David W. Andison

The choice of window size in approximating topographic surfaces from Digital Elevation Models

Quantitative surface analysis through quadratic modelling of Digital Elevation Models (DEMs) is a promising tool for automatically describing the physical environment in ecological studies of terrestrial landscapes. Fundamental topographic variables such as slope, aspect, plan and profile curvature can be simply calculated from the parameters of a conic equation fitted to a DEM window through the least-squares method. The scale of the analysis, defined by the size of the DEM window used to fit the conic equation, affects both the estimated value of the topographic variables and the propagation of elevation errors to derived topographic variables. The least-squares method is amenable to the analytical treatment of the propagation of elevation errors to the derived topographical variables. A general analytical model of error propagation is presented that accounts for the effects of window size and of spatial autocorrelation in elevation errors. The method is based on the Taylor approximation of the least-square fitting equation and allows for the presence of stationary autocorrelation in the elevation errors. In numerical simulations with DEMs from British Columbia, Canada, it is shown that increasing the size of evaluation windows effectively reduces the propagation of elevation errors to the derived topographic variables. However, this was obtained at the expense of topographic detail. A methodology is proposed to evaluate quantitatively the loss of topographic detail through a χ 2-test of the corrected residuals in the immediate neighbourhood of the evaluation point. This methodology, in combination with the analytical model of error propagation, can be used to select the scale or range of scales at which to calculate topographic variables from a DEM.

Evolution of Canada’s Boreal Forest Spatial Patterns as Seen from Space

Understanding the development of landscape patterns over broad spatial and temporal scales is a major contribution to ecological sciences and is a critical area of research for forested land management. Boreal forests represent an excellent case study for such research because these forests have undergone significant changes over recent decades. We analyzed the temporal trends of four widely-used landscape pattern indices for boreal forests of Canada: forest cover, largest forest patch index, forest edge density, and core (interior) forest cover. The indices were computed over landscape extents ranging from 5,000 ha (n = 18,185) to 50,000 ha (n = 1,662) and across nine major ecozones of Canada. We used 26 years of Landsat satellite imagery to derive annualized trends of the landscape pattern indices. The largest declines in forest cover, largest forest patch index, and core forest cover were observed in the Boreal Shield, Boreal Plain, and Boreal Cordillera ecozones. Forest edge density increased at all landscape extents for all ecozones. Rapidly changing landscapes, defined as the 90th percentile of forest cover change, were among the most forested initially and were characterized by four times greater decrease in largest forest patch index, three times greater increase in forest edge density, and four times greater decrease in core forest cover compared with all 50,000 ha landscapes. Moreover, approximately 18% of all 50,000 ha landscapes did not change due to a lack of disturbance. The pattern database results provide important context for forest management agencies committed to implementing ecosystem-based management strategies.

Predicting post-fire canopy mortality in the boreal forest from dNBR derived from time series of Landsat data

Regulatory and certification agencies need historical fire pattern information across the Canadian boreal forest to support natural disturbance-based management. Landsat-derived spectral indices have been used extensively to map burn severity in North America. However, satellite-derived burn severity is difficult to define and quantify, and relies heavily on ground truth data for validation, which hinders fire pattern analysis over broad scales. It is therefore critical to translate burn severity estimates into more quantifiable measurements of post-fire conditions and to provide more cost-effective methods to derive validation data. We assessed the degree to which Landsat-derived indices and ancillary data can be used to classify canopy mortality for 10 fires in the boreal forest of Alberta and Saskatchewan, Canada. Models based on two and three mortality classes had overall accuracies of 91 and 72% respectively. The three-level classification has more utility for resource management, with improved accuracy at predicting unburned and complete canopy mortality classes (93 and 66%), but is relatively inaccurate for the partial mortality class (56%). The results presented here can be used to assess the suitability of different canopy mortality models for forest fire management goals, to help provide objective, consistent and cost-effective results to analyse historical fire patterns across the Canadian boreal forest.

Quantifying local fire regimes using the Landsat data-archive: a conceptual framework to derive detailed fire pattern metrics from pixel-level information

ABSTRACT Free and open access to the Landsat archive has enabled the detection and delineation of an unprecedented number of fire events across the globe. Despite the availability and potential of these data, few studies have analysed residual vegetation patterns and/or partial mortality of fire across the Canadian boreal forest, and those available, are either incomplete or inaccurate. Further, they all differ in the methods and spatial language, which makes it difficult for managers to interpret fire patterns over large areas. There is an urgent need for methods to help unify fire pattern observations across the Canadian boreal forest. This study explores the capacity of the Landsat data archive when coupled with a recently developed fire mapping approach and a robust spatial language to characterize and compare tree mortality patterns across the boreal plains ecozone, Canada. With 507 fires 2.5 Mha mapped, this study represents the most comprehensive analysis of mortality patterns for study area. Summaries from this demonstration generated an accurate characterization of the fire patterns the various ecoregions based on seven key fire metrics. The comparison between ecoregions revealed differences in the amount of residual vegetation, which in turn suggested various climate, topography and/or vegetation ecosystem drivers.