Human Footprint

Human footprint models quantify the degree to which anthropogenic impacts to native environments diminish or degrade habitat suitability. Areas of high human footprint (e.g., dense urban areas, intensive agricultural areas, or surface mines) support very little if any native species while areas of low human footprint provide abundant native habitat and allow for natural processes to proceed without influence from human activities.

We follow methods similar to Theobald et al. (2020), which quantifies the magnitude of several anthropogenic impacts such as various classes of roads (interstate highways, primary roads, secondary roads, local roads), various classes of agriculture (irrigated row crops, dryland row crops, fallow, orchard, and pasture), timber harvest, population density, energy transmission lines, solar installations, wind turbines, canals, dams, quarries, and electrical power stations. Each impact also has a distance over which the impacts extend from their source, reflecting processes such as the spread of noise, light, invasive species, domestic animals, pollution, and other impacts that radiate outward from impacted areas. A fuzzy sum is used to combine impacts into a single human footprint model scaled from 0 to 1 (higher values indicate greater magnitude of human impacts). The locations of these impacts are derived from several sources, including TerrAdapt's landcover and disturbance models and OpenStreetMap.

Empirical validation of human footprint models is not strictly possible because they are expert based, but attempts have been made to relate model parameters to the degree to which a pixel is modified by human activities based on field data. The parameters we use were validated this way in Theobald (2013).