A simulated inventory of forest cover undertaken at several scales. A: Hypothetical tree locations and crown extents; crowns are considered to be impermeable. B to F: the results of an inventory carried out with grain-size determined by a disc of diameter equal to the vertical bar at the bottom right of each window. For each location on the grid, the percent of the disc area within crowns is calculated. The cell is classified as forest if this mean exceeds 40%. As grain size increased, the inventory results change substantially At grain size F, the inventory is clearly missing features which one feels are important, but at grain size B, the concept of iforest'is not being g the entire area would be determined as forest, despite the fact that only 50%otthswvas ddressed; small intervals between trees are clearly part of the forest. If grain-size increased beyond area is occupied by tree crowns B E冷A simulated inventory of forest cover undertaken at several scales. A:Hypothetical tree locations and crown extents; crowns are considered to be impermeable. B to F: the results of an inventory carried out with grain-size determined by a disc of diameter equal to the vertical bar at the bottom right of each window. For each location on the grid, the percent of the disc area within crowns is calculated. The cell is classified as ‘forest’ if this mean exceeds 40%. As grain size increased, the inventory results change substantially.At grain size F, the inventory is clearly missing features which one feels are important, but at grain size B, the concept of ‘forest’ is not being addressed; small intervals between trees are clearly part of the forest. If grain-size was increased beyond F, the entire area would be determined as forest, despite the fact that only 50% of the area is occupied by tree crowns. A B C D E F