Abstract: To clarify the allometric growth patterns and spatial differentiation characteristics of the height-diameter ratio(HDR)of Delonix regia in Shenzhen.A total of 33071 healthy individuals of D.regia across Shenzhen were selected as research materials.Multiple model fitting,standard major axis(SMA)regression and coupled multi-statistical framework approaches were applied to select the optimal height-diameter(H-D)allometric growth model,and quantify the influences of site factors including green space category,land surface cover type and terrain on tree HDR.The results revealed evident spatial heterogeneity in HDR distribution of Shenzhen-grown D.regia,with prominent divergences in growth strategies across various green space types.The Logistic function was identified as the best-performing H-D model,outperforming the power function,Weibull,Richards and logarithmic models in fitting goodness.Site factors exerted significant modulatory controls on the allometric growth of D.regia HDR.Terrain attributes and green space type acted as the primary driving factors,whereas impervious surfaces imposed no meaningful influence on HDR.Coupling multiple statistical frameworks greatly improved model fitting precision and eliminated defects arising from reliance on a single statistical model.This study systematically elucidates the allometric patterns and spatial differentiation mechanisms of HDR in Shenzhen’s D.regia population.It delivers quantitative references for differentiated maintenance schemes and structural safety regulation of D.regia,and offers methodological support for growth modeling and urban management of typical ornamental arbors in South subtropical cities.