Improved landscape ecology metrics for modeling, monitoring, and assessment of ecosystems with remote sensing

The goal of this dissertation was to develop improved metrics that show predictable or independent responses to characteristic variation in remote sensing data, specifically spatial resolution, while showing most sensitive responses to actual changes in landscape pattern. To achieve this goal two objectives were accomplished. The first objective was to develop landscape metrics that were sensitive to changes in fragmentation and patch shape complexity along various predictable gradients of change. The second objective was to develop landscape metrics that were insensitive or predictable with changes in spatial resolution.;The improved metrics developed were the Patch-Per-Unit area metric (PPU) and the Square Pixel metric (SqP). PPU measures the degree of fragmentation of patches on a landscape. SqP measures the shape complexity of patches on a landscape. These two metrics were analyzed and compared to two traditional metrics for fragmentation and patch shape complexity: Contagion and Fractal Dimension. The metrics were applied to four study sites which exhibited one of three predictable gradients of change: a spatial horizontal gradient; a vertical gradient; and a temporal gradient.;For all four study sites both PPU and SqP performed as predicted exhibiting the ability to distinguish between landcover types and landscape changes, unlike Contagion and Fractal Dimension. The PPU and SqP metrics were also found to be predictable with spatial resolution. The PPU metric exhibited a negative log log correlation with spatial resolution. Regression analysis of PPU for the log transformations yielded ;This dissertation has demonstrated the need for increased research on the integration of remote sensing, geographic information systems, and landscape ecology metrics for modeling, monitoring, and assessment of ecosystems. The long term value of such research will be realized as we employ these technologies and methods to achieve an improved understanding of the world in which we live.