Abstract The chemical composition of poplar wood showed strongly influences on pulping efficiency and paper performance, yet systematic predictive frameworks for clone evaluation remain limited. In this study, eight representative poplar clones widely cultivated in China were subjected to Kraft pulping and sheet preparation. Comprehensive characterization was performed including wood chemical composition, pulp yield, Kappa number, viscosity, and paper properties. 108 poplar yielded the highest burst index (4.79 kPa·m 2 ·g −1 ) and tear index (18.12 mN·m 2 ·g −1 ). The Ridge-regression models predicted pulp yield, Kappa number, viscosity and paper strength with R of 0.90–0.99, RMSE ≤ 0.33 and MAE ≤ 1.20. From the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) analysis, 108 poplar was the best overall clone (NCi = 0.230). To elucidate the interrelationships among wood chemistry, pulp behavior, and paper properties, twelve variables across three categories were analyzed. The Ridge regression was employed to overcome multicollinearity and achieved reliable prediction (R = 0.90–0.99, with low RMSE and MAE). Multi-criteria ranking using the TOPSIS further identified the 108 poplar as the best-performing clone overall. These findings established an effective integration of predictive modeling and decision analysis, offering a practical framework for clone selection and optimized utilization of poplar resources in the pulp and paper industry.
Meng et al. (Tue,) studied this question.