Development of stem diameter model using quantile regression

Turkish Journal of Forestry

Şerife Kalkanlı¹  Ramazan Ozçelik²  Onur Alkan² 

[1] ISPARTA UYGULAMALI BİLİMLER ÜNİVERSİTESİ, LİSANSÜSTÜ EĞİTİM ENSTİTÜSÜ, ORMAN MÜHENDİSLİĞİ (DR);Isparta Uygulamalı Bilimler Üniversitesi Orman fakültesi;

关键词: kantil regresyon;    kantil seti;    gövde çapı;    karaçam;    quantile regression;    quantile set;    stem diameter;    black pine;   

DOI  :  10.18182/tjf.955881

来源: DOAJ

【 摘 要 】

Tree volume estimates are one of the most important components of growth and yield models. Stem diameter models are one of the most modern approaches used for stem volume estimation. Different regression methods, especially the nonlinear least squares (ONLS) method, were used to develop stem diameter models. Recently, the quantile regression (QR) method has also been used in forestry applications for the development of taper models. In this study, a stem diameter model based on Max and Burkhart (1976) model was developed using ONLS and QR methods for natural black pine stands. For this purpose, sample tree data were obtained from two different black pine stands, and some of the data (group I) were used to develop the models and the rest of the data (group II) was used to test the models developed. In the study, the QR technique was used based on two different quantile sets (3QR and 5QR). The results were compared for the whole tree stem and ten different relative height classes using four different evaluation criteria. Evaluation statistics showed that both quantile regression models provided better results as compared to ONLS and 3QR model performed relatively better than 5QR. In conclusion, QR technique is an approach that can be used as an alternative to other regression techniques for stem diameter estimations.

【 授权许可】

Unknown