Fusing multidimensional hierarchical information into finer spatial landscape metrics

Area of Science:

• Ecology
• Geospatial Analysis
• Landscape Ecology

Background:

• Understanding landscape pattern effects on ecological processes is crucial.
• Traditional landscape pattern indicators (LPIs) primarily focus on horizontal structures, neglecting vertical relationships.
• This oversight can lead to biases and reduced accuracy in ecological assessments.

Purpose of the Study:

• To develop a multidimensional fusion method for LPIs that integrates hierarchy theory and information entropy.
• To address the limitations of existing LPIs by incorporating vertical landscape structure information.
• To enhance the spatial representational ability and accuracy of landscape pattern analysis.

Main Methods:

• Established a general fusion formula for simple LPIs using two-grade land use data.
• Applied hierarchy theory and information entropy to quantify vertical landscape structure.
• Derived three fusion landscape pattern indicators (FLIs) and validated them with a case study.

Main Results:

• The fusion method successfully captures fine spatial structure information.
• Significant differences between FLIs and traditional LPIs were observed in areas with complex vertical structures (e.g., ecotones).
• FLIs demonstrate superior spatial representational ability, retaining both coarse and fine-scale land use data details.

Conclusions:

• The proposed fusion method enhances the accuracy and detail of landscape pattern analysis.
• FLIs offer improved spatial representational ability compared to traditional LPIs.
• This approach is suitable for broader applications with various LPIs and data dimensions, potentially improving ecological models.