Gregory J. McDermid

Transcending scale dependence in identifying habitat with resource selection functions

Multi-scale resource selection modeling is used to identify factors that limit species distributions across scales of space and time. This multi-scale nature of habitat suitability complicates the translation of inferences to single, spatial depictions of habitat required for conservation of species. We estimated resource selection functions (RSFs) across three scales for a threatened ungulate, woodland caribou (Rangifer tarandus caribou), with two objectives: (1) to infer the relative effects of two forms of anthropogenic disturbance (forestry and linear features) on woodland caribou distributions at multiple scales and (2) to estimate scale-integrated resource selection functions (SRSFs) that synthesize results across scales for management-oriented habitat suitability mapping. We found a previously undocumented scale-specific switch in woodland caribou response to two forms of anthropogenic disturbance. Caribou avoided forestry cut-blocks at broad scales according to first- and second-order RSFs and avoided linear features at fine scales according to third-order RSFs, corroborating predictions developed according to predator-mediated effects of each disturbance type. Additionally, a single SRSF validated as well as each of three single-scale RSFs when estimating habitat suitability across three different spatial scales of prediction. We demonstrate that a single SRSF can be applied to predict relative habitat suitability at both local and landscape scales in support of critical habitat identification and species recovery.

The influence of patch-delineation mismatches on multi-temporal landscape pattern analysis

Investigations of land-cover change often employ metrics designed to quantify changes in landscape structure through time, using analyses of land cover maps derived from the classification of remote sensing images from two or more time periods. Unfortunately, the validity of these landscape pattern analyses (LPA) can be compromised by the presence of spurious change, i.e., differences between map products caused by classification error rather than real changes on the ground. To reduce this problem, multi-temporal time series of land-cover maps can be constructed by updating (projecting forward in time) and backdating (projecting backward in time) an existing reference map, wherein regions of change are delineated through bi-temporal change analysis and overlaid onto the reference map. However, this procedure itself creates challenges, because sliver patches can occur in cases where the boundaries of the change regions do not exactly match the land-cover patches in the reference map. In this paper, we describe how sliver patches can inadvertently be created through the backdating and updating of land-cover maps, and document their impact on the magnitude and trajectory of four popular landscape metrics: number of patches (NP), edge density (ED), mean patch size (MPS), and mean shape index (MSI). In our findings, sliver patches led to significant distortions in both the value and temporal behaviour of metrics. In backdated maps, these distortions caused metric trajectories to appear more conservative, suggesting lower rates of change for ED and inverse trajectories for NP, MPS and MSI. In updated maps, slivers caused metric trajectories to appear more extreme and exaggerated, suggesting higher rates of change for all four metrics. Our research underscores the need to eliminate sliver patches from any study dealing with multi-temporal LPA.

A Disturbance-Inventory Framework for Flexible and Reliable Landscape Monitoring

Remote sensing plays a key role in landscape monitoring, but our handling of these data in a multi-temporal time series is not yet fully developed. Of particular concern is the presence of spatial and thematic errors in independently created maps that distort measures of landscape pattern and constrain the reliability of change analysis. In addition, there is a need to incorporate continuous attributes of cover gradients for flexible map representations that support a variety of applications. In this paper, we present a framework for generating temporally and categorically dynamic land-cover maps that provide such a reliable and adaptable foundation. The centerpiece is a spatio-temporal disturbance-inventory database, created through semi-automated change detection and conditioned with boundary-matching procedures, which can be used to backdate and update both continuous and categorical reference maps. We demonstrate our approach using multi-annual Landsat imagery from a forested region in west-central Alberta, Canada, between the years 1998 and 2005.

Problems in remote sensing of landscapes and habitats

Wildlife habitat mapping strongly supports applications in natural resource management, environmental conservation, impacts of anthropogenic activity, perturbed ecosystem restoration, species-at-risk recovery and species inventory. Remote sensing has long been identified as a feasible and effective technology for large-area habitat mapping. However, existing and future uncertainties in remote sensing will definitely have a significant effect on the relevant scientific research. This article attempts to identify the current challenges and opportunities in remote sensing for large-area wildlife habitat mapping, and accordingly provide possible solutions and directions for further research.

Spectral, spatial, and geomorphometric variables for the remote sensing of slope processes☆

Abstract A combined spectral, spatial, and geomorphometric variable set was used to separate geomorphic surfaces at three scales, corresponding roughly to process domain (level I), landform (level II), and sublandform (level III). The best results were obtained using the combined data set, demonstrating two concepts: i) With the use of specialized processing techniques, digital variables can separate geomorphic surfaces at a variety of scales, and ii) considerable advantages can be gained by using multisource remote sensing data. Stepwise selection patterns indicated that geomorphometric variables were the most important contributors to discriminant functions at levels I and II. The reduced effectiveness of spectral and spatial measures reflects the highly variable nature of surface cover at the process domain and landform levels. At level III, variable contributions were more balanced, suggesting that high-quality landcover information from spectral and spatial variables are necessary for the consistent separation of sublandform units.

The influence of sampling density on geographically weighted regression: a case study using forest canopy height and optical data

Geographically weighted regression (GWR) extends the conventional ordinary least squares (OLS) regression technique by considering spatial nonstationarity in variable relationships and allowing the use of spatially varying coefficients in linear models. Previous forest studies have demonstrated the better performance of GWR compared to OLS when calibrated and validated at sampled locations where field measurements are collected. However, the use of GWR for remote-sensing applications requires generating estimates and evaluating the model performance for the large image scene, not just for sampled locations. In this study, we introduce GWR to estimate forest canopy height using high spatial resolution Quickbird (QB) imagery and evaluate the influence of sampling density on GWR. We also examine four commonly used spatial analysis techniques – OLS, inverse distance weighting (IDW), ordinary kriging (OK) and cokriging (COK) – and compare their performance with that using GWR. Results show that (i) GWR outperformed OLS at all sampling densities; however, they produced similar results at low sampling densities, suggesting that GWR may not produce significantly better results than OLS in remote-sensing operational applications where only a small number of field data are collected. (ii) The performance of GWR was better than those of IDW, OK and COK at most sampling densities. Among the spatial interpolation techniques we examined, IDW was the best to estimate the canopy height at most densities, while COK outperformed OK only marginally and produced larger canopy height estimation errors than both IDW and GWR. (iii) GWR had the advantage of generating canopy height estimation maps with more accurate estimates than OLS, and it preserved patterns of geographic features better than IDW, OK or COK.

A Conceptual Model for Multi-Temporal Landscape Monitoring in an Object-Based Environment

Remote sensing plays a critical role in contemporary monitoring programs, but our strategies for processing these data using automated procedures are not always reliable. In particular, the task of separating real from spurious changes remains problematic, especially in an object-based environment where differential errors in classification quality, spatial registration, scene illumination, resolution, and object delineation have forced some operators to adopt labor-intensive visual-interpretation strategies, or employ manual interaction on an object-by-object basis. In this paper, we present an updated summary of our new disturbance-inventory approach to land-cover monitoring that combines object-based classification and change-detection strategies with boundary-conditioning routines designed to maximize the spatial and thematic integrity of the finished products. With this approach, the final maps are only altered in regions of confirmed change, and spurious gaps, slivers, stretches, and encroachments are avoided. The approach constitutes an innovative, efficient, and transparent framework that can handle all the basic landscape dynamics, including feature appearance, disappearance, succession, expansion, and shrinkage, without the need for manual editing.

Topographic dependence of synthetic aperture radar imagery

Abstract The increasing availability of synthetic aperture radar (SAR) remote-sensing imagery for earth-science applications creates the need for reliable computer methods to improve the relationships between SAR observations and the Earths abiotic, biotic, and cultural resources. In this paper, the topographic effect on aerial and satellite SAR imagery is quantified and corrected using software modified from an earlier normalized-cosine package written for use with optical/infrared remote-sensing imagery. The basic idea is that the incidence angle and forest canopy interactions with the radar beam can be estimated using a digital elevation model (DEM) and near-coincident observations in the red and near-infrared portions of the spectrum. Four different study areas in Canada and three different types of SAR imagery are used to illustrate the topographic dependence, and the degree of accuracy that can be expected, following the application of these relatively simple radiometric corrections.

Disturbance capture and attribution through the integration of Landsat and IRS-1C imagery

A primary activity required to support sustainable forest management is the detection and mitigation of forest disturbances. These disturbances can be planned, through urbanization and harvesting, or unplanned, through insect infestations or fire. Detection and characterization of disturbance types are important, as different disturbances have different ecological effects and may require unique managerial responses. As such, it is necessary for forest managers to have as complete and current information as possible to support decision making. In this study, we developed a framework to automatically detect and label disturbances derived from remotely sensed images. Disturbances were detected through traditional image differencing of medium-resolution imagery (Landsat-7 Enhanced Thematic Mapper Plus (ETM+), resampled to 30 m) but were refined and augmented through comparison with edge features extracted from high spatial resolution satellite imagery (Indian Remote Sensing (IRS) satellite 1C panchromatic imagery, resampled to 5 m). By incorporating spectral information, derived composite band values (tasselled cap transformations), spatial and contextual information, and secondary datasets, we were able to capture and label disturbance features with a high level of overall agreement (91%). Areal features, such as harvest areas, are captured and labelled more reliably than linear features such as roads, with 92% and 72% agreement when compared with control data, respectively. By incorporating rule-based disturbance attribution with remote sensing change detection, we envision the update of land cover databases with reduced human intervention, aiding more rapid data integration and opportunities for timely managerial responses.

Mapping the distribution of whitebark pine (Pinus albicaulis) in Waterton Lakes National Park using logistic regression and classification tree analysis

Accurate spatial information on the distribution of whitebark pine, a keystone species in alpine environments across western Canada, is critical for the planning of conservation activities designed to ameliorate the damaging effects of blister rust, mountain pine beetle, and interspecific competition. We compared classification tree analysis and logistic regression analysis to explore their relative abilities to model whitebark pine presence and absence with medium-spatial-resolution satellite and topographic variables across a complex study site in Waterton Lakes National Park, Alberta. Both techniques were found to be effective, generating map products of roughly equal thematic quality (91% overall accuracy; kappa = 0.76). However, the logistic model was valuable in its ability to predict ratio-level probability surfaces, whereas the classification tree approach was simpler, faster, and found to generate a slightly more balanced model from an individual class accuracy perspective. End users selecting between the two techniques should make choices that balance flexibility with simplicity while always taking care to exercise sound modeling practices.