The application provides a
landscape pattern analysis suitable
granularity determination method and
system, comprising: generating a first multi-
granularity sequence according to a maximum area value rule (RMA) with a coarse step length from target area high-resolution
land cover or
land use vector data, and determining a
granularity upper limit threshold and a scale domain according to three conservation criteria of landscape composition, area and pattern form; generating a second multi-granularity sequence in the scale domain according to the RMA rule with a fine step length, and calculating a
scale effect sensitive
landscape pattern index of the second multi-granularity sequence; and performing
mutation analysis by using a non-parametric rank test (Mann-Kendall and Pettitt test), and taking the highest frequency effective
mutation point as the suitable granularity. The application solves the problems of traditional methods, such as ignoring
landscape pattern form loss, strong subjectivity of granularity judgment, data distribution dependence and the like by quantifying
information loss and non-parametric
mutation test, and improves the objectivity of granularity decision and the ecological application precision.