Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a... more Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a major concern for arctic communities and resource development as warming temperatures have led to increasing permafrost degradation. In order to effectively assess and mitigate permafrost disturbance risk, disturbance-prone areas can be predicted through the use of susceptibility modelling. Permafrost disturbance susceptibility modelling is a quantitative assessment of the relationship between the distribution of past permafrost disturbances and a set of influencing terrain attributes that may cause slope instability. In this study we develop a universal permafrost disturbance susceptibility model using a limited disturbance inventory to test the applicability of the model for a broader region in the Canadian High Arctic. Additionally, we use the model to explore the effect of influencing terrain parameters on disturbance occurrence. To account for a large range of landscape characteris...
Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a... more Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a major concern for arctic communities and resource development as warming temperatures have led to increasing permafrost degradation. In order to effectively assess and mitigate permafrost disturbance risk, disturbance-prone areas can be predicted through the use of susceptibility modelling. Permafrost disturbance susceptibility modelling is a quantitative assessment of the relationship between the distribution of past permafrost disturbances and a set of influencing terrain attributes that may cause slope instability. In this study we develop a universal permafrost disturbance susceptibility model using a limited disturbance inventory to test the applicability of the model for a broader region in the Canadian High Arctic. Additionally, we use the model to explore the effect of influencing terrain parameters on disturbance occurrence. To account for a large range of landscape characteris...
Detailed forest ecosystem classifications have been developed for large regions of northern Ontar... more Detailed forest ecosystem classifications have been developed for large regions of northern Ontario. These ecosystem classifications provide tools for ecosystem management that constitute part of a larger goal of integrated management of forest ecosystems for long-term sustainability. These classification systems provide detailed stand-level characterization of forest ecosystems at a local level. However, for ecological approaches to forest management to become widely accepted by forest managers, and these tools to be widely used, methods must be developed to characterize and map or model ecosystem classes at landscape scales for large regions. In this study, the site-specific Northwestern Ontario Forest Ecosystem Classification (NWO FEC) was adapted to provide a landscape-scale (1:20000) forest ecosystem classification for the Rinker Lake Study Area located in the Boreal Forest north of Thunder Bay, Ontario. Multispatial resolution remote sensing data were collected using the Compact Airborne Spectrographic Imager (CASI) and analysed using geostatistical techniques to obtain an understanding of the nature of the spatial dependence of spectral reflectance for selected forest ecosystems at high spatial resolutions. Based on these analyses it was determined that an optimal size of support for characterizing forest ecosystems (i.e., optimal spatial resolution), as estimated by the mean ranges of a series of experimental semivariograms, differed based on (i) wavelength; (ii) forest ecosystem class (and at low altitude as a function of mean maximum canopy diameter (MMCD)); and (iii) altitude of the remote sensing system. In addition, maximum semivariance as estimated from the sills of the experimental semivariograms increased with density of understory. Based on the estimates for optimal spatial resolutions for six landscape-scale forest ecosystem classes, a series of spectral-spatial features were derived from the high-altitude CASI data (4 metre spatial resolution) using spatial averaging. Linear discriminant analysis for various spectral-spatial and texture feature combinations indicated that a spatial resolution of approximately 6 m was optimal for discriminating the six-landscape scale ecosystem classes. Texture features, using second-order spatial statistics that were derived from the 4 m remote sensing data, also significantly improved discrimination of the classes over the original 4 m data. Finally, addition of terrain descriptors, particularly elevation within a local region, improved discrimination of the six landscape scale ecosystem classes. It has been demonstrated that in a low-relief boreal environment, addition of textural and geomorphometric variables to high-resolution CASI reflectance data provides improved discrimination of forest ecosystem classes. Although these improvements are statistically significant, the absolute classification accuracies are not at levels suitable for operational classification and mapping. The analysis presented here represents the initiation of a complex modelling approach that is necessary for improving forest ecosystem characterization and prediction using additional primary datasets and derived datasets that possess various levels of measurement. Not only are optimal or multispatial resolution remote sensing data required, but also appropriately scaled terrain and landscape features depicting soil texture, nutrient and moisture regimes. Incorporation of these types of terrain-specific variables with reflectance data should provide further improvement in forest ecosystem classification and modelling at landscape scales.
Carbon dioxide, water vapour, and energy fluxes vary spatially and temporally within forested env... more Carbon dioxide, water vapour, and energy fluxes vary spatially and temporally within forested environments. However, it is not clear to what extent they vary as a result of variability in the spatial distribution of biomass and elevation. The following study presents a new methodology for extracting changes in the structural characteristics of vegetation and elevation within footprint areas, for direct comparison with eddy covariance (EC) CO2 flux concentrations. The purpose was to determine whether within-site canopy structure and local elevation influenced CO2 fluxes in a mature jack pine (Pinus banksiana Lamb.) forest located in Saskatchewan, Canada. Airborne light detection and ranging (lidar) was used to extract tree height, canopy depth, foliage cover, and elevation within 30 min flux footprints. Within-footprint mean structural components and elevation were related to 30 min mean net ecosystem productivity (NEP) and gross ecosystem production (GEP). NEP and GEP were modeled using multiple regression, and when compared with measured fluxes, almost all periods showed improvements in the prediction of flux concentration when canopy structure and elevation were included. Increased biomass was related to increased NEP and GEP in June and August when the ecosystem was not limited by soil moisture. On a daily basis, fractional cover and elevation had varying but significant influences on CO 2 fluxes.
This study investigates the potential of discrete return light detection and ranging (lidar) data... more This study investigates the potential of discrete return light detection and ranging (lidar) data to characterize forest succession in a mixed mature forest in central Ontario using indices applied to the lidar point cloud. Derived indices include statistical indices, predicted Lorey's height (R 2 ϭ 0.86; RSME ϭ 2.36 m) and quadratic mean diameterat-breast-height (R 2 ϭ 0.68; RMSE ϭ 1.21 cm), canopy density indices and an information theory based complexity index. To assess how well these indices are able to capture the vertical structure of forest stands, they are compared to four stages of forest stand development. Best subsets regressions indicated that no single index is able to separate all four stages adequately. However, the predicted Lorey's height index is optimal for separating early from mid succession stages (p Ͻ.0001) and a combination of height and complexity indices performed best to discriminate between mid-and late-succession stages (p Ͻ.0001).
Literature from the past two decades documents how airborne LIDAR can be used to predict forest i... more Literature from the past two decades documents how airborne LIDAR can be used to predict forest inventory variables, such as basal area, volume, and biomass, at the plot and stand level. However, a key question that has yet to be fully addressed, and that the forest industry continues to ask as it considers operationalizing the use of LIDAR in forest resource inventories, is: What is the optimal point density for predicting forest inventory variables? For example, is a point density of 0.5 points/m 2 sufficient for making accurate predictions of forest inventory variables or is a point density of 3 points/m 2 the minimum required? To investigate this key question, a three-year project was launched in 2007 in Ontario, Canada. Field data for approximately 180 plots, sampling a broad range of forest types and conditions in Ontario, were collected over two study sites. Airborne LIDAR data, characterized by 3 points/m 2 were systematically decimated to produce datasets with point densities of approximately 1.5 and 0.5 points/m 2 . Models, developed using stepwise regression, were developed for each of the three lidar datasets to estimate several forest inventory variables including: (1) basal area (R 2 =0.25-0.94); (2) gross total volume (R 2 =0.46-0.95); (3) gross merchantable volume (R 2 =0.37-0.94); (4) stem density(R 2 =0.23-0.89); (5) quadratic mean diameter (R 2 =0.59-0.86); (6) total aboveground biomass (R 2 =0.26-0.93); (7) average height (R 2 =0.76-0.95); and (8) top height (R 2 =0.75-0.98). Aside from two cases, no decimation effect was found for predictions of forest variables, which suggests that a point density of 0.5 points/m 2 is sufficient for plot and stand level modelling under these forest conditions.
Sets of spectral, spectral-spatial, textural, and geomorphometric variables derived fram high spa... more Sets of spectral, spectral-spatial, textural, and geomorphometric variables derived fram high spatial resolution Compact Airborne Spectrographic Imager (CASI) and elevation data are tested to determine their ability to discriminate landscape-scale forest ecosystem classes for a study area in northern Ontario, Canada. First, linear discriminant analysis for various spectral and spectral-spatial variables indicated that a spatial resolution of approximately 6 m was optimal for discriminating six landscape-scale forest ecosystem classes. Second, texture features, using second-order spatial statistics, significantly improved discrimination of the classes over the original reflectance data. Finally, addition of terrain descriptors improved discrimination of the six forest ecosystem classes. It has been demonstrated that, in a low-to moderate-relief boreal environment, addition of textural and terrain variables to highresolution CASI reflectance data provides improved discrimination of forest ecosystem classes. P. Treitz was with the 1 J 0.591 2 J J 0.877 3 J J J 0.987 4 J J J J 1.081 5 J J J J J 1.136 6 J J J J J J 1.165 CASI Spectral Variable Spatial Resolution 1 2 3 4 5 6 CASI (580-601 nm) 4 meters 5 meters J J 6 meters J J CASI (663-678 nm) 4 meters J J 5 meters J J 6 meters J J CASI (744-750 nm) 5 meters 6 meters J J 7 meters J J J J *1. Aspen-Dominated Hardwood and Mixedwood; 2. White Spruce1 Balsam Fir Conifer and Mixedwood; 3. Cedar Mixedwood; 4. Upland Black SpruceIJack Pine; 5. Lowland Black Spruce; and 6. Wetland Black Spruce
The Kyoto Protocol has sharpened the focus on the possible role of forests in contributing to or ... more The Kyoto Protocol has sharpened the focus on the possible role of forests in contributing to or mitigating climate change. The understanding of carbon dynamics, and the prediction of future carbon stock changes, rely on an analysis of the past forest dynamics. Historical data may often be incomplete, approximate, or strongly generalized. To circumvent this problem, we have recently developed
IEEE International IEEE International IEEE International Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004, 2004
Abstract Interferometric synthetic aperture radar (InSAR) data from RADARSAT-1 have been examined... more Abstract Interferometric synthetic aperture radar (InSAR) data from RADARSAT-1 have been examined to assess their potential for mapping terrain and changes in snow cover characteristics, relative to the limiting effects of snow on foraging by endangered Peary caribou (Rangifer tarandus). Radar is one of the few observational tools that can provide information on the changing snow pack during the dark winter months. The goal of this research is to characterize the general ensemble of terrain characteristics that may affect ...
Thermokarst, active layer detachments (ALDs), and retrogressive thaw slumps (RTSs), representing ... more Thermokarst, active layer detachments (ALDs), and retrogressive thaw slumps (RTSs), representing three forms of permafrost degradation, constitute serious risks for infrastructure and have the potential to alter environmental and ecological conditions in Arctic regions. Environmental change and increased land development pressures require innovative cost-effective methods for assessing hazard prone areas. The overall research objective of this project is to design a landscape model to predict and characterize disturbance prone areas using key physiographic controls and geospatial modeling derived from satellite imagery to efficiently produce hazard susceptibility maps. Susceptibility maps identify disturbance prone areas, which are a fundamental component of hazard management and the basis for provision of measures aimed at reducing the risks resulting from permafrost degradation. To test the validity of this modelling approach and its applicability across the Arctic, methods are be...
This study examines the distribution of laser pulse returns obtained from coincident airborne and... more This study examines the distribution of laser pulse returns obtained from coincident airborne and terrestrial lidar surveys of a closed-canopy red pine (Pinus resinosa) plantation. The purpose of this study is to improve our understanding of laser pulse sampling within closed canopies so that estimates of forest structural variables (e.g., biomass, needle-leaf area, and base-of-live-crown) can be improved at the individual tree and stand levels using lidar. The results of this study indicate the following: (1) There is a statistically significant difference between field measurements of tree height and estimates derived from the maximum laser pulse return from airborne and terrestrial lidar. In both cases, maximum laser pulse returns underestimate tree height by 1 m, on average. (2) Both terrestrial and airborne lidar are unable to discern the base of the measured live crown. Laser pulse returns from airborne lidar are biased towards the top of the tree crown, i.e., lowest laser pul...
The ecosite unit in Ontario's boreal forest ecological land classification system is a polygon of... more The ecosite unit in Ontario's boreal forest ecological land classification system is a polygon of common vegetation type and soil conditions intended to provide a standardized provincial framework to inform meso-scale forestry and planning applications. To determine whether the physical factors used for ecosite classification relate to patterns in ecological function over finer spatial scales, we examined 14 soil properties in replicate boreal forest plots representing eight mineral soil ecosite classes and three organic soil ecosite classes in the Hearst Forest. Despite large differences in vegetation composition, we found few statistically significant differences in properties when compared for individual classes or for more general groupings based on vegetation type and soil texture or expected fertility status. However, some properties (soil organic carbon, total nitrogen, and C:N ratio) were approaching significance in the 0-10 cm depth increment, and there were distinct differences between organic soil and mineral soil sites. Overall, these results suggest few explicit links between ecosystem function and ecosite class at this scale of measurement, highlighting the potential importance of non-steady-state relationships between vegetation species and soil properties in disturbed forests and the potential need for finer-scale characterization to capture patterns in ecosystem function.
2006 IEEE International Symposium on Geoscience and Remote Sensing, 2006
Abstract-Variability in the gross primary product (GPP) algorithm derived from the Moderate Resol... more Abstract-Variability in the gross primary product (GPP) algorithm derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) is examined and compared with the same algorithm adjusted according to vegetation height influences on the light use efficiency (LUE) term within a section of the White Gull watershed, Saskatchewan Canada. GPP comparisons are made for pixels containing vegetation heights of 0 to 1m, 1 to 3m, 8 to 12m, and 12 to 24m. The results of this study find that the LUE term has an important influence ...
Vegetation in the Arctic is often sparse, spatially heterogeneous, and difficult to model. Synthe... more Vegetation in the Arctic is often sparse, spatially heterogeneous, and difficult to model. Synthetic Aperture Radar (SAR) has shown some promise in above-ground phytomass estimation at sub-arctic latitudes, but the utility of this type of data is not known in the context of the unique environments of the Canadian High Arctic. In this paper, Artificial Neural Networks (ANNs) were created to model the relationship between variables derived from high resolution multi-incidence angle RADARSAT-2 SAR data and optically-derived (GeoEye-1) Soil Adjusted Vegetation Index (SAVI) values. The modeled SAVI values (i.e., from SAR variables) were then used to create maps of above-ground phytomass across the study area. SAVI model results for individual ecological classes of polar semi-desert, mesic heath, wet sedge, and felsenmeer were reasonable, with r 2 values of 0.43, 0.43, 0.30, and 0.59, respectively. When the outputs of these models were combined to analyze the relationship between the model output and SAVI as a group, the r 2 value was 0.60, with an 8% normalized root mean square error (% of the total range of phytomass values), a positive indicator of a relationship. The above-ground phytomass model also resulted in a very strong relationship (r 2 = 0.87) between SAR-modeled and field-measured phytomass. A positive relationship was also found between optically derived SAVI values and field measured phytomass (r 2 = 0.79). These relationships demonstrate the utility of SAR data, compared to using optical data alone, for modeling OPEN ACCESS Remote Sens. 2014, 6 2135 above-ground phytomass in a high arctic environment possessing relatively low levels of vegetation.
Des methodes operetionnelles destinees aemeliorer la precision des resultets de classification de... more Des methodes operetionnelles destinees aemeliorer la precision des resultets de classification des cultures sont eveluees al'eide de donnees C-HH et C-HV recueillies Ie 70 juillet 7990 apartir d'un radar aantenne synthetique (RAS) eeroporte sur un site temoin situe dans Ie comte d'Oxiord, dans Ie sud de l'Onterio. Les donnees RAS brutes, les donnees tiltrees ainsi que des statistiques sur la texture des images ont ete clessitiees at'eide de methodes de classification utilisant i'une, les pixels et l'eutre, les champs dans Ie but de determiner leur eiticecite pour la classification des cultures. Les statistiques sur la texture qui sont derivees des densites de probebilite de la matrice de cooccurrence de niveaux de gris offrent, dans Ie cas des images RAS, un pouvoir de discrimination superieur acelui que presentent les statistiques des donnees brutes. La classification des champs permet, toutefois, d'obtenir des resultets nettement plus precis pour une foule de donnees d'entree.
Over the past two decades there has been an abundance of research demonstrating the utility of ai... more Over the past two decades there has been an abundance of research demonstrating the utility of airborne light detection and ranging (LiDAR) for predicting forest biophysical/inventory variables at the plot and stand levels. However, to date there has been little effort to develop a set of protocols for data acquisition and processing that would move governments or the forest industry towards cost-effective implementation of this technology for strategic and tactical (i.e., operational) forest resource inventories. The goal of this paper is to initiate this process by examining the significance of LiDAR data acquisition (i.e., point density) for modeling forest inventory variables for the range of species and stand conditions representing much of Ontario, Canada. Field data for approximately 200 plots, sampling a broad range of forest types and conditions across Ontario, were collected for three study sites. Airborne LiDAR data, characterized by a mean density of 3.2 pulses m −2 were systematically decimated to produce additional datasets with densities of approximately 1.6 and 0.5 pulses m −2 . Stepwise regression models, incorporating LiDAR height and density metrics, were developed for each of the three LiDAR datasets across a range of forest types OPEN ACCESS Remote Sens. 2012, 4 831 to estimate the following forest inventory variables: (1) average height (R 2 (adj) = 0.75-0.95);
Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a... more Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a major concern for arctic communities and resource development as warming temperatures have led to increasing permafrost degradation. In order to effectively assess and mitigate permafrost disturbance risk, disturbance-prone areas can be predicted through the use of susceptibility modelling. Permafrost disturbance susceptibility modelling is a quantitative assessment of the relationship between the distribution of past permafrost disturbances and a set of influencing terrain attributes that may cause slope instability. In this study we develop a universal permafrost disturbance susceptibility model using a limited disturbance inventory to test the applicability of the model for a broader region in the Canadian High Arctic. Additionally, we use the model to explore the effect of influencing terrain parameters on disturbance occurrence. To account for a large range of landscape characteris...
Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a... more Permafrost slope disturbance such as active layer detachments and retrogressive thaw slumps are a major concern for arctic communities and resource development as warming temperatures have led to increasing permafrost degradation. In order to effectively assess and mitigate permafrost disturbance risk, disturbance-prone areas can be predicted through the use of susceptibility modelling. Permafrost disturbance susceptibility modelling is a quantitative assessment of the relationship between the distribution of past permafrost disturbances and a set of influencing terrain attributes that may cause slope instability. In this study we develop a universal permafrost disturbance susceptibility model using a limited disturbance inventory to test the applicability of the model for a broader region in the Canadian High Arctic. Additionally, we use the model to explore the effect of influencing terrain parameters on disturbance occurrence. To account for a large range of landscape characteris...
Detailed forest ecosystem classifications have been developed for large regions of northern Ontar... more Detailed forest ecosystem classifications have been developed for large regions of northern Ontario. These ecosystem classifications provide tools for ecosystem management that constitute part of a larger goal of integrated management of forest ecosystems for long-term sustainability. These classification systems provide detailed stand-level characterization of forest ecosystems at a local level. However, for ecological approaches to forest management to become widely accepted by forest managers, and these tools to be widely used, methods must be developed to characterize and map or model ecosystem classes at landscape scales for large regions. In this study, the site-specific Northwestern Ontario Forest Ecosystem Classification (NWO FEC) was adapted to provide a landscape-scale (1:20000) forest ecosystem classification for the Rinker Lake Study Area located in the Boreal Forest north of Thunder Bay, Ontario. Multispatial resolution remote sensing data were collected using the Compact Airborne Spectrographic Imager (CASI) and analysed using geostatistical techniques to obtain an understanding of the nature of the spatial dependence of spectral reflectance for selected forest ecosystems at high spatial resolutions. Based on these analyses it was determined that an optimal size of support for characterizing forest ecosystems (i.e., optimal spatial resolution), as estimated by the mean ranges of a series of experimental semivariograms, differed based on (i) wavelength; (ii) forest ecosystem class (and at low altitude as a function of mean maximum canopy diameter (MMCD)); and (iii) altitude of the remote sensing system. In addition, maximum semivariance as estimated from the sills of the experimental semivariograms increased with density of understory. Based on the estimates for optimal spatial resolutions for six landscape-scale forest ecosystem classes, a series of spectral-spatial features were derived from the high-altitude CASI data (4 metre spatial resolution) using spatial averaging. Linear discriminant analysis for various spectral-spatial and texture feature combinations indicated that a spatial resolution of approximately 6 m was optimal for discriminating the six-landscape scale ecosystem classes. Texture features, using second-order spatial statistics that were derived from the 4 m remote sensing data, also significantly improved discrimination of the classes over the original 4 m data. Finally, addition of terrain descriptors, particularly elevation within a local region, improved discrimination of the six landscape scale ecosystem classes. It has been demonstrated that in a low-relief boreal environment, addition of textural and geomorphometric variables to high-resolution CASI reflectance data provides improved discrimination of forest ecosystem classes. Although these improvements are statistically significant, the absolute classification accuracies are not at levels suitable for operational classification and mapping. The analysis presented here represents the initiation of a complex modelling approach that is necessary for improving forest ecosystem characterization and prediction using additional primary datasets and derived datasets that possess various levels of measurement. Not only are optimal or multispatial resolution remote sensing data required, but also appropriately scaled terrain and landscape features depicting soil texture, nutrient and moisture regimes. Incorporation of these types of terrain-specific variables with reflectance data should provide further improvement in forest ecosystem classification and modelling at landscape scales.
Carbon dioxide, water vapour, and energy fluxes vary spatially and temporally within forested env... more Carbon dioxide, water vapour, and energy fluxes vary spatially and temporally within forested environments. However, it is not clear to what extent they vary as a result of variability in the spatial distribution of biomass and elevation. The following study presents a new methodology for extracting changes in the structural characteristics of vegetation and elevation within footprint areas, for direct comparison with eddy covariance (EC) CO2 flux concentrations. The purpose was to determine whether within-site canopy structure and local elevation influenced CO2 fluxes in a mature jack pine (Pinus banksiana Lamb.) forest located in Saskatchewan, Canada. Airborne light detection and ranging (lidar) was used to extract tree height, canopy depth, foliage cover, and elevation within 30 min flux footprints. Within-footprint mean structural components and elevation were related to 30 min mean net ecosystem productivity (NEP) and gross ecosystem production (GEP). NEP and GEP were modeled using multiple regression, and when compared with measured fluxes, almost all periods showed improvements in the prediction of flux concentration when canopy structure and elevation were included. Increased biomass was related to increased NEP and GEP in June and August when the ecosystem was not limited by soil moisture. On a daily basis, fractional cover and elevation had varying but significant influences on CO 2 fluxes.
This study investigates the potential of discrete return light detection and ranging (lidar) data... more This study investigates the potential of discrete return light detection and ranging (lidar) data to characterize forest succession in a mixed mature forest in central Ontario using indices applied to the lidar point cloud. Derived indices include statistical indices, predicted Lorey's height (R 2 ϭ 0.86; RSME ϭ 2.36 m) and quadratic mean diameterat-breast-height (R 2 ϭ 0.68; RMSE ϭ 1.21 cm), canopy density indices and an information theory based complexity index. To assess how well these indices are able to capture the vertical structure of forest stands, they are compared to four stages of forest stand development. Best subsets regressions indicated that no single index is able to separate all four stages adequately. However, the predicted Lorey's height index is optimal for separating early from mid succession stages (p Ͻ.0001) and a combination of height and complexity indices performed best to discriminate between mid-and late-succession stages (p Ͻ.0001).
Literature from the past two decades documents how airborne LIDAR can be used to predict forest i... more Literature from the past two decades documents how airborne LIDAR can be used to predict forest inventory variables, such as basal area, volume, and biomass, at the plot and stand level. However, a key question that has yet to be fully addressed, and that the forest industry continues to ask as it considers operationalizing the use of LIDAR in forest resource inventories, is: What is the optimal point density for predicting forest inventory variables? For example, is a point density of 0.5 points/m 2 sufficient for making accurate predictions of forest inventory variables or is a point density of 3 points/m 2 the minimum required? To investigate this key question, a three-year project was launched in 2007 in Ontario, Canada. Field data for approximately 180 plots, sampling a broad range of forest types and conditions in Ontario, were collected over two study sites. Airborne LIDAR data, characterized by 3 points/m 2 were systematically decimated to produce datasets with point densities of approximately 1.5 and 0.5 points/m 2 . Models, developed using stepwise regression, were developed for each of the three lidar datasets to estimate several forest inventory variables including: (1) basal area (R 2 =0.25-0.94); (2) gross total volume (R 2 =0.46-0.95); (3) gross merchantable volume (R 2 =0.37-0.94); (4) stem density(R 2 =0.23-0.89); (5) quadratic mean diameter (R 2 =0.59-0.86); (6) total aboveground biomass (R 2 =0.26-0.93); (7) average height (R 2 =0.76-0.95); and (8) top height (R 2 =0.75-0.98). Aside from two cases, no decimation effect was found for predictions of forest variables, which suggests that a point density of 0.5 points/m 2 is sufficient for plot and stand level modelling under these forest conditions.
Sets of spectral, spectral-spatial, textural, and geomorphometric variables derived fram high spa... more Sets of spectral, spectral-spatial, textural, and geomorphometric variables derived fram high spatial resolution Compact Airborne Spectrographic Imager (CASI) and elevation data are tested to determine their ability to discriminate landscape-scale forest ecosystem classes for a study area in northern Ontario, Canada. First, linear discriminant analysis for various spectral and spectral-spatial variables indicated that a spatial resolution of approximately 6 m was optimal for discriminating six landscape-scale forest ecosystem classes. Second, texture features, using second-order spatial statistics, significantly improved discrimination of the classes over the original reflectance data. Finally, addition of terrain descriptors improved discrimination of the six forest ecosystem classes. It has been demonstrated that, in a low-to moderate-relief boreal environment, addition of textural and terrain variables to highresolution CASI reflectance data provides improved discrimination of forest ecosystem classes. P. Treitz was with the 1 J 0.591 2 J J 0.877 3 J J J 0.987 4 J J J J 1.081 5 J J J J J 1.136 6 J J J J J J 1.165 CASI Spectral Variable Spatial Resolution 1 2 3 4 5 6 CASI (580-601 nm) 4 meters 5 meters J J 6 meters J J CASI (663-678 nm) 4 meters J J 5 meters J J 6 meters J J CASI (744-750 nm) 5 meters 6 meters J J 7 meters J J J J *1. Aspen-Dominated Hardwood and Mixedwood; 2. White Spruce1 Balsam Fir Conifer and Mixedwood; 3. Cedar Mixedwood; 4. Upland Black SpruceIJack Pine; 5. Lowland Black Spruce; and 6. Wetland Black Spruce
The Kyoto Protocol has sharpened the focus on the possible role of forests in contributing to or ... more The Kyoto Protocol has sharpened the focus on the possible role of forests in contributing to or mitigating climate change. The understanding of carbon dynamics, and the prediction of future carbon stock changes, rely on an analysis of the past forest dynamics. Historical data may often be incomplete, approximate, or strongly generalized. To circumvent this problem, we have recently developed
IEEE International IEEE International IEEE International Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004, 2004
Abstract Interferometric synthetic aperture radar (InSAR) data from RADARSAT-1 have been examined... more Abstract Interferometric synthetic aperture radar (InSAR) data from RADARSAT-1 have been examined to assess their potential for mapping terrain and changes in snow cover characteristics, relative to the limiting effects of snow on foraging by endangered Peary caribou (Rangifer tarandus). Radar is one of the few observational tools that can provide information on the changing snow pack during the dark winter months. The goal of this research is to characterize the general ensemble of terrain characteristics that may affect ...
Thermokarst, active layer detachments (ALDs), and retrogressive thaw slumps (RTSs), representing ... more Thermokarst, active layer detachments (ALDs), and retrogressive thaw slumps (RTSs), representing three forms of permafrost degradation, constitute serious risks for infrastructure and have the potential to alter environmental and ecological conditions in Arctic regions. Environmental change and increased land development pressures require innovative cost-effective methods for assessing hazard prone areas. The overall research objective of this project is to design a landscape model to predict and characterize disturbance prone areas using key physiographic controls and geospatial modeling derived from satellite imagery to efficiently produce hazard susceptibility maps. Susceptibility maps identify disturbance prone areas, which are a fundamental component of hazard management and the basis for provision of measures aimed at reducing the risks resulting from permafrost degradation. To test the validity of this modelling approach and its applicability across the Arctic, methods are be...
This study examines the distribution of laser pulse returns obtained from coincident airborne and... more This study examines the distribution of laser pulse returns obtained from coincident airborne and terrestrial lidar surveys of a closed-canopy red pine (Pinus resinosa) plantation. The purpose of this study is to improve our understanding of laser pulse sampling within closed canopies so that estimates of forest structural variables (e.g., biomass, needle-leaf area, and base-of-live-crown) can be improved at the individual tree and stand levels using lidar. The results of this study indicate the following: (1) There is a statistically significant difference between field measurements of tree height and estimates derived from the maximum laser pulse return from airborne and terrestrial lidar. In both cases, maximum laser pulse returns underestimate tree height by 1 m, on average. (2) Both terrestrial and airborne lidar are unable to discern the base of the measured live crown. Laser pulse returns from airborne lidar are biased towards the top of the tree crown, i.e., lowest laser pul...
The ecosite unit in Ontario's boreal forest ecological land classification system is a polygon of... more The ecosite unit in Ontario's boreal forest ecological land classification system is a polygon of common vegetation type and soil conditions intended to provide a standardized provincial framework to inform meso-scale forestry and planning applications. To determine whether the physical factors used for ecosite classification relate to patterns in ecological function over finer spatial scales, we examined 14 soil properties in replicate boreal forest plots representing eight mineral soil ecosite classes and three organic soil ecosite classes in the Hearst Forest. Despite large differences in vegetation composition, we found few statistically significant differences in properties when compared for individual classes or for more general groupings based on vegetation type and soil texture or expected fertility status. However, some properties (soil organic carbon, total nitrogen, and C:N ratio) were approaching significance in the 0-10 cm depth increment, and there were distinct differences between organic soil and mineral soil sites. Overall, these results suggest few explicit links between ecosystem function and ecosite class at this scale of measurement, highlighting the potential importance of non-steady-state relationships between vegetation species and soil properties in disturbed forests and the potential need for finer-scale characterization to capture patterns in ecosystem function.
2006 IEEE International Symposium on Geoscience and Remote Sensing, 2006
Abstract-Variability in the gross primary product (GPP) algorithm derived from the Moderate Resol... more Abstract-Variability in the gross primary product (GPP) algorithm derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) is examined and compared with the same algorithm adjusted according to vegetation height influences on the light use efficiency (LUE) term within a section of the White Gull watershed, Saskatchewan Canada. GPP comparisons are made for pixels containing vegetation heights of 0 to 1m, 1 to 3m, 8 to 12m, and 12 to 24m. The results of this study find that the LUE term has an important influence ...
Vegetation in the Arctic is often sparse, spatially heterogeneous, and difficult to model. Synthe... more Vegetation in the Arctic is often sparse, spatially heterogeneous, and difficult to model. Synthetic Aperture Radar (SAR) has shown some promise in above-ground phytomass estimation at sub-arctic latitudes, but the utility of this type of data is not known in the context of the unique environments of the Canadian High Arctic. In this paper, Artificial Neural Networks (ANNs) were created to model the relationship between variables derived from high resolution multi-incidence angle RADARSAT-2 SAR data and optically-derived (GeoEye-1) Soil Adjusted Vegetation Index (SAVI) values. The modeled SAVI values (i.e., from SAR variables) were then used to create maps of above-ground phytomass across the study area. SAVI model results for individual ecological classes of polar semi-desert, mesic heath, wet sedge, and felsenmeer were reasonable, with r 2 values of 0.43, 0.43, 0.30, and 0.59, respectively. When the outputs of these models were combined to analyze the relationship between the model output and SAVI as a group, the r 2 value was 0.60, with an 8% normalized root mean square error (% of the total range of phytomass values), a positive indicator of a relationship. The above-ground phytomass model also resulted in a very strong relationship (r 2 = 0.87) between SAR-modeled and field-measured phytomass. A positive relationship was also found between optically derived SAVI values and field measured phytomass (r 2 = 0.79). These relationships demonstrate the utility of SAR data, compared to using optical data alone, for modeling OPEN ACCESS Remote Sens. 2014, 6 2135 above-ground phytomass in a high arctic environment possessing relatively low levels of vegetation.
Des methodes operetionnelles destinees aemeliorer la precision des resultets de classification de... more Des methodes operetionnelles destinees aemeliorer la precision des resultets de classification des cultures sont eveluees al'eide de donnees C-HH et C-HV recueillies Ie 70 juillet 7990 apartir d'un radar aantenne synthetique (RAS) eeroporte sur un site temoin situe dans Ie comte d'Oxiord, dans Ie sud de l'Onterio. Les donnees RAS brutes, les donnees tiltrees ainsi que des statistiques sur la texture des images ont ete clessitiees at'eide de methodes de classification utilisant i'une, les pixels et l'eutre, les champs dans Ie but de determiner leur eiticecite pour la classification des cultures. Les statistiques sur la texture qui sont derivees des densites de probebilite de la matrice de cooccurrence de niveaux de gris offrent, dans Ie cas des images RAS, un pouvoir de discrimination superieur acelui que presentent les statistiques des donnees brutes. La classification des champs permet, toutefois, d'obtenir des resultets nettement plus precis pour une foule de donnees d'entree.
Over the past two decades there has been an abundance of research demonstrating the utility of ai... more Over the past two decades there has been an abundance of research demonstrating the utility of airborne light detection and ranging (LiDAR) for predicting forest biophysical/inventory variables at the plot and stand levels. However, to date there has been little effort to develop a set of protocols for data acquisition and processing that would move governments or the forest industry towards cost-effective implementation of this technology for strategic and tactical (i.e., operational) forest resource inventories. The goal of this paper is to initiate this process by examining the significance of LiDAR data acquisition (i.e., point density) for modeling forest inventory variables for the range of species and stand conditions representing much of Ontario, Canada. Field data for approximately 200 plots, sampling a broad range of forest types and conditions across Ontario, were collected for three study sites. Airborne LiDAR data, characterized by a mean density of 3.2 pulses m −2 were systematically decimated to produce additional datasets with densities of approximately 1.6 and 0.5 pulses m −2 . Stepwise regression models, incorporating LiDAR height and density metrics, were developed for each of the three LiDAR datasets across a range of forest types OPEN ACCESS Remote Sens. 2012, 4 831 to estimate the following forest inventory variables: (1) average height (R 2 (adj) = 0.75-0.95);
Uploads
Papers by Paul Treitz