SpatialBoundaries.jl: edge detection using spatial wombling

Abstract

Spatial wombling is an approach for detecting edges within a defined two‐dimensional landscape. This is achieved by calculating the rate and direction of change through the interpolation of points. This not only gives an approximation as to the shape of the landscape but can also be used to identify candidate boundaries cells that delimit a shift from one state to another within the landscape. Here we introduce the SpatialBoundaries.jl package for Julia, which has been developed to implement the wombling algorithm for datasets that are spatially referenced for both uniformly or randomly sampled landscapes. From a practical perspective, the wombling functionality allow the user to answer two questions: how much and in which direction does the variable of interest change within the landscape? Whereas the boundaries functionality identifies candidate boundary cells. We conclude by providing a working example of the package using the various woody plant layers for Britain and Ireland from the EarthEnv database.