A study on the effects of tree spatial distribution on the accuracy of canopy clumping index retrieved by remote sensing

Clumping Index (CI) is a key parameter for estimating the leaf area index (LAI) in forest canopy and distinguishing between "sunlit foliage" and "shaded foliage" in forest canopy. It is of great significance for studying total primary productivity and evapotranspiration of forests. The method of canopy CI retrieved by multi-angle satellite data has been widely used. However, it is unclear whether the method of using the normalized difference between hotspot and darkspot (NDHD) to retrieve CI by remote sensing is suitable for forests with different tree spatial distributions (TSDs). In this study, the accuracy of canopy CI retrieved by remote sensing under different TSDs was investigated by using 1 km × 1 km within the Saihanba Forest Farm as the study area, and using the DART model to calculate the multi-angle gap fractions and reflectance under different TSDs. The results showed that: (1) The CI results of retrieval of the plots with clustered TSDs were generally larger than the CI results of validation, while the CI results of retrieval of the plots with regular TSDs were generally smaller than the CI results of validation. (2) The accuracies of the CI results of retrieval were small for plots with strong clustered TSDs or strong regular TSDs. The results of this study play an important role in understanding the influence of TSDs in canopy CI retrieved by remote sensing, as well as in assessing forest productivity and studying carbon cycling in terrestrial ecosystems.