Research overview: local convex hull (LoCoH) home-range estimators

Local convex hull methods, known as LoCoH, build a home range from many small convex polygons constructed around each location and its nearest neighbors, then merge them into a utilization distribution. This overview explains how, unlike kernel methods, LoCoH can sharply delineate hard boundaries such as cliffs, lake edges, or unused interiors, producing ranges with realistic holes and edges. Variants differ in how neighbors are selected, by fixed number, fixed radius, or adaptive criteria. For cats inhabiting topographically constrained landscapes, where movement is bounded by terrain, LoCoH can outperform smoother estimators. The overview also notes its sensitivity to parameter choices and outliers. LoCoH represents one branch of the broader effort to make home-range estimates respect real landscape structure.