Comparative study of the possibility estimation of some structural quantitative attributes of Caspian forests using Radar and integrating Radar and Lidar data

The purpose of this study was to compare the estimation of the structural attributes of stand volume, basal area, and tree stem density per hectare of the Caspian forests using Radar data and integration of Radar and Lidar data in some parts of the district I and II the ShastKalateh forest in the Golestan province. Forest structural data were measured and computed from 307 circular plots. The required pre-processing and processing was performed using raw data of Radar (2009) and Lidar (2011), and the corresponding values of sample plots were extracted on all Radar and Lidar derived indices. The modeling was performed using extracted Radar features as individual and also using Lidar and Radar extracted features as integrated with the non-parametric random forest algorithm in 75% of samples. The modeling validity was performed using 25% of the remained samples by absolute and relative root mean square error (RMSe) and Bias. The percentage RMSe and the Bias values using Radar data were obtained form stand volume (44.09% and -0.99%), basal area per hectare (35.72% and -3.15%) and tree stem density per hectare (42.73% and 3.52%), respectively, and using the integration of Radar and Lidar data for stand volume (37.23% and 0.76%), basal area per hectare (31.37% and -3.14%), and tree density per hectare (36.44% and 0.95%). The results showed that the integration of Radar and Lidar data could improve the estimates, especially in the stand volume, compared to using Radar data as individually.