Quantification of the Relationship Between the Height – Diameter of (Pinus brutia Ten), Calabrian Pine Trees in Four Different Microsites in Duhok Governorate. Kurdistan Region, Iraq

Abstract

A sample of 120 trees was purposely selected from four different microsites (Behere, Swaratoka, Zawita, and Semel), (30 trees from each microsite) in the Duhok governorate. 100 of which were used for model calibration and the rest were used for validation of the selected regression equation. This study consisted of two main parts. In the first part ,In the beginning, a scatter diagram was conducted for each microsite to detect the type of relationship between height and diameter at breast, which in turn will decrease the number of regression models that will be tested. Accordingly, 4 regression equations were developed for each of the four microsites separately. These models can be used to see how the ratio of height diameter in each of the studied sites is inter-correlated and which of them is the most appropriate for producing Calabrian pine trees. In the second part, all microsites were then treated as one sample for estimating the parameters of 25 regression models using Excel and Statographic packageThe developed regression models underwent a screening process in order to find the most appropriate one to be used for the prediction of the height of Calabrian pine grown in the four mentioned microsites in Duhok governorate. Many measures of precision among them coefficient of determination, Bias%, Mean absolute error, Ohtomo’s unbiased test, Furnival Index, and AIC criterion were used for testing the performance of the developed equations in the prediction of the height. At last, the equation was selected, as the best regression model. This equation shows that there is a linear relationship between and a curvilinear relationship between D and H. The study showed that the height–diameter ratio was highest in Swaratoka and was the most appropriate microsite among the rest, followed by Behere, then Zawita and Semel came in last place.

Comparison Between Linear and Non-Linear Regression Models in the Prediction of the Height of Gall Oak Trees (Quercus infectoria Olive.) in Duhok Governorate

“…(Nearest value to one)  Measure of precision for regression models having different forms of the dependent variable. The coefficient of determination can be applied to test the performance ability of equations with a homogeneous form of the dependent variable [8,[26][27][28][29][30][31][32]. If the competed regression models appeared in different forms then the following criteria can be used:…”

“…The best equation is that having the values of m, and k closest to zero and one respectively. [31] proposed a new version of Ohtomo's test in which the values of both m and k in addition to the value of the coefficient of determination and MAE is taken into account in calculating the value of the new version [ ] as follow:…”

“…According to [8,24,26,28,27,29,30,31,32] It is not possible to use of the ability of regression models unless their dependent variables appear in the same form.…”

“…As mentioned earlier that the most précised regression model is that having the lowest values of RMSE, MAE and Bias%, therefore equation 3 and equation 30 were also excluded as competitive candidates the remained equations were (2,21, and 26) which underwent both Furnival Index and modified Ohtomos test proposed by Salih 2021[31]. Table4.…”