Figure 9 The Moran scatter plot and spatial autocorrelation clustering map of Shannon-Wiener index.

3.2.4 The Correlation between Vegetation Indices and Vegetation Diversity

Significant correlations between vegetation and diversity indices were observed at both the 10m × 10m and 30m × 30m scales (Figure 10), indicating the possibility of establishing diversity assessment models. At the 10m scale, all six vegetation indices, except the Margalef index with no correlation to NDVI and SAVI, exhibited significant relationships with the four diversity indices. Notably, the correlation coefficients between Pielou, Simpson, Shannon, and DVI, as well as MSAVI, consistently exceeded 0.54 (Table 3). However, at the 30m scale, not all correlations observed at the smaller scale were present. NDVI and MSAVI showed weak or no correlation with the four diversity indices, while Pielou, Simpson, and Shannon exhibited significant positive correlations with the remaining vegetation indices. Overall, whether at the 10m × 10m or 30m × 30m scale, DVI showed higher correlation coefficients with Shannon, with values of 0.617 (R2=0.38, P<0.01) and 0.725 (R2=0.53, P<0.01), respectively. The diversity assessment model using DVI as the independent variable and Shannon as the dependent variable can be represented as: y=1.667x+0.8364 (10m × 10m) and y=2.3452x+1.0473 (30m×30m), explaining 38% and 53% of the variation, respectively.