Authors:email@example.com (M. Penner et al) Abstract: Canadian Journal of Remote Sensing, Volume 39, Issue 05, Page 426-443, October 2013.
Parametric and nonparametric predictions of forest inventory attributes from airborne LiDAR data are compared for a forest management unit in boreal Ontario. For the parametric approach, seemingly unrelated regression models were calibrated by forest type (SUR) and for all forest types combined (SUR_All). For the nonparametric approach, randomForest (RF) and k-nearest neighbours (kNN) were implemented. Calibration data consisted of 442 circular 0.04 ha plots covering a range of development stages within eight forest types. Results were validated on 64 independent plots distributed across the same forest types. Predicted variables included top height, merchantable basal area, and gross merchantable volume. In general, RF and SUR predictions were the most accurate and precise, whereas kNN and SUR_All predictions were less reliable. Prediction accuracy and precision varied markedly with forest type, with no single method producing results that were consistently best. None of the methods extrapolated well, underscoring the need to capture the full range of population variation during calibration. Parametric predictions were improved by forest-type stratification, necessitating a population forest-type layer prior to application. In contrast, forest type was not an important predictor in the nonparametric solutions. RF can offer significant operational advantages over parametric regression without loss of accuracy or precision. PubDate: Wed, 04 Dec 2013 08:00:00 GMT
Authors:firstname.lastname@example.org (Mark J. Ducey et al) Abstract: Canadian Journal of Remote Sensing, Volume 39, Issue 05, Page 410-425, October 2013.
Nondetection of trees is a serious problem for the use of terrestrial laser scanning (TLS) in forest inventory applications. The use of multiple coregistered scans can reduce nondetection but may not eliminate it, and it carries substantial field and post-processing costs. We examined and extended previously developed theoretical approaches to modeling nondetection. The results suggested that tree size as well as multiple stand structural characteristics may be factors, but the theoretical models do not lend themselves to empirical estimation. We then used distance sampling techniques to identify detection probabilities and develop adjusted estimates for trees per hectare and basal area in nine forest stands in southern Norway. The results compared favorably with field estimates based on fixed-area plots. The estimated detection probabilities indicate that correction for nondetection is needed unless the search for trees is limited to very small distances from the scanner. Distance sampling appears promising when TLS is used in the context of temporary-plot forest inventories. PubDate: Wed, 04 Dec 2013 08:00:00 GMT
Authors:email@example.com (Alejandro Lorenzo Gil et al) Abstract: Canadian Journal of Remote Sensing, Volume 39, Issue 05, Page 396-409, October 2013.
This paper compares two types of digital terrain models (DTMs) with ground elevation measures collected through field work in a dense forest area on the island of Tenerife (Canary Islands, Spain). The first was an existing DTM derived from altimetric features obtained by manual photogrammetric restitution. The second DTM was computed from aerial LiDAR data with a nadir density of 0.8 points·m−2. Both DTMs have a pixel size of 5 m. The field work consisted of measuring three elevation profiles by land surveying techniques using a total station survey and taking into account different vegetation covers. The analysis of the profiles by means of nonparametric techniques showed an accuracy at the 95th percentile between 0.54 m and 24.26 m for the photogrammetry-derived DTM and between 0.22 m and 3.20 m for the LiDAR-derived DTM. Plotting the elevation profiles allowed for the visual detection of locations where the models failed. The LiDAR data were able to reflect more accurately the true ground surface in areas of dense vegetation, especially in places where the ground was invisible to photogrammetric operators as in the case of Canarian pine forest with understory. PubDate: Wed, 04 Dec 2013 08:00:00 GMT
Authors:firstname.lastname@example.org (Martin A. Montes-Hugo et al) Abstract: Canadian Journal of Remote Sensing, Volume 0, Issue 0, Page 1-11, e-First articles.
Historical regional budgets of suspended particulate matter (SPM) in surface waters (i.e., 0–20 m depth) of the St. Lawrence Estuary (SLE) are not very accurate because of the lack of in situ measurements. The aim of this study was to improve the accuracy of SPM budgets in the SLE based on spaceborne ocean color measurements by investigating one kind of uncertainty on satellite-derived SPM concentration and related with the influence of the bottom depth and (or) bottom reflectivity (hereafter bottom effects) on water-leaving radiance. Theoretical results suggest that bias on optically derived SPM concentration due to bottom effects is greater in the lower estuary and can be up to 36.6% when the water depth is ≤10 m, sediments are dominated by sand, and water turbidity is relatively low (beam attenuation coefficient PubDate: Wed, 27 Nov 2013 16:17:02 GMT
Authors:email@example.com (Mikko Vastaranta et al) Abstract: Canadian Journal of Remote Sensing, Volume 0, Issue 0, Page 1-14, e-First articles.
Airborne laser scanning (ALS) has demonstrated utility for forestry applications and has renewed interest in other forms of remotely sensed data, especially those that capture three-dimensional (3-D) forest characteristics. One such data source results from the advanced processing of high spatial resolution digital stereo imagery (DSI) to generate 3-D point clouds. From the derived point cloud, a digital surface model and forest vertical information with similarities to ALS can be generated. A key consideration is that when developing forestry related products such as a canopy height model (CHM), a high spatial resolution digital terrain model (DTM), typically from ALS, is required to normalize DSI elevations to heights above ground. In this paper we report on our investigations into the use of DSI-derived vertical information for capturing variations in forest structure and compare these results to those acquired using ALS. An ALS-derived DTM was used to provide the spatially detailed ground surface elevations to normalize DSI-derived heights. Similar metrics were calculated from the vertical information provided by both DSI and ALS. Comparisons revealed that ALS metrics provided a more detailed characterization of the canopy surface including canopy openings. Both DSI and ALS metrics had similar levels of correlation with forest structural attributes (e.g., height, volume, and biomass). DSI-based models predicted height, diameter, basal area, stem volume, and biomass with root mean square (RMS) accuracies of 11.2%, 21.7%, 23.6%, 24.5%, and 23.7%, respectively. The respective accuracies for the ALS-based predictions were 7.8%, 19.1%, 17.8%, 17.9%, and 17.5%. Change detection between ALS-derived CHM (time 1) and DSI-derived CHM (time 2) provided change estimates that demonstrated good agreement (r = 0.71) with two-date, ALS only, change outputs. For the single-layered, even-aged stands under investigation in this study, the DSI-derived vertical information is an appropriate and cost-effective data source for estimating and updating forest information. The accuracy of DSI information is based on a capability to measure the height of the upper canopy envelope with performance analogous to ALS. Forest attributes that are well captured and subsequently modeled from height metrics are best suited to estimation from DSI metrics, whereas ALS is more suitable for capturing stand density. Further investigation is required to better understand the performance of DSI-derived height products in more complex forest environments. Furthermore, the difference in variance captured between ALS and DSI-derived CHM also needs to be better understood in the context of change detection and inventory update considerations. PubDate: Wed, 20 Nov 2013 17:40:20 GMT
Authors:firstname.lastname@example.org (Ying Zhang et al) Abstract: Canadian Journal of Remote Sensing, Volume 0, Issue 0, Page 1-15, e-First articles.
A “hybrid”, high throughput image processing system is described to provide updated urban information on Canadian topographic maps from moderate resolution image sources such as SPOT. Two features are of particular interest; Buildings to Scale (BTS) (i.e., large buildings to be represented by individual symbols), and Built-up Areas (BUA), the majority of urban developed encompassing smaller structures (typically residential areas) that are represented on maps in a generalized way by a shading or tint. This paper addresses mapping of the latter, hereafter referred to as BUA mapping. A data-driven approach that utilizes a linear vegetation index to reduce data dimensionality and facilitate automation was developed. Full scene interpretation can be completed in approximately 2–3 minutes even on low-cost desktop computers. Accuracy of the BUA mapping output is comparable with that previously provided through traditional labour-intensive means (i.e., visual interpretation of aerial photography). Our process also provides a mapping of high intensity urban areas such as industrial parks and urban cores, thus providing search areas for subsequent extraction of large buildings. Finally, it is argued that the automated techniques described here have applicability in international urban mapping initiatives and the methodology can be adapted to the mapping of natural features such as water bodies and forests. PubDate: Wed, 20 Nov 2013 17:39:56 GMT
Authors:email@example.com (Douglas K. Bolton et al) Abstract: Canadian Journal of Remote Sensing, Volume 0, Issue 0, Page S1-S13, e-First articles.
Carbon storage in forest aboveground biomass is a critical, yet difficult, component of the global carbon cycle to estimate. Canopy height, a key indicator of carbon storage, can be estimated from Light Detection and Ranging (Lidar) waveforms collected by the Geoscience Laser Altimeter System (GLAS) aboard the Ice, Cloud, and land Elevation Satellite (ICESat). Although globally distributed, GLAS does not provide spatially exhaustive coverage. Therefore, accurate methods of extrapolation are necessary to produce wall-to-wall global canopy height maps from these data. In this analysis, we compare two of these global GLAS-derived height products to canopy height estimates derived from 25000 km of discrete return airborne Lidar data over Canada's boreal forests. We selected the 95th percentile of first return height from airborne Lidar as a measure of canopy height to relate against estimates from the global GLAS-derived products. The agreement between the global GLAS-derived products and airborne Lidar-derived height estimates varied between the two products (average ecozone RMSE = 3.9 and 7.4 m), demonstrating that differences in data selection, processing, and extrapolation can influence height estimates derived from GLAS data. Where large differences existed between the global GLAS-derived products and the airborne Lidar-derived height estimates, the GLAS-derived products tended to predict taller canopies. Removing GLAS waveforms on steep terrain appeared to be a superior approach to reducing errors in height estimates, as the global GLAS-derived product that filtered these waveforms was in closer agreement with airborne Lidar-derived height estimates in regions of rough terrain (RMSE = 3.2–8.5 m compared with 8.1–13.8 m). Differences in the spatial resolution of canopy height estimates, coupled with varying definitions of canopy height within each product, should be considered when interpreting the results of this analysis. Investigating the relationship between small-footprint Lidar data and published canopy height products can identify the approaches that lead to the most accurate estimates of aboveground biomass and can help determine why discrepancies in height estimates exist between various model approaches, data and underlying environmental conditions. PubDate: Wed, 20 Nov 2013 17:38:57 GMT
Authors:firstname.lastname@example.org (Huaguo Huang et al) Abstract: Canadian Journal of Remote Sensing, Volume 0, Issue 0, Page S1-S13, e-First articles.
Our objective was to assess the effect of multiple scattering on lidar radiative transfer. We have developed a time-dependent radiosity-based model (RBL) to simulate the propagation of lidar pulses through forest canopies. This 3-D model enables simulation of lidar waveforms with varied topography and clumping vegetation. The incidence angle can also be specified. This new model has the potential to provide better approximations of return waveforms. The prototype is being tested using data from the Scanning Lidar Imager of Canopies by Echo Recovery (SLICER). Waveforms simulated by RBL resemble SLICER waveforms (R2 > 0.90) over a jack pine canopy and a black spruce canopy. There is also good agreement (R2 > 0.95) when the model results are compared with a time-dependent radiative transfer model. Results to date indicate that multiply scattered photons do increase the intensity of the reflected signal, especially the portion originating from the lower levels of the canopy. A sensitivity analysis enabled assessment of the effects of leaf area index, slope, and canopy height on multiple scattering. PubDate: Mon, 21 Oct 2013 18:29:49 GMT