Search

The maintenance and restoration of wetland habitat is a priority conservation action for most waterfowl and other wetland-dependent species in North America. Despite much progress in targeting habitat management in staging and wintering areas, methods to identify and target high-quality breeding habitats that result in the greatest potential for wildlife are still required. This is particularly true for species that breed in remote, inaccessible areas such as the American black duck, an intensively managed game bird in Eastern North America. Although evidence suggests that black ducks prefer productive, nutrient-rich waterbodies, such as beaver ponds, information about the distribution and quality of these habitats across the vast boreal forest is lacking with accurate identification remaining a challenge. Continuing advancements in remote sensing technologies that provide spatially extensive and temporally repeated information are particularly useful in meeting this information gap. In this study, we used multi-source remotely sensed information and a fuzzy analytical hierarchy process to map the spatial distribution of beaver ponds in Ontario. The use of multi-source data, including a Digital Elevation Model, a Sentinel-2 Multi-Spectral Image, and RadarSat 2 Polarimetric data, enabled us to identify individual beaver ponds on the landscape. Our model correctly identified an average of 83.0% of the known beaver dams and 72.5% of the known beaver ponds based on validation with an independent dataset. This study demonstrates that remote sensing is an effective approach for identifying beaver-modified wetland features and can be applied to map these and other wetland habitat features of interest across large spatial extents. Furthermore, the systematic acquisition strategy of the remote sensors employed is well suited for monitoring changes in wetland conditions that affect the availability of habitats important to waterfowl and other wildlife.

Terrain displacement due to the seasonal thaw of the active layer above permafrost can be sensitive to climate change; however, its accurate characterization remains a challenge. This study aimed to improve the measurement of the subsidence or vertical ground surface displacement using differential synthetic aperture radar interferometry (InSAR). Existing methods for reliable phase unwrapping are hindered by the decorrelation between time-series of SAR acquisitions that can result due to the heterogeneity and structural sensitivity of permafrost landscapes to external conditions. In this study, an advanced phase unwrapping method was proposed, in which three types of regions, namely non-residue, sparse residue, and dense residue objects, were obtained from wrapped interferogram and residue map using a segmentation method. Two variants of Polynomial-Based Region Growing Phase Unwrapping (PBRGPU) were developed, which are sparse-residue Object-based PBRGPU(SOP) and dense-residue Object-based PBRGPU(DOP). The results demonstrated that the proposed method outperformed the existing phase unwrapping methods by partially suppressing the decorrelation phase and enhancing robustness for complex terrain deformation in the absence of measured field data. Both the PBRGPU variants and segmentation strategies compose the object-based unwrapping method for permafrost, and also provide a new framework by combining the segmentations and scenarios for phase unwrapping for permafrost regions.

This study aims to exploit multisource remotely sensed data to improve land cover classification of an area dominated by extensive wetlands with surface cover complexity strongly shaped by permafrost, fine-scale geomorphological and topographic characteristics. Accurate and precise mapping tools are critically needed to track change in wetland ecosystems under climate change. Wetland classification is challenging due to its high intraclass and low interclass variations in remote sensing features. Furthermore, the distribution and seasonal thaw patterns in permafrost influence vegetation cover at broad and fine scales within the study area of Hudson Bay Lowland, Canada, creating further complexity in classification. A systematic analysis was performed to evaluate the contribution of various remote sensing features (e.g., spectral, temporal, and structural features) and to determine vital datasets for effective monitoring of subarctic wetland-dominated ecosystems. Prediction uncertainty was comprehensively studied and reported together with classification accuracy. A decision-level fusion method based on the Dempster–Shaffer (DS) theory was developed. Different classes (i.e., focal elements under DS theory) were considered in this classification, using Sentinel-1 and Sentinel-2 data separately. The overall accuracy of the classification of 13 wetland classes was 0.952, significantly improved compared with the accuracies obtained by using individual data sources and by using feature-based fusion. Furthermore, the percentage of uncertain pixels is reduced as well.

With the recent thrust to convert forests in Ontario’s Clay Belt to agricultural land, a vital need arises to assess the attendant effects on carbon and greenhouse gas (GHG) emissions. This paper examines the possible effect of land conversion on soil organic carbon and GHG emissions within a study area in Northern Ontario, Canada, during the next two decades under different land management schemes. The study established a framework to conduct simulations with the DNDC model for agricultural lands and the CBM for forested areas. The methodology involves a unique change detection method for models’ land cover and disturbance inputs. The work highlights the improvement in carbon simulation accuracy from better inputs to carbon models. Furthermore, it addresses modalities to ensure fewer uncertainties are introduced while merging data from multiple geospatial data sources. The simulations demonstrated that the carbon sequestration potential in the forests was almost double the soil organic carbon accumulation in the agricultural lands. Validations done for the estimation of carbon sequestered included comparisons of the carbon model outputs from field survey data from 2018–2021. In most sites, the carbon amounts from the computer models compared to those from the field survey, within limits of error. The average uncertainties in GHG emissions ranged from ~0.5% to 12.8%.

Variations within local topography can effectively impact the location of tree species within naturally forested areas. Furthermore, the uncertainty of prediction for classification can vastly differ amongst topography and the overlying tree species groupings. This study investigated the supplementation of a suite of topographic features corresponding to morphometry and hydrological considerations, in addition to multispectral imagery and other LiDAR-derived features, at fine (2 m) spatial resolution for a pixel-based tree species classification of a forested region of the boreal biome in northern Ontario, Canada. The study area conforms to the Abitibi River Forest (ARF) and consists of the tree species of black spruce (Picea mariana), balsam fir (Abies balsamea), trembling aspen (Populus tremuloides), balsam poplar (Populus balsamifera), tamarack (Larix laricina), white spruce (Picea glauca), and eastern white cedar (Thuja occidentalis). Random forest (RF) and support vector machines (SVMs) were implemented for the classification. Topographic features, specifically those conforming to channel base level, valley depth, and multi-resolution valley bottom flatness (MRVBF), were among the most important features for species predictors. The RF and SVM methods were trained on pixels of pure stands (composed of 70%+ of same tree species) for the tree species groupings, which were split by site level. Modelling accuracies for both the pixel and site level were reported, with the best model attaining an overall site level accuracy and corresponding Cohen’s kappa score of 0.79 and 0.69 for classification, respectively. Entropy maps were generated to characterize the uncertainty of prediction, and substantiate that the regions of lowest uncertainty correspond to wetlands, which are dominated by black spruce (Picea mariana). A modified entropy map was calculated from the normalized top two probabilities of tree species groupings predicted per pixel, so as to better highlight regions of prediction uncertainty. A prediction map for the second most-likely tree species groupings was also computed, which supports the presence of balsam fir (Abies balsamea) as a secondary tree species throughout the ARF region.

Methodologies were developed to classify wetlands (Open Bog, Treed Bog, Open Fen, Treed Fen, and Swamps) from remotely sensed data using advanced classification algorithms through two hierarchical approaches. The data utilized included multispectral optical and thermal data (Landsat-5, and Landsat-8), radar imagery (Sentinel-1), and a digital elevation model. Goals were to determine the best way to combine imagery to classify wetlands through hierarchically based classification approaches to produce more accurate and efficient maps compared to standard classification. Algorithms used were Random Forest (RF), and Naïve Bayes. A hierarchically based RF classification methodology produced the most accurate classification result (91.94%). The hierarchically based approaches also improved classification accuracies for low-quality data, as defined through feature analysis, when compared to a nonhierarchical classifier. The hierarchical approaches also produced a significant increase in classification accuracy for the Naïve Bayes classifier versus the standard approach (∼12% increase) while not significantly increasing computation time – comparable in accuracy to the RF tests for around 20% the computational effort. Preselection of spectral bands, polarizations and other input parameters (Normalized Difference Vegetation Index, Normalized Difference Water Index, albedo, slope, etc.) using log-normal or RF variable importance analysis was very effective at identifying low-quality features and features which were of higher quality.

With the wide availability of remotely sensed data from various sensors, fusion-based tree species classification approaches have emerged as a prominent and ongoing research topic. However, most recent studies primarily focused on combining multisource data at the feature level, while few systematically examined their positive or negative contributions to tree species classification. This study aimed to investigate fusion approaches at the feature and decision levels deployed with support vector machine and random forest algorithms to classify five dominant tree species: Norway maple, honey locust, Austrian pine, white spruce, and blue spruce in individual crowns. Spectral, textural, and structural features derived from multispectral imagery (MSI), a very high-resolution panchromatic image (PAN), and LiDAR data were systematically exploited to assess their contributions to accurate classifications. Among the various classification schemes that were explored, both feature- and decision-level fusion approaches demonstrated significant improvements in tree species classification compared with the utilization of MSI (0.7), PAN (0.74), or LiDAR (0.8) in isolation. Notably, the decision-level fusion approach achieved the highest overall accuracies (0.86 for SVM and 0.84 for RF) and kappa coefficients (0.82 for SVM and 0.79 for RF). The misclassification analysis of fusion approaches highlighted the potential and flexibility of decision-level fusion in tree species classification.

In this study, an object-oriented, decision-level fusion method is proposed for tree species classification based on spectral, textural, and structural features derived from multi-spectral and panchromatic imagery and Light Detection And Ranging (LiDAR) data. Murphy's average method based on the Dempster Shafer theory (DST) was used to calculate the combined mass function for decision making purposes. For individual feature groups, the mass functions were calculated using the support vector machine (SVM) classification method. The species examined included Norway maple, honey locust, Austrian pine, blue spruce, and white spruce. In addition to these species, a two- or three-species compound class was included in the decision process based on the normalized entropy in the presence of conflict that was itself determined according to whether individual groups of features were consistent. The developed method provided a mechanism to identify tree crowns, which could not be classified to one single species with a high confidence due to the conflict among feature groups. Data used in this study were obtained for the Keele Campus of York University, Toronto, Ontario. Among the 223 test crowns, 204 crowns were assigned to one single species, and the overall classification accuracy was 0.89. A decision could not be made for 19 crowns with confidence, and as a result, a two- or three-species compound class was assigned. The classification accuracy was higher than that obtained using SVM classification based on individual and combined spectral, structural, and textural features.

The goal of this research was to improve wetland classification by fully exploiting multi-source remotely sensed data. Three distinct classifiers were designed to distinguish individual or compound wetland categories using random forest (RF) classification. They were determined, in part, to best use the available remotely sensed features in order to maximize that information and to maximize classification accuracy. The results from these classifiers were integrated according to Dempster–Shafer theory (D–S theory). The developed method was tested on data collected from a study area in Northern Alberta, Canada. The data utilized were Landsat-8 and Sentinel-2 (multi-spectral), Sentinel-1 (synthetic aperture radar—SAR), and digital elevation model (DEM). Classification of fen, bog, marsh, swamps, and upland resulted in an overall accuracy of 0.93 using the proposed methodology, an improvement of 5% when compared to a traditional classification method based on the aggregated features from these data sources. It was noted that, with the traditional method, some pixels were misclassified with a high level of confidence (>85%). Such misclassification was significantly reduced (by ~10%) by the proposed method. Results also showed that some features important in separating compound wetland classes were not considered important using the traditional method based on the RF feature selection mechanism. When used in the proposed method, these features increased the classification accuracy, which demonstrated that the proposed method provided an effective means to fully employ available data to improve wetland classification.

Deep learning (DL) has shown promising performances in various remote sensing applications as a powerful tool. To explore the great potential of DL in improving the accuracy of individual tree species (ITS) classification, four convolutional neural network models (ResNet-18, ResNet-34, ResNet-50, and DenseNet-40) were employed to classify four tree species using the combined high-resolution satellite imagery and airborne LiDAR data. A total of 1503 samples of four tree species, including maple, pine, locust, and spruce, were used in the experiments. When both WorldView-2 and airborne LiDAR data were used, the overall accuracies (OA) obtained by ResNet-18, ResNet-34, ResNet-50, and DenseNet-40 were 90.9%, 89.1%, 89.1%, and 86.9%, respectively. The OA of ResNet-18 was increased by 4.0% and 1.8% compared with random forest (86.7%) and support vector machine (89.1%), respectively. The experimental results demonstrated that the size of input images impacted on the classification accuracy of ResNet-18. It is suggested that the input size of ResNet models can be determined according to the maximum size of all tree crown sample images. The use of LiDAR intensity image was helpful in improving the accuracies of ITS classification and atmospheric correction is unnecessary when both pansharpened WorldView-2 images and airborne LiDAR data were used.

Mapping and understanding the differences in land cover and land use over time is an essential component of decision-making in sectors such as resource management, urban planning, and forest fire management, as well as in tracking of the impacts of climate change. Existing methods sometimes pose a barrier to the effective monitoring of changes in land cover and land use, since a threshold parameter is often needed and determined based on trial and error. This study aimed to develop an automatic and operational method for change detection on a large scale from Moderate Resolution Imaging Spectroradiometer (MODIS) data. Super pixels were the basic unit of analysis instead of traditional individual pixels. T2 tests based on the feature vectors of temporal Normalized Difference Vegetation Index (NDVI) and land surface temperature were used for change detection. The developed method was applied to data over a predominantly vegetated area in northern Ontario, Canada spanning 120,000 sq. km from 2001–2016. The accuracies ranged between 78% and 88% for the NDVI-based test, from 74% to 86% for the LST-based test, and from 70% to 86% for the joint method compared with manual interpretation. Our proposed method for detecting land cover change provides a functional and viable alternative to existing methods of land cover change detection as it is reliable, repeatable, and free from uncertainty in establishing a threshold for change.

The Asian Emerald Ash Borer beetle (EAB, Agrilus planipennis Fairmaire) can cause damage to all species of Ash trees (Fraxinus), and rampant, unchecked infestations of this insect can cause significant damage to forests. It is thus critical to assess and model the spread of the EAB in a manner that allows authorities to anticipate likely areas of future tree infestation. In this study, a generalized linear mixed model (GLMM), combining the features of the commonly used generalized linear model (GLM) and a random effects model, was developed to predict future EAB spread patterns in Southern Ontario, Canada. The GLMM was designed to deal with autocorrelation in the data. Two random effects were established based on the geographic information provided with the EAB data, and a method based on statistical inference was proposed to identify the most significant factors associated with the distribution of the EAB. The results of the model showed that 95% of the testing data were correctly classified. The predictive performance of the GLMM was substantially enhanced in comparison with that obtained by the GLM. The influence of climatic factors, such as wind speed and anthropogenic activities, had the most significant influence on the spread of the EAB.

Mapping the distribution and persistence of surface water in a timely fashion has broad value for tracking dynamic events like flooding, and for monitoring the effects of climate and human activities on natural resource values and biodiversity. Traditionally, surface water is mapped from optical imagery using semi-automatic approaches. However, this process is time-consuming and the accuracy of results can vary among image interpreters. In recent years, Synthetic Aperture Radar (SAR) images have been increasingly used. Microwave signals sensitive to water content make SAR systems useful for mapping surface water, saturated soils, and flooded vegetation. In this study, a fully automatic method based on robust stepwise thresholding was developed to map and track the change in the extent of surface water using Polarimetric SAR data. The application of this method in both Radarsat-2 and Sentinel-1 data in central Ontario, Canada demonstrates that the developed robust stepwise thresholding approach could facilitate rapid mapping of open water areas with a promising accuracy of over 95%. In addition, the time-series extent of surface water extracted from May 2008 to August 2016 reveals the dynamic nature of surface inundation, and the trend was consistent with the local precipitation data.

In current individual tree crown (ITC) delineation methods for high-resolution multispectral imagery, either a spectral band or a brightness component of the multispectral image is employed in delineation with reference to edges or shapes of crowns, whereas spectra of tree crowns are seldom taken into account. Such methods normally perform well in coniferous forests with obvious between-crown shadows, but fail in dense deciduous or mixed forests, in which tree crowns are close to each other, between-crown shadows and boundaries are unobvious, whereas adjacent tree crowns may be of distinguishable spectra. In order to effectively delineate crowns in dense deciduous or mixed forests, a new ITC delineation method using both brightness and spectra of the image is proposed in this study. In this method, a morphological gradient map of the image is first generated, treetops of multi-scale crowns are extracted from the gradient map and refined regarding the spectral differences between neighboring crowns, the gradient map is segmented using a watershed approach with treetops as markers, and the resulting segmentation map is refined to yield a crown map. Evaluated on images of a rainforest and a deciduous forest, the proposed method more accurately delineated adjacent broad-leaved tree crowns with similar brightness but different spectra than the other two typical ITC delineation algorithms, achieving a delineation accuracy of up to 76% in the rainforest and 63% in the deciduous forest.

A critical step in object-oriented geospatial analysis (OBIA) is image segmentation. Segments determined from a lower-spatial resolution image can be used as the context to analyse a corresponding image at a higher-spatial resolution. Due to inherent differences in perceptions of a scene at different spatial resolutions and co-registration, segment boundaries from the low spatial resolution image need to be adjusted before being applied to the high-spatial resolution image. This is a non-trivial task due to considerations such as noise, image complexity, and determining appropriate boundaries, etc. An innovative method was developed in the study to solve this. Adjustments were executed for each boundary pixel based on the minimization of an energy function characterizing local homogeneity. It executed adjustments based on a structure which rewarded movement towards edges, and superior changes towards homogeneity. The developed method was tested on a set of Quickbird, ASTER and a lower resolution, resampled, Quickbird image, over a study area in Ontario, Canada. Results showed that the adjusted-segment boundaries obtained from the lower resolution imagery aligned well with the features in the Quickbird imagery.

Wetlands play a key role in regional and global environments, and are critically linked to major issues such as climate change, wildlife habitat, biodiversity, water quality protection, and global carbon and methane cycles. Remotely-sensed imagery provides a means to detect and monitor wetlands on large scales and with regular frequency. In this project, methodologies were developed to classify wetlands (Open Bog, Treed Bog, Open Fen, Treed Fen, and Swamps) from multi-source remotely sensed data using advanced classification algorithms. The data utilized included multispectral optical and thermal data (Landsat-5) and Radar imagery from RADARSAT-2 and Sentinel-1. The goals were to determine the best way to combine the aforementioned imagery to classify wetlands, and determine the most significant image features. Classification algorithms investigated in this study were Naive Bayes, K-Nearest Neighbor (K-NN), Support Vector Machine (SVM), and Random Forest (RF). Based on the test results in the study area in Northern Ontario, Canada (49°31′.34N, 80°43′37.04W), a RF based classification methodology produced the most accurate classification result (87.51%). SVM, in some cases, produced results of comparable or better accuracy than RF. Our work also showed that the use of surface temperature (an untraditional feature choice) could aid in the classification process if the image is from an abnormally warm spring. This study found that wetlands were best classified using the NDVI (Normalized Difference Vegetative Index) calculated from optical imagery obtained in the spring months, radar backscatter coefficients, surface temperature, and ancillary data such as surface slope, computed through either an RF or SVM classifier. It was also found that preselection of features using Log-normal or RF variable importance analysis was an effective way of identifying low quality features and to a lesser extent features which were of higher quality.

Despite its popularity, the development of an embedded real-time multisensor kinematic positioning and navigation system discourages many researchers and developers due to its complicated hardware environment setup and time consuming device driver development. To address these issues, this paper proposed a multisensor kinematic positioning and navigation system built on Linux with Real Time Application Interface (RTAI), which can be constructed in a fast and economical manner upon popular hardware platforms. The authors designed, developed, evaluated and validated the application of Linux/RTAI in the proposed system for the integration of the low cost MEMS IMU and OEM GPS sensors. The developed system with Linux/RTAI as the core of a direct geo-referencing system provides not only an excellent hard real-time performance but also the conveniences for sensor hardware integration and real-time software development. A software framework is proposed in this paper for a universal kinematic positioning and navigation system with loosely-coupled integration architecture. In addition, general strategies of sensor time synchronization in a multisensor system are also discussed. The success of the loosely-coupled GPS-aided inertial navigation Kalman filter is represented via post-processed solutions from road tests.

Individual tree species identification is important for urban forest inventory and ecology management. Recent advances in remote sensing technologies facilitate more detailed estimation of individual urban tree characteristics. This study presents an approach to improve the classification of individual tree species via longitudinal profiles from very high spatial resolution airborne imagery. The longitudinal profiles represent the side view tree shape, which play a very important role in individual tree species on-site identification. Decision tree classification was employed to conduct the final classification result. Using this profile approach, six major species (Maple, Ash, Birch, Oak, Spruce, Pine) of trees on the York University (Ontario, Canada) campus were successfully identified. Two decision trees were constructed, one knowledge-based and one derived from gain ratio criteria. The classification accuracy achieved were 84% and 86%, respectively.

Terrestrial LiDAR provides many disciplines with an effective and efficient means of producing realistic three-dimensional models of real world objects. With the advent of mobile terrestrial LiDAR, this ability has been expanded to include the rapid collection of three-dimensional models of large urban scenes. For all its usefulness, it does have drawbacks. One of the major problems faced by the LiDAR industry today is the automatic removal of outlying data points from LiDAR point clouds. This paper discusses the development and combined implementation of two methods of performing outlier detection in georeferenced point clouds. These methods made use of the raw data available from most time-of-flight mobile terrestrial LiDAR scanners in both the temporal and spatial domains. The first method involved a moving fixed interval smoother derived from the well-known position velocity acceleration Kalman Filter. The second method fitted a quadratic curved surface to sections of LiDAR data. The combined use of these routines is discussed through examples with real LiDAR data.

In this study, full-waveform LiDAR data were exploited to detect weak returns backscattered by the bare terrain underneath vegetation canopies and thus improve the generation of a digital terrain model (DTM). Building on the methods of progressive generation of triangulation irregular network (TIN) model reported in the literature, we proposed an integrated approach where echo detection, terrain identification, and TIN generation were carried out iteratively. The proposed method was tested on a dataset collected by a Riegl LMS Q-560 scanner over a study area near Sault Ste. Marie, Ontario, Canada (46°33′56′′N, 83°25′18′′W). The results demonstrated that more terrain points under shrubs could be identified, and the generated DTMs exhibited more details in the terrain than those obtained using the progressive TIN method. In addition, 1275 points across this study area were surveyed on the ground and used to validate the proposed approach. The estimated elevations were shown to have a strong linear relationship with the measured ones, with R2 values above 0.98, and the RMSEs (Root Mean Squared Errors) between them were less than 0.15 m even for areas with hilly terrains underneath vegetation canopies.

In this study, we demonstrate the potential of using high spatial resolution airborne imagery to characterize the structural development stages of forest canopies. Four forest succession stages were adopted: stand initiation, young multistory, understory reinitiation, and old growth. Remote sensing metrics describing the spatial patterns of forest structures were derived and a Random Forest learning algorithm was used to classify forest succession stages. These metrics included texture variables from Gray Level Co-occurrence Measures (GLCM), range and sill from the semi-variogram, and the fraction of shadow and its spatial distribution. Among all the derived variables, shadow fractions and the GLCM variables of contrast, mean, and dissimilarity were the most important for characterizing the forest succession stages (classification accuracy of 89%). In addition, a LiDAR (Light Detection and Ranging) derived forest structural index (predicted Lorey’s height) was employed to validate the classification result. The classification using imagery spatial variables was shown to be consistent with the LiDAR derived variable (R2 = 0.68 and Root Mean Square Error (RMSE) = 2.39). This study demonstrates that high spatial resolution imagery was able to characterize forest succession stages with promising accuracy and may be considered an alternative to LiDAR data for this kind of application. Also, the results of stand development stages build a framework for future wildlife habitat mapping.