Joan E. Luther

Pixel-Based Image Compositing for Large-Area Dense Time Series Applications and Science

Abstract Free and open access to the more than 40 years of data captured in the Landsat archive, combined with improvements in standardized image products and increasing computer processing and storage capabilities, have enabled the production of large-area, cloud-free, surface reflectance pixel-based image composites. Best-available-pixel (BAP) composites represent a new paradigm in remote sensing that is no longer reliant on scene-based analysis. A time series of these BAP image composites affords novel opportunities to generate information products characterizing land cover, land cover change, and forest structural attributes in a manner that is dynamic, transparent, systematic, repeatable, and spatially exhaustive. Herein, we articulate the information needs associated with forest ecosystem science and monitoring in a Canadian context, and indicate how these new image compositing approaches and subsequent derived products can enable us to address these needs. We highlight some of the issues and opportunities associated with an image compositing approach and demonstrate annual composite products at a national-scale for a single year, with more detailed analyses for two prototype areas using 15 years of Landsat data. Recommendations concerning how to best link compositing decisions to the desired use of the composite (and the information need) are presented, along with future research directions. Résumé L’accès libre et gratuit à plus de 40 ans de données dans l’archive Landsat combiné à l’amélioration des produits d’imagerie standardisés et l’augmentation des capacités de traitement et de stockage informatiques ont permis la production d’images composites basées sur les pixels de réflectance de surface de grande superficie sans nuages. Les composites du « meilleur pixel disponible » (best-available-pixel; BAP) représentent un nouveau paradigme en matière de télédétection qui ne dépend plus de l’analyse par scène. Une série chronologique de ces images composites BAP offre de nouvelles occasions de générer des produits d’information qui caractérisent la couverture terrestre, le changement de la couverture terrestre et les attributs structurels de la forêt d’une manière dynamique, transparente, systématique, répétable et spatialement exhaustive. Ici, nous articulons les besoins d’information liés à la science et à la surveillance des écosystèmes forestiers dans un contexte canadien, et nous indiquons comment ces nouvelles approches de composition d’image et les produits qui en découlent peuvent nous permettre de répondre à ces besoins. Nous soulignons quelques-uns des problèmes et des possibilités associés à une approche de composition d’image et nous démontrons des produits composites annuels à l’échelle nationale pour une année, avec des analyses plus détaillées pour deux zones prototypes utilisant 15 ans de données Landsat. Des recommandations concernant la meilleure façon de lier des décisions de composition d’images à l’utilisation souhaitée du composite (et le besoin d’information) ainsi que les orientations futures de la recherche sont présentées.

Development of an Index of Balsam Fir Vigor by Foliar Spectral Reflectance

The potential to measure indices of forest vigor from foliar spectral reflectance was assessed in a range of balsam fir conditions achieved through stand manipulations. In order of increasing vigor, treatments consisted of root pruning, control, thinning, and thinning in combination with fertilization. Concentrations of chlorophyll and nitrogen increased significantly from the low to high vigor plots, and shoot lengths increased by more than three times. Balsam fir shoots from each treatment were measured for reflectance under laboratory conditions by a portable spectroradiometer with a spectral range from 350 nm to 2500 nm. Reflectance decreased consistently with vigor throughout the spectral region analyzed. Most wavebands showed significant effects of the treatments, with the most significant wavebands in the region of chlorophyll absorption between 500 nm and 740 nm. Foliage age class and sample date also affected reflectance in these wavebands. Reflectance indices, defined as the ratio and normalized difference of the most significant waveband at 711 nm and the least significant waveband at 913 nm, were affected by the treatments (P<0.0001). The normalized difference index correlated strongly with concentrations of chlorophyll a (r2⩾0.75), chlorophyll b (r2⩾0.66), nitrogen (r2⩾0.52), and shoots lengths (r2⩾0.69), thereby comprising a spectral index of balsam fir vigor that integrates several aspects of forest physiological condition.

Forecasting the susceptibility and vulnerability of balsam fir stands to insect defoliation with Landsat Thematic Mapper data

Abstract The potential of remote sensing to aid in the forecast of insect defoliation and the associated impact on forest stands was tested by examining the relationships between i) preoutbreak forest structure and growth characteristics, ii) multispectral values recorded by the Landsat Thematic Mapper (TM) sensor, and iii) subsequent patterns of defoliation and volume loss in the balsam fir ( Abies balsamea L. Mill) forests of Western Newfoundland. Forest structure and growth characteristics were measured in the field in 1992–1993 and reconstructed to represent conditions 3 years prior to an infestation of black-headed budworm, Acleris variana (Fern.), which began in 1988. Preoutbreak Landsat TM data were acquired on 8 August 1985. Spectral values were converted to approximate reflectances using PCI atmospheric correction routines and related to the field measurements. Logistic regression techniques were used to develop models for predicting forest susceptibility (probability of attack) and vulnerability (response to attack) and to assess the accuracy of the susceptibility and vulnerability forecasts. In general, susceptible stands were younger and had lower basal area and tree density, but higher leaf area index than stands that were not attacked. Vulnerable stands were older, had lower vigor and growth efficiency indices (stemwood growth per unit of leaf area), and also displayed the highest near infrared spectral values. The best predictions combined selected spectral measurements with provincial forest inventory data. These models produced classifications with accuracies of 81%, 67%, and 78% for predicting susceptibility and pre- and postoutbreak vulnerability, respectively.

Use of vector polygons for the accuracy assessment of pixel-based land cover maps

Identifying appropriate validation sources for large-area land cover products is a challenge, with logistical constraints frequently necessitating the use of preexisting data sources. Several issues exist when comparing polygon (vector-based) datasets to raster imagery: geolocational mismatches, differences in features or classes mapped, disparity between the scale of polygon delineation and the spatial resolution of the image, and temporal discrepancies. To evaluate the potential impact of using vector coverages to assess the accuracy of pixel-based land cover maps, five evaluation protocols are applied to test sites located in British Columbia and Newfoundland and Labrador, Canada. One protocol directly compared the land cover of the sample unit to the land cover of the forest inventory polygon within which the sample unit fell, two protocols used different regions around the sample unit to define the land cover class, and two protocols were based on homogeneity criteria that restricted the selection of sample units. For the protocols tested, the overall accuracy values ranged from 34% to 58%. Given the broad range of accuracies achieved, the results suggest that caution is needed when making spatially explicit comparisons between raster and vector datasets. When possible, the use of purpose-collected validation data is recommended for the accuracy assessment of maps derived from remotely sensed data; if preexisting vector-based data are the only option for the validation, approaches accounting for the heterogeneity of classes within a given polygon are recommended.

Development of large-area land cover and forest change indicators using multi-sensor Landsat imagery: Application to the Humber River Basin, Canada

Abstract Monitoring ecological indicators is important for assessing impacts of human activities on ecosystems. A means of identifying and applying appropriate indicators is a prerequisite for: environmental assessment; better assessment and understanding of ecosystem health; elucidation of biogeochemical trends; and more accurate predictions of future responses to global change, particularly those due to anthropogenic disturbance. The challenge is to derive meaningful indicators of change that capture the complexities of ecosystems yet can be monitored consistently over large areas and across time. In this study, methods for monitoring indicators of land cover (LC) and forest change were developed using multi-sensor Landsat imagery. Mapping and updating procedures were applied to the Humber River Basin (HRB) in Newfoundland and Labrador, one of four test sites in Canada selected for testing the development of national-scale methods. Procedures involved unsupervised clustering and labeling of baseline imagery, followed by image-to-image spectral clustering to derive binary change masks within which new LC types were classified for non-baseline imagery. Updated maps were compatible with the baseline map and reflected change in LC for three time periods: 1976–1990, 1990–2001, and 2001–2007. From the LC products, several change indicators were quantified including: forest depletion, forest regeneration, forest change, net forest change, and annual rates of change. The procedures were validated using field plots to assess the accuracy of the 2007 LC product (74.2% for 10 LC classes) and change classes observed from 2001 to 2007 (87.8% for four change classes: depletion, regeneration, non-treed class no change, and treed class no change). Methods were considered to be highly efficient and operationally feasible over large areas spanning multiple Landsat scenes. Specific results for the test site provided trend information supporting land and resource management in the HRB region.

Moving Toward Consistent ALS Monitoring of Forest Attributes across Canada

As airborne laser scanning (ALS) gains wider adoption to support forest operations in Canada, the consistency and quality of derivative products that support long-term monitoring and planning are becoming a key issues for managers. The Canadian Consortium for Lidar Environmental Applications Research (C-CLEAR) has supported almost 200 projects across Canada since 2000, with forest-related studies being a dominant theme. In 2010 and 2011, field operations were mobilized to support 13 ALS projects spanning almost the full longitudinal gradient of Canada’s forests. This paper presents case studies for seven plus an overview of some best practices and data processing workflow tools that have resulted from these consortium activities. Although the projects and research teams are spread across Canada, the coordination and decade of experience provided through C-CLEAR have brought common methodological elements to all. It is clear that operational, analytical and reporting guidelines that adhere to community accepted standards are required if the benefits promised by ALS forestry are to be realized. A national Lidar Institute that builds upon the C-CLEAR model and focuses on developing standards, guidelines, and certified training would address this need.

Recent changes at the northwest margin of the Barnes Ice Cap, Baffin Island, N. W. T. , Canada

A climate change monitoring site has been established at the northwest margin of the Barnes Ice Cap, in the vicinity of the Lewis Glacier. Three years of climatic data (1989 to 1992) and field observations, supplemented by satellite imagery, provide the basis for updating previous studies of local change at the ice cap margin, including climatology and substrate colonization by lichens. Climatic data from the ice cap summit permit extrapolation of seasonal temperatures from the ice marginal station as a basis for equilibrium line altitude estimates. Results are discussed against the background of studies from the 1960s. Retreat of the Lewis Glacier continues at about 25 m yr[sup [minus]1], whereas other areas of the northwest margin are retreating by 10 to 30 m yr[sup [minus]1]. Lower regional summer temperatures over the past three decades have not significantly slowed the recession that has been underway in this sector of the ice cap for the past three centuries. 41 refs., 7 figs., 2 tabs.

Modeling and mapping forest biomass using forest inventory and Landsat TM data: results from the Foothills Model Forest, Alberta

Forest biomass information is needed for reporting of selected indicators of sustainable forest management and for models that estimate carbon budgets and forest productivity, particularly within the context of a changing climate. In collaboration with the Canadian Space Agency, a strategy for mapping Canadas forest biomass has been developed as part of the Earth Observation for Sustainable Development of Forests (EOSD) project. This paper reports on the results derived from an application of this strategy to a pilot study area in the Foothills Model Forest, Alberta. Methods to estimate forest biomass have been developed using tree-level inventory plot data that is then extrapolated to the stand level by statistical relationships between biomass density and stand structural characteristics. These ground-based biomass estimates serve as source data that are related to stand structure derived from classified Landsat TM data. Models developed from inventory data to estimate biomass density attained adjusted R/sup 2/ values that ranged from 0.60 to 0.77 for 5 species groups, and tests with an independent validation sample compared favourably for all species (deciduous, lodgepole pine, mixed species, white spruce/fir), except black spruce/larch. Landsat-derived forest biomass was statistically and moderately correlated to the inventory-derived biomass with values of 0.63, 0.68, and 0.70 for conifer, deciduous, and mixed species, respectively. Research areas were identified from both inventory and remote sensing perspectives that will lead to incremental improvements in biomass estimation.

Earth observation for sustainable development of forests (EOSD)-a national project

A group of forested nations agreed upon a set of criteria and indicators to measure the sustainable development of forests. Of the 83 indicators, remote sensing could measure in whole or in part about 25 indicators. Moreover, such measurements could also support elements of Canadas national forest inventory and reporting requirements for the Kyoto protocol on reforestation, afforestation, and deforestation (RAD). The authors have commenced a new project, EOSD, to monitor the sustainable development of Canadas forests from space. Two federal partners, the Canadian Forest Service and the Canadian Space Agency, are creating the ten-year project in cooperation with the provinces and territories. The project will support, with space based technology, Canadas priorities and international commitments for monitoring the sustainable development of its forests and for meeting core forest information needs of the Kyoto Protocol. The status and major changes in the composition, distribution, structure and function of forests over time will be quantified. The remote sensing observations will also form part of a three-stage new national forest inventory. It is intended that products and data will be widely available via intelligent information systems. This paper will describe the products to be produced, and outline the project implementation.

Validation of L-Architect model for balsam fir and black spruce trees with structural measurements

The fine reconstruction of tree structure provides important information that is relevant to forest ecological processes and may be linked to wood quality attributes. Terrestrial laser scanners (TLS) provide detailed and accurate 3-D data and have the potential to enhance forest inventories with fine-scale information on structure. However, in forests, TLS data are limited by the effects of object occlusion and wind. Therefore, adequate tree reconstruction is not possible. In previous studies, we used the architectural model L-Architect (LiDAR data to tree Architecture) to overcome the limitations of TLS to characterize the 3-D tree structure. In this study, L-Architect was validated using structural measurements for two coniferous species found in Newfoundland (Canada), namely, balsam fir (Abies balsamea) and black spruce (Picea mariana). The model reproduced realistic tree structures using TLS data and allometric relationships to define the total amount of foliage following two main steps: branch growth and addition of foliage within the crown. Results from L-Architect were compared with measurements of branching structure from the main stem (angle of insertion, length, and diameter) and vertical distribution of foliage indicated by the shoot vertical distribution for six sampled trees. Averaged normalized root-mean square errors of 21%, 20%, 25%, and 21% were obtained for branch diameter, branch length, angle of branch insertion, and vertical distribution of foliage, respectively.