Search

Abstract: British Columbia is currently experiencing the largest mountain pine beetle (Dendroctonus ponderosae Hopkins) epidemic on record. The spatial extent of this infestation highlights the need for large-area forest management. [...]

Wildfire is an important disturbance agent in Canada''s boreal forest. Optical remotely sensed imagery (e.g., Landsat TM/ETM+), is well suited for capturing horizontally distributed forest conditions, structure, and change, while Light Detection and Ranging (LIDAR) data are more appropriate for capturing vertically distributed elements of forest structure and change. The integration of optical remotely sensed imagery and LIDAR data provides improved opportunities to characterize post-fire conditions. The objective of this study is to compare changes in forest structure, as measured with a discrete return profiling LIDAR, to post-fire conditions, as measured with remotely sensed data. Our research is focused on a boreal forest fire that occurred in May 2002 in Alberta, Canada. The Normalized Burn Ratio (NBR), the differenced NBR (dNBR), and the relative dNBR (RdNBR) were calculated from two dates of Landsat data (August 2001 and September 2002). Forest structural attributes were derived from two spatially coincident discrete return LIDAR profiles acquired in September 1997 and 2002 respectively. Image segmentation was used to produce homogeneous spatial patches analogous to forest stands, with analysis conducted at this patch level. In this study area, which was relatively homogenous and dominated by open forest, no statistically significant relationships were found between pre-fire forest structure and post-fire conditions (r <0.5; p >0.05). Post-fire forest structure and absolute and relative changes in forest structure were strongly correlated to post-fire conditions (r ranging from −0.507 to 0.712; p <0.0001). Measures of vegetation fill (VF) (LIDAR capture of cross-sectional vegetation amount), post-fire and absolute change in crown closure (CC), and relative change in average canopy height, were most useful for characterizing post-fire conditions. Forest structural attributes generated from the post-fire LIDAR data were most strongly correlated to post-fire NBR, while dNBR and RdNBR had stronger correlations with absolute and relative changes in the forest structural attributes. Absolute and relative changes in VF and changes in CC had the strongest positive correlations with respect to dNBR and RdNBR, ranging from 0.514 to 0.715 (p <0.05). Measures of average inter-tree distance and volume were not strongly correlated to post-fire NBR, dNBR, or RdNBR. No marked differences were found in the strength or significance of correlations between post-fire structure and the post-fire NBR, dNBR, RdNBR, indicating that for the conditions present in this study area all three burn severity indices captured post-fire conditions in a similar manner. Finally, the relationship between post-fire forest structure and post-fire condition was strongest for dense forests (>60% crown closure) compared to open (26–60%) and sparse forests (10–25%). Forest structure information provided by LIDAR is useful for characterizing post-fire conditions and burn induced structural change, and will complement other attributes such as vegetation type and moisture, topography, and long-term weather patterns, all of which will also influence variations in post-fire conditions. [Copyright &y& Elsevier]

Abstract: The mountain pine beetle Dendroctonus ponderosae Hopkins is endemic to lodgepole pine, Pinus contorta var. latifolia Engelmann, forests in western Canada. However, the current beetle epidemic in this area highlights the challenges faced by forest managers tasked with prioritizing stands for mitigation activities such as salvage harvesting and direct control methods. In western Canada, the operational risk rating system for mountain pine beetle is based on biological knowledge gained from a rich legacy of stand-scale field studies. Due to the large spatial (millions of hectares affected) and temporal (over 10 years) extents of the current epidemic, new research into large-area mountain pine beetle processes has revealed further insights into the landscape-scale characteristics of beetle infested forests. In this paper, we evaluated the potential for this new knowledge to augment an established system for rating the short-term risk of tree mortality in a stand due to mountain pine beetle. New variables explored for utility in risk rating include direct shortwave radiation, site index, diameter at breast height, the temporal trends in local beetle populations, Biogeoclimatic Ecosystem Classification and beetle–host interaction variables. Proportional odds ordinal regression was used to develop a model for the Vanderhoof Forest District in west-central British Columbia. Prediction on independent data was assessed with the area under the receiver operator curve (AUC), indicating good discriminatory power (AUC=0.84) for predicting levels of mountain pine beetle-caused pine mortality. [Copyright &y& Elsevier]

Abstract: Large area land cover products generated from remotely sensed data are difficult to validate in a timely and cost effective manner. As a result, pre-existing data are often used for validation. Temporal, spatial, and attribute differences between the land cover product and pre-existing validation data can result in inconclusive depictions of map accuracy. This approach may therefore misrepresent the true accuracy of the land cover product, as well as the accuracy of the validation data, which is not assumed to be without error. Hence, purpose-acquired validation data is preferred; however, logistical constraints often preclude its use — especially for large area land cover products. Airborne digital video provides a cost-effective tool for collecting purpose-acquired validation data over large areas. An operational trial was conducted, involving the collection of airborne video for the validation of a 31,000 km2 sub-sample of the Canadian large area Earth Observation for Sustainable Development of Forests (EOSD) land cover map (Vancouver Island, British Columbia, Canada). In this trial, one form of agreement between the EOSD product and the airborne video data was defined as a match between the mode land cover class of a 3 by 3 pixel neighbourhood surrounding the sample pixel and the primary or secondary choice of land cover for the interpreted video. This scenario produced the highest level of overall accuracy at 77% for level 4 of classification hierarchy (13 classes). The coniferous treed class, which represented 71% of Vancouver Island, had an estimated user''s accuracy of 86%. Purpose acquired video was found to be a useful and cost-effective data source for validation of the EOSD land cover product. The impact of using multiple interpreters was also tested and documented. Improvements to the sampling and response designs that emerged from this trial will benefit a full-scale accuracy assessment of the EOSD product and also provides insights for other regional and global land cover mapping programs. [Copyright &y& Elsevier]

Change detection approaches, such as computing change in spectral indexes through time, are a mature and established science, which is increasingly being applied in operational remote sensing programs. The quality and consistency of the changes detected using these approaches are linked, however, to the processing of the imagery required to address issues related to image radiometry, normalization, and computation of the spectral indexes. These processing steps are typically undertaken independently, providing opportunities for computation errors, increasing disk storage needs, and consuming processing time. In this letter, we present an approach for combining these processing steps to facilitate a more streamlined and computationally efficient approach to change detection using Landsat-5 and -7. The individual elements of the algorithm (raw Landsat-5 or -7, to calibrated Landsat-7, to top-of-atmosphere reflectance, to tasselled cap components) are described, followed by a description and illustration of the protocol to algebraically combine the elements. Rather than producing intermediate outputs, the sequentially integrated data processing protocol operates in memory and produces only the desired outputs. The proposed approach mitigates opportunities for inappropriate scaling between processing steps, the consistency of which is especially important for threshold-based change detection procedures. In addition, savings in both processing time and disk storage are afforded through the combination of processing steps, with processing of the time-1 images reduced from three to two stages and five to two stages for the time-2 images, resulting in savings of 50% and 69% in computing times and disk space requirements, respectively [ABSTRACT FROM PUBLISHER]

Abstract: Accurate spatial information on the location and extent of mountain pine beetle infestation is critical for the planning of mitigation and treatment activities. Areas of mixed forest and variable terrain present unique challenges for the detection and mapping of mountain pine beetle red-attack damage, as red-attack has a more heterogeneous distribution under these conditions. In this study, mountain pine beetle red-attack damage was detected and mapped using a logistic regression approach with a forward stepwise selection process and a set of calibration data representing samples of red-attack and non-attack from the study area. Variables that were considered for inclusion in the model were the enhanced wetness difference index (EWDI) derived from a time series of Landsat remotely sensed imagery, elevation, slope, and solar radiation (direct, diffuse, and global). The output from the logistic regression was a continuous probability surface, which indicated the likelihood of red-attack damage. Independent validation data were used to assess the accuracy of the resulting models. The final model predicted red-attack damage with an accuracy of 86%. These results indicate that for this particular site, with mixed forest stands and variable terrain, remotely sensed and ancillary spatial data can be combined, through logistic regression, to create a mountain pine beetle red-attack likelihood surface that accurately identifies damaged forest stands. The use of a probabilistic approach reduces dependence upon the definition of change by the application of thresholds (upper and lower bounds of change) at the image processing stage. Rather, a change layer is generated that may be interpreted liberally or conservatively, depending on the information needs of the end user. [Copyright &y& Elsevier]

Copyright of Forestry Chronicle is the property of Canadian Institute of Forestry and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Seven Landsat images near Prince George, British Columbia, Canada, representing a range of within-year and between-year dates, were acquired to assess spectral variability and the concomitant impact upon hyperclustering classification results. Top-of-atmosphere (TOA) radiometric corrections, dark target subtractions and geometric corrections were applied to the imagery. Following application of an unsupervised hyperclustering procedure which employed the K-means classifier, post-classification comparisons examined the differences in spectral response patterns for several target classes, and area summaries were generated to compare the variability in the total area of classes as identified in each image. Finally, the kappa coefficient of agreement was used to quantify the degree of correspondence between the classified images. The results indicated that the spectral variability of the within-year image set exceeded the variability in the between-year image set and differences in class area were highly variable over the range of image acquisition dates. These findings suggest that off-year imagery (acquired on or near anniversary dates) may be preferred to off-season imagery when building large-area Landsat mosaics for land cover classification using the hyperclustering procedure. [ABSTRACT FROM AUTHOR]

Modeling of forest carbon (C) dynamics requires precise information regarding when a disturbance occurred and the age of regeneration present. Generally, this information is obtained in the age class attribute of forest inventories; however forest inventories can become quickly outdated when disturbance events are not continuously integrated into the database. In this study, Landsat ETM+ image data and Tasseled Cap index values were used to estimate the age of lodgepole pine (Pinus contorta) stands from the approximate time of disturbance to 20 years of regeneration. An image segmentation procedure aided the removal of pixels representative of residual forest and other non-characteristic stand conditions within forest inventory polygons, in order to isolate the pixels representative of regenerating harvested forest. A stepwise multivariate regression procedure was used to estimate stand age for harvested areas, and an R2 of 0.68 (with a standard error of less than 2.4 years) was computed. This transferable approach provides useful information for forest C accounting when the year of disturbance, or the age of subsequent regeneration, is required for estimating C stocks. [Copyright &y& Elsevier]

A priority of the Canadian Forest Service and Canadian Space Agency joint project, Earth Observation for Sustainable Development of Forests (EOSD), is the production of a land cover map of the forested area of Canada based upon Landsat data. The land cover will be produced through a partnership of federal, provincial and territorial governments, universities, and industry. The short-term goal of EOSD is to complete a land cover map representing year 2000 forested area conditions by early 2006. Over the longer term, EOSD will aim to produce land cover products to capture changes in forest conditions over time to support national and international reporting requirements. The forested area of Canada represents approximately half of Canada’ s landmass, requiring over 450 scenes for complete coverage (with overlap minimized). EOSD is working with provincial and territorial mapping agencies that have on-going land cover mapping programs to optimize production capacity. It is envisioned that the combined output of EOSD and provincial and territorial land cover mapping programs will be integrated with maps developed by other sectors and agencies (such as agriculture) to produce a complete representation of the land cover of Canada. Large-area land cover mapping using remote sensing is a relatively new phenomenon. Advances in data storage capabilities, computing power, and increases in the affordability of data have allowed for large-area projects to be undertaken in ways previously not possible. The manner in which a large-area mapping project is approached is related to a number of factors including the spatial extent of the area of interest, the spatial resolution of the selected sensor, and the products which are to be generated. In this communication we report on the strategy, methods, and status of the EOSD land cover mapping program of the forested area of Canada. [ABSTRACT FROM AUTHOR]

The identification and classification of mountain pine beetle, Dentroctonus ponderosa (Hopkins), red-attack damage patterns in a mature lodgepole pine (Pinus contorta) forest located in the Fort St. James Forest District, British Columbia, was accomplished using 1999 Landsat TM satellite imagery, 1999 mountain pine beetle field and aerial survey point data, and G15 forest inventory data. Unrelated variance in the observed spectral response at mountain pine beetle field and aerial survey points was reduced following image stratification with the GIS forest inventory data and removal of other factors uncharacteristic of red-attack damage. Locations of known mountain pine beetle infestation were used to train a maximum-likelihood algorithm; overall classification accuracy was 73 percent, based on an assessment of 360 independent validation points. If local stand variability is reduced prior to signature generation, accuracies and map products can be useful for those involved in active forest management decisionmaking regarding mountain pine beetle infestations. [ABSTRACT FROM AUTHOR]

Numerous large-area, multiple image-based, multiple sensor land cover mapping programs exist or have been proposed, often within the context of national forest monitoring, mapping and modelling initiatives, worldwide. Common methodological steps have been identified that include data acquisition and preprocessing, map legend development, classification approach, stratification, incorporation of ancillary data and accuracy assessment. In general, procedures used in any large-area land cover classification must be robust and repeatable; because of data acquisition parameters, it is likely that compilation of the maps based on the classification will occur with original image acquisitions of different seasonality and perhaps acquired in different years and by different sensors. This situation poses some new challenges beyond those encountered in large-area single image classifications. The objective of this paper is to review and assess general medium spatial resolution satellite remote sensing land cover classification approaches with the goal of identifying the outstanding issues that must be overcome in order to implement a large-area, land cover classification protocol. [ABSTRACT FROM AUTHOR]

Abstract: Regression has been widely applied in Light Detection And Ranging (LiDAR) remote sensing to spatially extend predictions of total aboveground biomass (TAGB) and other biophysical properties over large forested areas. Sample (field) plot size has long been considered a key sampling design parameter and focal point for optimization in forest surveys, because of its impact on sampling effort and the estimation accuracy of forest inventory attributes. In this study, we demonstrate how plot size and co-registration error interact to influence the estimation of LiDAR canopy height and density metrics, regression model coefficients, and the prediction accuracy of least-squares estimators of TAGB. We made use of simulated forest canopies and synthetic LiDAR point clouds, so that we could maintain strict control over the spatial scale and complexity of forest scenes, as well as the magnitude and type of planimetric error inherent in ground-reference and LiDAR datasets. Our results showed that predictions of TAGB improved markedly as plot size increased from 314 (10m radius) to 1964m2 (25m radius). The co-registration error (spatial overlap) between ground-reference and LiDAR samples negatively impacted the estimation of LiDAR metrics, regression model fit, and the prediction accuracy of TAGB. We found that larger plots maintained a higher degree of spatial overlap between ground-reference and LiDAR datasets for any given GPS error, and were therefore more resilient to the ill effects of co-registration error compared to small plots. The impact of co-registration error was more pronounced in tall, spatially heterogeneous stands than short, homogeneous stands. We identify and briefly discuss three possible ways that LiDAR data could be used to optimize plot size, sample selection, and the deployment of GPS resources in forest biomass surveys. [ABSTRACT FROM AUTHOR]

Abstract: Quantifying aboveground biomass in forest ecosystems is required for carbon stock estimation, aspects of forest management, and further developing a capacity for monitoring carbon stocks over time. Airborne Light Detection And Ranging (LiDAR) systems, of all remote sensing technologies, have been demonstrated to yield the most accurate estimates of aboveground biomass for forested areas over a wide range of biomass values. However, these systems are limited by considerations including large data volumes and high costs. Within the constraints imposed by the nature of the satellite mission, the GeoScience Laser Altimeter System (GLAS) aboard ICESat has provided data conferring information regarding forest vertical structure for large areas at a low end user cost. GLAS data have been demonstrated to accurately estimate forest height and aboveground biomass especially well in topographically smooth areas with homogeneous forested conditions. However in areas with dense forests, high relief, or heterogeneous vegetation cover, GLAS waveforms are more complex and difficult to consistently characterize. We use airborne discrete return LiDAR data to simulate GLAS waveforms and to subsequently deconstruct coregistered GLAS waveforms into vegetation and ground returns. A series of waveform metrics was calculated and compared to topography and vegetation information gleaned from the airborne data. A model to estimate maximum relief directly from waveform metrics was developed with an R 2 of 0.76 (n =110), and used for the classification of the maximum relief of the areas sensed by GLAS. Discriminant analysis was also conducted as an alternative classification technique. A model was also developed estimating forest canopy height from waveform metrics for all of the data (R 2 =0.81, n =110) and for the three separate relief classes; maximum relief 0–7m (R 2 =0.83, n =44), maximum relief 7–15m (R 2 =0.88, n =41) and maximum relief >15m (R 2 =0.75, n =25). The moderate relief class model yielded better predictions of forest height than the low relief class model which is attributed to the increasing variability of waveform metrics with terrain relief. The moderate relief class model also yielded better predictions than the high relief class model because of the mixing of vegetation and terrain signals in waveforms from high relief footprints. This research demonstrates that terrain can be accurately modeled directly from GLAS waveforms enabling the inclusion of terrain relief, on a waveform specific basis, as supplemental model input to improve estimates of canopy height. [Copyright &y& Elsevier]

Abstract: Areas of similar ecology are often delineated based on homogenous topography, temperature, and land cover. Once delineated, these zones become the basis for a wide variety of scientific research and management activities. For instance, in Canada, ecozones are commonly utilized ecological management units delineated using geographic, topographic, and climatic information aided by spring and summer vegetation conditions. Snow cover has an influence on local and regional hydrological conditions and climate, as well as on animal habitats. As such, we posit that inclusion of winter conditions, incorporating spatial- and temporal-variation in snow cover is an additional element for consideration when delineating areas with homogenous conditions. In our analysis we use satellite passive microwave brightness temperatures from 19years of Special Sensor Microwave/Imager (SSM/I) measurements to produce a daily time-series on snow cover, and demonstrate how these data can be used to delineate areas of similar winter conditions. We use splines and curve fitting to generalize the dense time-series (of over 6900days) to a set of metrics, and select three for use in cluster-based generalization of snow cover regimes: annual maximum difference between 37 and 19GHz SSM/I measurements (with differences in magnitudes indicative of snow accumulation), variation of 37–19GHz brightness temperatures (indicative of snow cover variability), and variation in the rate of brightness temperature change during the snow melt season (indicative of seasonal change). Our results indicate that these metrics produce spatial units that are unique, and not captured by conventional ecological management units, while also producing spatial units that cohere to those generated from summer conditions. Spatial units that are found to have spatial cohesion between summer and winter data sources are located in regions where the amount of snow tends to be low, and snow cover variability minimal. We propose that snow cover regimes may be used to augment typical vegetation-based ecological zonations or to provide insights on hydrology and animal habitat conditions. Inclusion of winter conditions is especially important when areal delineations are used to monitor impacts of climate change, and as a baseline for monitoring changes in snow cover amount, extent, and/or distribution. [Copyright &y& Elsevier]

Imputations of missing values and optimal smoothing with massive data arrays poses a computational challenge since ordinary kriging becomes infeasible. Imputation and smoothing with standard algorithms like inverse distance weighted nearest neighbour interpolation (IDW) and interpolation on triangulated irregular networks (TIN/IP) fail to incorporate the spatial structure and ignore information beyond the neighbourhood. Multiresolution spatial models (MRSM) or approximate kriging methods adapted to handling massive data sets can be expected to do better than IDW and TIN/IP in terms of mean square errors of prediction (MSEP). We illustrate a MRSM that is efficient, computationally fast, and easy to implement. In two forestry examples with imputation of LiDAR range values the MRSM achieved a lower MSEP than IDW, TIN/IP, and fixed ranked kriging. MRSM appear as especially attractive for the construction of a DTM from last return LiDAR pulses. A third example demonstrates MRSM for efficient smoothing. [Copyright &y& Elsevier]