Objective To select the optimal dominant height estimation method for natural Picea schrenkiana forests in the Tianshan Mountains, this study compared seven methods to provide a scientific basis for regional forest site quality evaluation.

Method Based on three 1 hm² sample plots in the Tianshan region of Xinjiang, seven stand dominant height estimation methods were constructed using the traditional method, modified largest tree method, small plot estimation, and U-estimation method as foundations, combined with two dominant tree selection strategies (largest DBH trees and tallest trees). The relationships among methods were analyzed through correlation and difference tests. Different sub-plot sizes and density gradients were established to determine the threshold ranges where dominant height estimation stabilized. Uncertainty was quantified using bias and stratified standard deviation, and correlation analysis with stand volume, mean DBH, biomass, and other factors was conducted to comprehensively select the optimal method.

Result  (1) Correlation coefficients among the different dominant height estimation methods ranged from 0.76 to 0.99, with significant differences detected among them. (2) The threshold plot area for stable dominant height estimation was 0.06–0.13 hm², and the stand density threshold was 500–1,000 trees/hm². (3) The modified largest tree method (5.99) and U-estimation method (6.02) using the tallest tree selection strategy produced the smallest uncertainty. (4) All dominant height estimation methods showed strong correlations with stand factors, and the modified largest tree method (tallest tree strategy) was identified as the optimal method for natural Picea schrenkiana forests.

Conclusion  The modified largest tree method (tallest tree strategy) identified in this study is suitable for dominant height estimation in natural Picea schrenkiana forests and can provide technical support for site quality evaluation of natural forests in this region.