Estimation of DBH at Forest Stand Level Based on Multi-Parameters and Generalized Regression Neural Network

The diameter at breast height (DBH) is an important factor used to estimate important forestry indices like forest growing stock, basal area, biomass, and carbon stock. The traditional DBH ground surveys are time-consuming, labor-intensive, and expensive. To reduce the traditional ground surveys, this study focused on the prediction of unknown DBH in forest stands using existing measured data. As a comparison, the tree age was first used as the only independent variable in establishing 13 kinds of empirical models to fit the relationship between the age and DBH of the forest subcompartments and predict DBH growth. Second, the initial independent variables were extended to 19 parameters, including 8 ecological and biological factors and 11 remote sensing factors. By introducing the Spearman correlation analysis, the independent variable parameters were dimension-reduced to satisfy very significant conditions (p ≤ 0.01) and a relatively large correlation coefficient (r ≥ 0.1). Finally, the remaining independent variables were involved in the modeling and prediction of DBH using a multivariate linear regression (MLR) model and generalized regression neural network (GRNN) model. The (root-mean-squared errors) RMSEs of MLR and GRNN were 1.9976 cm and 1.9655 cm, respectively, and the R2 were 0.6459 and 0.6574 respectively, which were much better than the values for the 13 traditional empirical age–DBH models. The use of comprehensive factors is beneficial to improving the prediction accuracy of both the MLR and GRNN models. Regardless of whether remote sensing image factors were included, the experimental results produced by GRNN were better than MLR. By synthetically introducing ecological, biological, and remote sensing factors, GRNN produced the best results with 1.4688 cm in mean absolute error (MAE), 13.78% in MAPE, 1.9655 cm for the RMSE, 0.6574 for the R2, and 0.0810 for the Theil’s inequality coefficient (TIC), respectively. For modeling and prediction based on more complex tree species and a wider range of samples, GRNN is a desirable model with strong generalizability.

[1]  Zhang Shu-hui.  Comparative analysis on the mapping with aerial image and satellite image , 2008 .

[2]  G. Scrinzi,et al.  Development of a neural network model to update forest distribution data for managed alpine stands , 2007 .

[3]  Masami Kaneko,et al.  Comparison of image data acquired with AVHRR, MODIS, ETM+ and ASTER over Hokkaido, Japan , 2003 .

[4]  R. I. Lumbres,et al.  Model fitting and validation of six height–DBH equations for Pinus kesiya Royle ex Gordon in Benguet Province, Philippines , 2013 .

[5]  Luming Fang,et al.  A Levenberg–Marquardt Backpropagation Neural Network for Predicting Forest Growing Stock Based on the Least-Squares Equation Fitting Parameters , 2018, Forests.

[6]  Donald F. Specht,et al.  A general regression neural network , 1991, IEEE Trans. Neural Networks.

[7]  Giovanni Correia Vieira,et al.  Prognoses of diameter and height of trees of eucalyptus using artificial intelligence. , 2018, The Science of the total environment.

[8]  Environmental factors and underlying mechanisms of tree community assemblages of pine-oak mixed forests in the Qinling Mountains, China , 2016, Journal of Plant Biology.

[10]  Daniel Henrique Breda Binoti,et al.  Estimation of inside-bark diameter and heartwood diameter for Tectona grandis Linn. trees using artificial neural networks , 2011, European Journal of Forest Research.

[11]  Annabel Porté,et al.  Modelling mixed forest growth: a review of models for forest management , 2002 .

[12]  Dasheng Wu,et al.  Dynamic Estimation of Forest Volume Based on Multi-Source Data and Neural Network Model , 2015 .

[13]  Maria J. Diamantopoulou,et al.  Artificial neural networks as an alternative tool in pine bark volume estimation , 2005 .

[14]  G. Mallinis,et al.  Assessing the relationships between growing stock volume and Sentinel-2 imagery in a Mediterranean forest ecosystem , 2017 .

[15]  Hong Wei,et al.  A study on simulating predictive model of mean DBH for bamboo stands. , 2000 .

[16]  L. Yu,et al.  Impact of climate on individual tree radial growth based on generalized additive model , 2014 .

[17]  Claudia Notarnicola,et al.  Potential of ALOS2 and NDVI to Estimate Forest Above-Ground Biomass, and Comparison with Lidar-Derived Estimates , 2016, Remote. Sens..

[18]  Changhui Peng,et al.  TRIPLEX: a generic hybrid model for predicting forest growth and carbon and nitrogen dynamics , 2002 .

[19]  R J Mullins,et al.  Application of an artificial neural network. , 2001, The Journal of trauma.

[20]  M. Pach.  The influence of admixture and co-dominant species on the height and DBH of silver fir (Abies alba Mill.) and on the growing stock of fir stands in the Carpathian Forest-Natural Region , 2010 .

[21]  F. Bravo,et al.  Multinomial logit estimation of a diameter growth matrix model of two Mediterranean pine species in Spain , 2011, Annals of Forest Science.

[22]  Pablo J. Zarco-Tejada,et al.  Tree height quantification using very high resolution imagery acquired from an unmanned aerial vehicle (UAV) and automatic 3D photo-reconstruction methods , 2014 .

[23]  Christiane Schmullius,et al.  Growing stock volume estimation from L-band ALOS PALSAR polarimetric coherence in Siberian forest , 2014 .

[24]  Venkat Lakshmi,et al.  Comparison of Normalized Difference Vegetation Index Derived from Landsat, MODIS, and AVHRR for the Mesopotamian Marshes Between 2002 and 2018 , 2019, Remote. Sens..

[25]  J. Louw,et al.  Spatiotemporally interactive growth dynamics in selected South African forests: Edaphoclimatic environment, crowding and climate effects , 2011 .

[26]  M. Diamantopoulou,et al.  Estimating tree bole volume using artificial neural network models for four species in Turkey. , 2010, Journal of environmental management.

[27]  Jingyun Fang,et al.  Effects of size, neighbors, and site condition on tree growth in a subtropical evergreen and deciduous broad‐leaved mixed forest, China , 2015, Ecology and evolution.

[28]  Jianzhou Wang,et al.  A hybrid model for PM₂.₅ forecasting based on ensemble empirical mode decomposition and a general regression neural network. , 2014, The Science of the total environment.