. In general terms, intensity-based models provided more accurate predictions of the biomass fractions. Height models selected were mainly based on a percentile of the height distribution. Intensity models selected included variables that consider the percentage of the intensity accumulated at different height percentiles, which implicitly take into account the height distribution. The general models derived considering all species together were based on height combined with intensity data. These models yielded R² values greater than 0.58 for the different biomass fractions considered and RMSE values of 28.89, 18.28 and 1.51 Mg ha⁻¹ for aboveground, branch and foliage biomass, respectively. Results greatly improved for species-specific models using the main species present in each plot, with R² values greater than 0.85, 0.70 and 0.90 for black pine, Spanish juniper and Holm oak, respectively, and with lower RMSE for the biomass fractions. Reductions in LiDAR point density had only a small effect on the results obtained, except for those models based on a variation of the Canopy Reflection Sum, which was weighted by the mean point density. Based on the species-specific equations derived, Holm oak dominated plots showed the highest average carbon contained by aboveground biomass and branch biomass 44.66 and 31.42 Mg ha^− 1 respectively, while for foliage biomass carbon, Spanish juniper showed the highest average value (3.04 Mg ha^− 1).     

Volume estimation in a Eucalyptus plantation using multi-source remote sensing and digital terrain data: a case study in Minas Gerais State,Brazil

In this study, we tested the effectiveness of stand age, multispectral optical imagery obtained from the Landsat 8 Operational Land Imager (OLI), synthetic aperture radar (SAR) data acquired by the Sentinel-1B satellite, and digital terrain attributes extracted from a digital elevation model (DEM), in estimating forest volume in 351 plots in a 1,498 ha Eucalyptus plantation in northern Minas Gerais state, Brazil. A Random Forest (RF) machine learning algorithm was used following the Principal Component Analysis (PCA) of various data combinations, including multispectr al and SAR texture variables and DEM-based geomorphometric derivatives. Using multispectral, SAR or DEM variables alone (i.e. Experiments (ii)–(iv)) did not provide accurate estimates of volume (RMSE (Root Mean Square Error) > 32.00 m³ ha^?1) compared to predictions based on age since planting of Eucalyptus stands (Experiment (i)). However, when these datasets were individually combined with stand age (i.e. Experiments (v)–(vii)), the RF models resulted in better volume estimates than those obtained when using the individual multispectral, SAR and DEM datasets (RMSE < 28.00 m³ ha^?1). Furthermore, a model that integrated the selected variables of these data with stand age (Experiment (viii)) improved volume estimation significantly (RMSE = 22.33 m³ ha^?1). The large and increasing area of Eucalyptus forest plantations in Brazil and elsewhere suggests that this new approach to volume estimation has the potential to support Eucalyptus plantation monitoring and forest management practices.     

TanDEM-X elevation model data for canopy height and aboveground biomass retrieval in a tropical peat swamp forest

It was demonstrated in the past that radar data is useful to estimate aboveground biomass due to their interferometric capability. Therefore, the potential of a globally available TanDEM-X digital elevation model (DEM) was investigated for aboveground biomass estimation via canopy height models (CHMs) in a tropical peat swamp forest. However, CHMs based on X-band interferometers usually require external terrain models. High accurate terrain models are not available on global scale. Therefore, an approach exclusively based on TanDEM-X and the decrease of accuracy compared to an approach utilizing a high accurate terrain model is assessed. In addition, the potential of X-band interferometric heights in tropical forests needs to be evaluated. Therefore, two CHMs were derived from an intermediate TanDEM-X DEM (iDEM; as a precursor for WorldDEM^TM) alone and in combination with lidar measurements used as terrain model. The analysis showed high accuracies (root mean square error [RMSE] = 5 m) for CHMs based on iDEM and reliable estimation of aboveground biomass. The iDEM CHM, exclusively based on TanDEM-X, achieved a poor R² of 0.2, nonetheless resulted in a cross-validated RMSE of 54 t ha^?1 (16%). The low R² suggested that the X-band height alone was not sufficient to estimate an accurate CHM, and thus the need for external terrain models was confirmed. A CHM retrieved from the difference of iDEM and an accurate lidar terrain model achieved a considerably higher correlation with aboveground biomass (R² = 0.68) and low cross-validated RMSE of 24.5 t ha^?1 (7.5%). This was higher or comparable to other aboveground biomass estimations in tropical peat swamp forests. The potential of X-band interferometric heights for CHM and biomass estimation was thus confirmed in tropical forest in addition to existing knowledge in boreal forests.     

联合无人机激光雷达和高光谱数据反演玉米叶面积指数

叶面积指数（Leaf Area Index, LAI）是作物长势监测及产量估算的重要指标,准确高效的LAI反演对农田经济的宏观管理具有重要作用。研究探索了联合无人机激光雷达（Light Detec-tion and Ranging, LiDAR） 和高光谱数据反演玉米叶面积指数的潜力,并分析了LiDAR数据不同采样尺寸、高度阈值、点密度对LAI反演精度的影响同时确定三者的最优值。该研究分别从重采样的LiDAR数据和高光谱影像中提取了LiDAR变量和植被指数,然后基于偏最小二乘回归（Partial Least Square Regression,PLSR）和随机森林（Random Forest, RF） 回归两种算法分别利用LiDAR变量、植被指数、联合LiDAR变量和植被指数构建预测模型,并确定反演玉米LAI的最优预测模型。结果表明：反演玉米LAI的最优采样尺寸、高度阈值、点密度分别为5.5 m、0.55 m、18 points/m²,研究发现最高的点密度（420 points/m²）并没有产生最优的玉米LAI反演精度,因此单独依靠增加点密度的方法提高LAI的反演精度并不可靠。基于LiDAR变量获得的LAI反演精度（PLSR：R²=0.874,RMSE=0.317;RF：R²=0.942,RMSE=0.222）高于基于植被指数获得的LAI反演精度（PLSR： R²=0.741,RMSE=0.454;RF：R²=0.861,RMSE=0.338）,而使用组合变量构建预测模型的反演精度（PLSR：R²=0.885, RMSE=0.304;RF：R²=0.950,RMSE=0.203）优于使用单一变量建立的LAI预测模型,其中利用联合LiDAR变量和植被指数建立的随机森林回归模型为最优预测模型。因此,将两种数据源融合在提高植被LAI反演精度方面具有一定的潜力。     

Estimating the leaf area index in Indian tropical forests using Landsat-8 OLI data

Leaf area index (LAI) is a key vegetation biophysical parameter and is extensively used in modelling of phenology, primary production, light interception, evapotranspiration, carbon, and nitrogen dynamics. In the present study, we attempt to spatially characterize LAI for natural forests of Western Ghats India, using ground based and Landsat-8 Operational Land Imager (OLI) sensor satellite data. For this, 41 ground-based LAI measurements were carried out across a gradient of tropical forest types, viz. dry, moist, and evergreen forests using LAI-2200 plant canopy analyser, during the month of March 2015. Initially, measured LAI values were regressed with 15 spectral variables, including nine spectral vegetation indices (SVIs) and six Landsat-8 surface reflectance (ρ) variables using univariate correlation analysis. Results showed that the red (ρ_red), near-infrared (ρ_NIR), shortwave infrared (ρ_SWIR1, ρ_SWIR2) reflectance bands (R² > 0.6), and all SVIs (R² > 0.7) except simple ratio (SR) have the highest and second highest coefficient of determination with ground-measured LAI. In the second step, to select significant (high R², low root mean square error (RMSE), and p-level < 0.05) SVIs to determine the best representative model, stepwise multiple linear regression (SMLR) was implemented. The results indicate that the SMLR model predicted LAI with better coefficient of determination (R² = 0.83, RMSE = 0.78) using normalized difference vegetation index, enhanced vegetation index, and soil-adjusted vegetation index variables compared to the univariate approach. The predicted SMLR model was used to estimate a spatial map of LAI. It is desirable to evaluate the stability and potentiality of regional LAI models in natural forest ecosystems against the operationally accepted Moderate Resolution Imaging Spectroradiometer (MODIS) global LAI product. To do this, the Landsat-8 pixel-based LAI map was resampled to 1 km resolution and compared with the MODIS derived LAI map. Results suggested that Landsat-8 OLI-based VIs provide significant LAI maps at moderate resolution (30 m) as well as coarse resolution (1 km) for regional climate models.     

Testing LiDAR models of fractional cover across multiple forest ecozones

Four LiDAR-based models of canopy fractional cover (FC_LiDAR) have been tested against hemispherical photography fractional cover measurements (FC_HP) and compared across five ecozones, eight forest species and multiple LiDAR survey configurations. The four models compared are based on: i) a canopy-to-total first returns ratio (FC_LiDAR(FR)) method; ii) a canopy-to-total returns ratio (FC_LiDAR(RR)); iii) an intensity return ratio (FC_LiDAR(IR)); and iv) a Beer's Law modified (two-way transmission loss) intensity return ratio (FC_LiDAR(BL)). It is found that for the entire dataset, the FC_LiDAR(RR) model demonstrates the lowest overall predictive capability of overhead FC (annulus rings 1-4) (r² = 0.70), with a slight improvement for the FC_LiDAR(FR) model (r² = 0.74). The intensity-based FC_LiDAR(IR) model displays the best results (r² = 0.78). However, the FC_LiDAR(BL) model is considered generally more useful (r² = 0.75) because the associated line of best fit passes through the origin, has a slope near unity and produces a mean estimate of FC_HP within 5%. Therefore, FC_LiDAR(BL) requires the least calibration across a broad range of forest cover types. The FC_LiDAR(FR) and FC_LiDAR(RR) models, on the other hand, were found to be sensitive to variations in both canopy height and sensor pulse repetition frequency (or pulse power); i.e. changing the repetition frequency led to a systematic shift of up to 11% in the mean FC_LiDAR(RR) estimates while it had no effect on the intensity-based FC_LiDAR(IR) or FC_LiDAR(BL) models. While the intensity-based models were generally more robust, all four models displayed at least some sensitivity to variations in canopy structural class, suggesting that some calibration of FC_LiDAR might be necessary regardless of the model used. Short (< 2 m tall) or open canopy forest plots posed the greatest challenge to accurate FC estimation regardless of the model used.