Landscape Permeability developed by TerrAdapt for the Washington Habitat Connectivity Action Plan. This model assessed the degree to which each pixel was connected to its neighbors, highlighting opportunities for local movements regardless of the presence of defined cores.

The least-cost corridor models represented the most efficient routes between core areas given the resistance in the landscape. However, the distribution of corridors is highly sensitive to the location of the core areas, and habitat outside of cores is not represented as a source for movement in core-corridor models. As a complementary alternative method to map connectivity that does not depend on first mapping core areas, we used permeability models. This method assesses the degree to which a pixel is connected to its neighbors in a probabilistic way. We used the resistance model for each ecosystem as the input after rescaling it to range from 0 to 1 as resistance decreased from 150 to 1. Thus, pixels with the minimum resistance (1) were always included in the calculation. For pixels with a probability of less than 1, they were compared to a random number per pixel. If the probability exceeded the random number, the pixel was retained, otherwise, it was dropped. Then, we calculate the patch size of all retained pixels that were contiguous. Finally, we summarized the patch sizes across 300 replicate runs, each with a different randomly generated probability surface. The result represents the spectrum of areas that are always connected (contiguous patches of the lowest resistance value) and areas that are increasingly less likely to be connected due to the resistance in the landscape.

Landscape Permeability developed by TerrAdapt for the Washington Habitat Connectivity Action Plan. This model assessed the degree to which each pixel was connected to its neighbors, highlighting opportunities for local movements regardless of the presence of defined cores.

The least-cost corridor models represented the most efficient routes between core areas given the resistance in the landscape. However, the distribution of corridors is highly sensitive to the location of the core areas, and habitat outside of cores is not represented as a source for movement in core-corridor models. As a complementary alternative method to map connectivity that does not depend on first mapping core areas, we used permeability models. This method assesses the degree to which a pixel is connected to its neighbors in a probabilistic way. We used the resistance model for each ecosystem as the input after rescaling it to range from 0 to 1 as resistance decreased from 150 to 1. Thus, pixels with the minimum resistance (1) were always included in the calculation. For pixels with a probability of less than 1, they were compared to a random number per pixel. If the probability exceeded the random number, the pixel was retained, otherwise, it was dropped. Then, we calculate the patch size of all retained pixels that were contiguous. Finally, we summarized the patch sizes across 300 replicate runs, each with a different randomly generated probability surface. The result represents the spectrum of areas that are always connected (contiguous patches of the lowest resistance value) and areas that are increasingly less likely to be connected due to the resistance in the landscape.